ASM Handbook, Volume 15, Casting (2008)

Citation preview

© 2008 ASM International. All Rights Reserved. ASM Handbook Volume 15: Casting (#05115G)

www.asminternational.org

W

ASM Handbook Volume 15 Casting Prepared under the direction of the ASM International Handbook Committee

Editorial Committee Srinath Viswanathan, University of Alabama, Chair Diran Apelian, Worcester Polytechnic Institute Raymond J. Donahue, Mercury Marine Babu DasGupta, National Science Foundation Michael Gywn, ATI Inc John L. Jorstad, J.L.J. Technologies, Inc Raymond W. Monroe, Steel Founders’ Society of America Mahi Sahoo, CANMET Materials Technology Laboratory Thomas E. Prucha, American Foundry Society Daniel Twarog, North American Die Casting Association Steve Lampman, Project Editor Charles Moosbrugger, Editor Eileen DeGuire, Editor Madrid Tramble, Senior Production Coordinator Ann Britton, Editorial Assistant Diane Whitelaw, Production Coordinator Kathryn Muldoon, Production Assistant Scott D. Henry, Senior Product Manager Bonnie R. Sanders, Manager of Production

Editorial Assistance Elizabeth Marquard Heather Lampman WPR Indexing Service

Materials Park, Ohio 44073-0002 www.asminternational.org

© 2008 ASM International. All Rights Reserved. ASM Handbook Volume 15: Casting (#05115G)

www.asminternational.org

Copyright # 2008 by ASM InternationalW All rights reserved No part of this book may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording, or otherwise, without the written permission of the copyright owner. First printing, December 2008

This book is a collective effort involving hundreds of technical specialists. It brings together a wealth of information from worldwide sources to help scientists, engineers, and technicians solve current and long-range problems. Great care is taken in the compilation and production of this Volume, but it should be made clear that NO WARRANTIES, EXPRESS OR IMPLIED, INCLUDING, WITHOUT LIMITATION, WARRANTIES OF MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE, ARE GIVEN IN CONNECTION WITH THIS PUBLICATION. Although this information is believed to be accurate by ASM, ASM cannot guarantee that favorable results will be obtained from the use of this publication alone. This publication is intended for use by persons having technical skill, at their sole discretion and risk. Since the conditions of product or material use are outside of ASM’s control, ASM assumes no liability or obligation in connection with any use of this information. No claim of any kind, whether as to products or information in this publication, and whether or not based on negligence, shall be greater in amount than the purchase price of this product or publication in respect of which damages are claimed. THE REMEDY HEREBY PROVIDED SHALL BE THE EXCLUSIVE AND SOLE REMEDY OF BUYER, AND IN NO EVENT SHALL EITHER PARTY BE LIABLE FOR SPECIAL, INDIRECT OR CONSEQUENTIAL DAMAGES WHETHER OR NOT CAUSED BY OR RESULTING FROM THE NEGLIGENCE OF SUCH PARTY. As with any material, evaluation of the material under end-use conditions prior to specification is essential. Therefore, specific testing under actual conditions is recommended. Nothing contained in this book shall be construed as a grant of any right of manufacture, sale, use, or reproduction, in connection with any method, process, apparatus, product, composition, or system, whether or not covered by letters patent, copyright, or trademark, and nothing contained in this book shall be construed as a defense against any alleged infringement of letters patent, copyright, or trademark, or as a defense against liability for such infringement. Comments, criticisms, and suggestions are invited, and should be forwarded to ASM International. Library of Congress Cataloging-in-Publication Data ASM International ASM Handbook Includes bibliographical references and indexes Contents: v.1. Properties and selection—irons, steels, and high-performance alloys—v.2. Properties and selection—nonferrous alloys and special-purpose materials—[etc.]—v.21. Composites 1. Metals—Handbooks, manuals, etc. 2. Metal-work—Handbooks, manuals, etc. I. ASM International. Handbook Committee. II. Metals Handbook. TA459.M43 1990 620.1’6 90-115 SAN: 204-7586 ISBN-13: 978-0-87170-711-6 ISBN-10: 0-87170-711-X

ASM InternationalW Materials Park, OH 44073-0002 www.asminternational.org Printed in the United States of America Multiple copy reprints of individual articles are available from Technical Department, ASM International.

© 2008 ASM International. All Rights Reserved. ASM Handbook Volume 15: Casting (#05115G)

www.asminternational.org

Dedicated to the memory of JOSEPH R. DAVIS (1954–2007) Editor, Metals Handbook, 1982–1991

© 2008 ASM International. All Rights Reserved. ASM Handbook Volume 15: Casting (#05115G)

www.asminternational.org

Foreword

In this revision of ASM HandbookW Volume 15 on casting, ASM International is indebted to the volunteer efforts of the Volume 15 Editorial Committee and over 150 participants who helped as authors or reviewers. Their professional commitment and efforts represent a continuing devotion to the practice of metalcasting and the publication of peer consensus information on it. Special thanks are extended to Srinath Viswanathan, University of Alabama, for recruiting an outstanding Editorial Committee with Diran Apelian, Worcester Polytechnic Institute; Babu DasGupta, National Science Foundation, Raymond J. Donahue, Mercury Marine; Michael Gywn, ATI Inc.; John L. Jorstad, J.L.J. Technologies, Inc.; Raymond W. Monroe, Steel Founders’ Society of America; Thomas E. Prucha, American Foundry Society; Kumar Sadayappan and Mahi Sahoo, CANMET Materials Technology Laboratory; Edward S. Szekeres, Casting Consultants Incorporated; and Daniel Twarog, North American Die Casting Association. We thank them and the other contributors for this publication.

Dianne Chong President ASM International Stanley C. Theobald Managing Director ASM International

iv

© 2008 ASM International. All Rights Reserved. ASM Handbook Volume 15: Casting (#05115G)

www.asminternational.org

Policy on Units of Measure

By a resolution of its Board of Trustees, ASM International has adopted the practice of publishing data in both metric and customary U.S. units of measure. In preparing this Handbook, the editors have attempted to present data in metric units based primarily on Syste`me International d’Unite´s (SI), with secondary mention of the corresponding values in customary U.S. units. The decision to use SI as the primary system of units was based on the aforementioned resolution of the Board of Trustees and the widespread use of metric units throughout the world. For the most part, numerical engineering data in the text and in tables are presented in SI-based units with the customary U.S. equivalents in parentheses (text) or adjoining columns (tables). For example, pressure, stress, and strength are shown both in SI units, which are pascals (Pa) with a suitable prefix, and in customary U.S. units, which are pounds per square inch (psi). To save space, large values of psi have been converted to kips per square inch (ksi), where 1 ksi = 1000 psi. The metric tonne (kg  103) has sometimes been shown in megagrams (Mg). Some strictly scientific data are presented in SI units only. To clarify some illustrations, only one set of units is presented on artwork. References in the accompanying text to data in the illustrations are presented in both SI-based and customary U.S. units. On graphs and charts, grids corresponding to SI-based units usually appear along the left and bottom edges. Where appropriate, corresponding customary U.S. units appear along the top and right edges. Data pertaining to a specification published by a specification-writing group may be given in only the units used in that specification or in dual units, depending on the nature of the data. For example, the typical yield strength of steel sheet made to a specification written in customary U.S.

units would be presented in dual units, but the sheet thickness specified that specification might be presented only in inches. Data obtained according to standardized test methods for which the standard recommends a particular system of units are presented in the units of that system. Wherever feasible, equivalent units are also presented. Some statistical data may also be presented in only the original units used in the analysis. Conversions and rounding have been done in accordance with IEE/ ASTM SI-10, with attention given to the number of significant digits in the original data. For example, an annealing temperature of 1570  F contains three significant digits. In this case, the equivalent temperature would be given as 855  C; the exact conversion to 854.44  C would not be appropriate. For an invariant physical phenomenon that occurs at a precise temperature (such as the melting of pure silver), it would be appropriate to report the temperature as 961.93  C or 1763.5  F. In some instances (especially in tables and data compilations), temperature values in  C and  F are alternatives rather than conversions. The policy of units of measure in this Handbook contains several exceptions to strict conformance to IEEE/ASTM SI-10; in each instance, the exception has been made in an effort to improve the clarity of the Handbook. The most notable exception is the use of g/cm3 rather than kg/m3 as the unit of measure for density (mass per unit volume). SI practice requires that only one virgule (diagonal) appear in units formed by combination of several basic units. Therefore, all of the units preceding the virgule are in the numerator and all units following the virgule are in the denominator of the expression; no parentheses are required to prevent ambiguity.

v

© 2008 ASM International. All Rights Reserved. ASM Handbook Volume 15: Casting (#05115G)

www.asminternational.org

Preface

From the preceding Volume 5 (1970) of the 8th Edition Metals Handbook and the 9th Edition Metals Handbook Volume 15 (1988) on casting, this Handbook provides an update on the continuing advances in casting technologies and applications. Casting, as both a science and practical tool in art and technology, is enormously varied in scope. It is impossible to capture the full scope of casting technology in one volume.

on sand casting is expanded relative to the previous edition with the intention of providing a reference that may be helpful as a communication tool between product designers and metalcasters in developing successful and economical products. This Volume consists of 18 sections. The first section introduces the historical development of metal casting, as well as to the advantages of castings over parts produced by other manufacturing processes, their applications, and the current market size of the industry. This includes an article on “Metalcasting Technology and the Purchasing Process” written by Al Spada and the technical staff of the American Foundry Society. Then, the principles and practice of melt processing are described in the next three sections followed by a section on principles of solidification including nucleation kinetics, fundamentals of growth, transformation behavior, and microstructure development. Solid-state processing of casting, such as heat treat treatment and hot isostatic pressing, are also introduced. This is followed by a section on the “Modeling and Analysis of Casting Processes.”

The main focus of this Volume is on the products and processes of foundry (shape) casting, although primary (ingot or continuous casting) of steel and aluminum are also covered. In addition, continuous casting of copper is described in an article, as copper continuous casting was a precursor to steel and aluminum continuous casting in some respects. Some of the articles on melt processing, such as the articles on “Electric Arc Furnace Melting” and “Steel Melt Processing,” also briefly describe primary production of cast metal. Shape casting of metal is dominated by cast iron, which constitutes just over 70% of the worldwide production of castings on a tonnage basis (See Table 2 in the first article “History and Trends of Metal Casting”). This is followed by steel, copper-alloy, and aluminum-alloy castings, which make up about 25% on the worldwide tonnage of casting production. Magnesium and zinc are on the order of 1% or less. The dominance of just a few alloys in shape casting is due to the fact that successful and economic shape casting typically involves alloy compositions near a eutectic. The lower melting points and narrower freezing range of near-eutectic compositions promote better castability.

Like the previous edition, traditional subjects such as patterns, molding and casting processes, foundry equipment, and processing considerations are extensively covered in the next sections. As noted, coverage on sand casting has been consolidated and expanded. For example, the major method of shell molding is described in an article—based on an update of a still largely valid 1970 handbook (Volume 5) article. New updates are also provided on processes growing in use, such as squeeze casting, lost-foam casting, semisolid metal forming, and low-pressure casting. The latter is particularly important in producing quality products, as described by John Campbell in the article “Filling and Feeding Concepts.”

Since the 1988 edition of Volume15, several developments have occurred (see Table 5 in the first article “History and Trends of Metal Casting”). Of these developments, computer technology continues to shorten development time and help simulate the casting process. Automation and robotic technology also has improved the productivity and process control of casting. In terms of processes, semisolid processing, squeeze casting, lost-foam, vacuum molding, and various dies casting technologies continue to improve and finds new applications. These important topics are updated in this Volume. For nonferrous alloys, high-pressure die casting of aluminum is a major area of expansion and update in this volume.

Finally, the last five sections describe the major types of cast alloys in term of processing and the properties and characteristics of cast ferrous and nonferrous alloys. Emphasis is placed on cast iron, cast steel, aluminum, copper, and zinc. The last section covers the quality aspects of cast products and the processing of castings. It is hoped that this Handbook is a useful work of peer-consensus reference information for the producers, designers, and buyers of castings. Many thanks are extended to all the contributors and the editors who worked on this Volume. This publication would not have been possible without their commitment and effort.

In addition, coverage on sand casting is expanded and consolidated in this Volume with major articles on “Green Sand Molding,” No-Bake Sand Molding,” and “Shell Molding and Shell Coremaking.” Bonded sand mold casting, although well-established for many years, is the most widely used method of casting on a tonnage basis. Improvement in methods and materials continue to provide better yields, productivity, and product quality. The sand system is also a major factor in the economics of large-volume, production casting. Coverage

Srinath Viswanathan University of Alabama Chair, Volume 15 Editorial Committee

vi

© 2008 ASM International. All Rights Reserved. ASM Handbook Volume 15: Casting (#05115G)

www.asminternational.org

Officers and Trustees of ASM International (2007–2008)

Dianne Chong President and Trustee The Boeing Company Roger J. Fabian Vice President and Trustee Bodycote Thermal Processing Lawrence C. Wagner Immediate Past President and Trustee Texas Instruments Inc (retired) Paul L. Huber Treasurer and Trustee Seco/Warwick Corporation Stanley C. Theobald Secretary and Managing Director ASM International

Trustees

Sue S. Baik-Kromalic Honda of America Christopher C. Berndt Swinburne University of Technology Brady G. Butler University of Utah Danniel P. Dennies The Boeing Company Leigh C. Duren Worcester Polytechnic Institute Pradeep Goyal Pradeep Metals Ltd.

Digby D. Macdonald Penn State University Subhash Mahajan Arizona State University Charles A. Parker Honeywell Aerospace Megan M. Reynolds Washington State University Mark F. Smith Sandia National Laboratories Jon D. Tirpak ATI

Members of the ASM Handbook Committee (2007–2008) Larry D. Hanke (Chair 2006–; Member 1994–) Materials Evaluation and Engineering Inc. Kent L. Johnson (Vice Chair 2006–; Member 1999–) Engineering Systems Inc. Viola L. Acoff (2005–) University of Alabama David E. Alman (2002–2008) National Energy Technology Laboratory Tim Cheek (2004–) DELTA (v) Forensic Engineering Lichun Leigh Chen (2002–) Technical Materials Incorporated Sarup K. Chopra (2007–) Nook Industries Craig Clauser (2005–) Craig Clauser Engineering Consulting Incorporated Craig V. Darragh (1989–)

The Timken Company Jon L. Dossett (2006–) Consultant David U. Furrer (2006–) Rolls-Royce Corporation Lee Gearhart (2005–2008) Moog Inc. Ernest W. Klechka (2006–) Lloyd Register Capstone Inc. Alan T. Male (2003–) University of Kentucky William L. Mankins (1989–) Metallurgical Services Inc. Dana J. Medlin (2005–2008) South Dakota School of Mines and Technology Joseph W. Newkirk (2005–) University of Missouri-Rolla Cory J. Padfield (2006–) Hyundai America Technical Center Inc.

Toby V. Padfield (2004–) ZF Sachs Automotive of America Charles A. Parker (2007–) Honeywell Aerospace Elwin L. Rooy (2007–) Elwin Rooy & Associates Karl P. Staudhammer (1997–) Los Alamos National Laboratory Kenneth B. Tator (1991–) KTA-Tator Inc. George F. Vander Voort (1997–) Buehler Ltd.

Chairs of the ASM Handbook Committee J.F. Harper (1923–1926) (Member 1923–1926) W.J. Merten (1927–1930) (Member 1923–1933) L.B. Case (1931–1933) (Member 1927–1933) C.H. Herty, Jr. (1934–1936) (Member 1930–1936) J.P. Gill (1937) (Member 1934–1937) R.L. Dowdell (1938–1939) (Member 1935–1939) G.V. Luerssen (1943–1947) (Member 1942–1947) J.B. Johnson (1948–1951) (Member 1944–1951) E.O. Dixon (1952–1954) (Member 1947–1955) N.E. Promisel (1955–1961) (Member 1954–1963)

R.W.E. Leiter (1962–1963) (Member 1955–1958, 1960–1964) D.J. Wright (1964–1965) (Member 1959–1967) J.D. Graham (1966–1968) (Member 1961–1970) W.A. Stadtler (1969–1972) (Member 1962–1972) G.J. Shubat (1973–1975) (Member 1966–1975) R. Ward (1976–1978) (Member 1972–1978) G.N. Maniar (1979–1980) (Member 1974–1980) M.G.H. Wells (1981) (Member 1976–1981) J.L. McCall (1982) (Member 1977–1982) L.J. Korb (1983) (Member 1978–1983)

vii

T.D. Cooper (1984–1986) (Member D.D. Huffman (1986–1990) (Member D.L. Olson (1990–1992) (Member 1989–1992) R.J. Austin (1992–1994) (Member W.L. Mankins (1994–1997) (Member M.M. Gauthier (1997–1998) (Member C.V. Darragh (1999–2002) (Member Henry E. Fairman (2002–2004) (Member Jeffrey A. Hawk (2004–2006) (Member Larry D. Hanke (2006–2008) (Member

1981–1986) 1982–) 1982–1988, 1984–1985) 1989–) 1990–2000) 1989–) 1993–) 1997–) 1994–)

© 2008 ASM International. All Rights Reserved. ASM Handbook Volume 15: Casting (#05115G)

www.asminternational.org

Authors and Contributors

University of Birmingham, UK Kevin R. Anderson Mercury Marine

Paul G Campbell Metaullics Systems Division Pyrotek, Inc.

Princewill N. Anyalebechi Padnos College of Engineering & Computing, Grand Valley State University

Robert D Carnahan Thixomat Inc.

Diran Apelian Metal Processing Institute, Worcester Polytechnic Institute

Sujoy Chaudhury Honeywell International, Inc. K. K. Chawla University of Alabama At Birmingham

Alan P. Druschitz University of Alabama at Birmingham Cor van Ettinger Gieterij Doesburg Henry “Ed” Fairman Cincinnati Metallurgical Consultants Edward W. Flynn General Motors

Leigh Chen Technical Materials Incorporated

Jim Frost ACIPCO (American Cast Iron Pipe Company)

Henry Bakemeyer Die Casting Design and Consulting

Yeou-Li Chu Ryobi Die Casting

Richard Fruehan Carnegie Mellon University

Bryan Baker Vulcan Engineering Co.

Craig D. Clauser Craig Clauser Engineering Consulting Incorporated

Rafael Gallo Foseco Metallurgical

Lars Arnberg Norwegian University Of Science & Technology

Steve L. Cockroft The University of British Columbia

Charles-Andre´ Gandin Centre de Mise en Forme des Mate´riaux (CEMEF)

Chris Cooper Vulcan Engineering Co.

George Giles Morganite Crucible Inc

James D. Cotton The Boeing Co.

Peter C. Glaws The Timken Company

Randall Counselman Oldenburg Group Incorporated

Martin E. Glicksman University of Florida

Joe Bigelow CONTECH U. S. LLC.

Jonathan A. Dantzig University of Illinois-Urbana

George Goodrich Stork Climax Research Services

Ron Bird Stainless Foundry & Engineering

Atef Daoud Central Metallurgical Research and Development Institute, Cairo, Egypt

Frank E. Goodwin International Lead Zinc

Stewart A Ballantyne Allvac Allegheny Technologies Company Scott A. Balliett Latrobe Specialty Steel Company Christoph Beckermann University Of Iowa, The Michel Bellet Ecole des Mines de Paris, Centre de Mise en Forme des Mate´riaux (CEMEF)

Malcolm Blair Steel Founders Society of America William Boettinger NIST James A. Brock Sr. Alcoa Harold D. Brody University Of Connecticut Craig C. Brown Stork Technimet, Inc. Zach Brown CONTECH U.S. LLC Andreas Buhrig-Polackzek Gieberei-Institut, Germany Willam A. Butler General Motors (Retired) John Campbell

Craig V. Darragh The Timken Company Babu DasGupta National Science Foundation Bob Dawson Honeywell Aerospace MCOE Raymond Decker Thixomat Incorporated Raymond J Donahue Mercury Marine Daniel Dos Santos Gießerei Institut der RWTH Aachen Jon Dossett Metallurgical Consultant Jean Marie Drezet Ecole Polytechnique Fe´de´rale de Lausanne

viii

Lee Gouwens CMI Novacast Inc. Ronald Graham ATI Wah Chang Patrick Grant Oxford University Daniel E.Groteke Q.C. Designs Inc. Mustafa Guclu Stanley Associates Inc. Richard Gundlach Stork Climax Research Services Nikhil Gupta Polytechnic University Mike Gwyn ATI, Inc. John Hall

© 2008 ASM International. All Rights Reserved. ASM Handbook Volume 15: Casting (#05115G)

Hall Permanent Mold Machines Qingyou Han Purdue University Larry D. Hanke Materials Evaluation and Engineering Inc Justin Heimsch Madison Kipp Corporation

www.asminternational.org

Makhlouf M. Makhlouf Metal Processing Institute, Worcester Polytechnic Institute William L. Mankins Metallurgical Services Incorporated Stephen J. Mashl Bodycote HIP Andover

Intermet-PCPC Robert Pischel FOSECO Metallurgical Inc. David R. Poirier University of Arizona David Poweleit Steel Founders’ Society of America

Juan Carlos Heinrich The University of New Mexico

Ragnvald H. Mathiesen Norwegian University of Science and Technology

Gregg Holman Wah Chang

Robert J. McInerney BuhlerPrince

Thomas E. Prucha American Foundry Society

Robert A. Horton Precision Metalsmiths Incorporated

Scott McIntyre Grede Foundrie

Donald J. Hurtuk H.C. Starck Inc.

Brad McKee Deloro Stellite Inc.

Peter Quested Industry and Innovation Division, National Physical Laboratory, UK

Alain Jacot Laboratoire de Simulation des Mate´riaux

Harold T. Michels Copper Development Association

John L. Jorstad J.L.J. Technologies, Inc.

Tony Midea Foseco Metallurgical Inc.

Ursula R. Kattner N I S T/Metallurgy Division

Paul H Mikkola Metal Casting Technology Incorporated

Seymour Katz S. Katz Associates

Asbjrn Mo SINTEF Materials and Chemistry

Jay S. Keist Worcester Polytechnic Institute

Raymond W. Monroe Steel Founders’ Society of America

Graham Keough Consarc Corporation

Y.V. Murty Cellular Materials International, Inc.

Matthew Krane Purdue University

Laurentiu Nastac Concurrent Technologies Corporation

Mary Beth Krusiak Sand Technology Co. LLC

David V. Neff Metaullics Systems Division Pyrotek, Inc.

T.A. Kuhn Alcoa Primary Metals

Charles Nelson Morris Bean and Company

Wilfried Kurz Ecole Polytechnique Federale De Lausanne (EPFL) Switzerland

Itsuo Ohnaka Osaka Sangyo University, Japan

Selcuk Kuyucak CANMET Materials Technology Laboratory Vic La Fay Hill & Griffith Company Robin A. Lampson Retech Systems LLC

Cory Padfield American Axle & Manufacturing Qingyue Pan SPX Corporation

William L. Powell ThyssenKrupp-Waupaca (Retired)

Wayne Rasmussen Consultant Martin Reeves Striko-Dynarad Allen Richardson SPX Corporation Pradeep Rohatgi University of Wisconsin-Milwaukee Elwin L. Rooy Elwin Rooy and Associates Gary Ruff SMW Automotive Corporation Kumar (Muthukumarasamy) Sadayappan CANMET Materials Technology Laboratory Mahi Sahoo CANMET Materials Technology Laboratory Kenneth Savage Magnesium Elektron Gary Scholl Metal Casting Technology Incorporated Ted Schorn Enkei America, Inc. Benjamin F. Schultz University of Wisconsin-Milwakee David Schwam Case Werstern Reserve University

Charles A. Parker Honeywell Aerospace

William D. Scott AAA Alchemy

David Lee Howmet Corp.

Kent Peaslee Missouri University of Science and Technology (Missouri S&T)

Ben Q. Li University of Michigan Dearborn

Pehlke, Robert D. Univ of Michigan

Sumanth Shankar McMaster University

Alan Luo General Motors R&D Center

John H. Perepezko University of Wisconsin, Madison

Geoffrey K. Sigworth Alcoa Primary Metals

Dan Maas Ex-One

Ralph Perkul Foundry Solutions and Design LLC

Steve Sikkenga Cannon-Muskegon Corporation

James F. Major Alcan International Ltd

Gordon H. Peters

Brian V. Smith

ix

John Serra Carpenter Brothers Inc.

© 2008 ASM International. All Rights Reserved. ASM Handbook Volume 15: Casting (#05115G)

General Motors Corporation, Powertrain Division Jerry Sokolowski University of Windsor Al Spada American Foundry Society Karl Staudhammer Los Alamos National Laboratory Doru M. Stefanescu The Ohio State University Ingo Steinbach RWTH-Aachen, ACCESS e. V. Karthik Subbiah SPX Product Solutions Group Erin H. Sunseri Iowa State University Tony Suschil Foseco Inc Richard Sutherlin ATI Wah Chang Edward S.Szekeres Casting Consultants Incorporated Brian G. Thomas

www.asminternational.org

University of Illinois Michael L Tims Concurrent Technologies Corporation David W. Tripp TIMET

Srinath Viswanathan University of Alabama Birmingham Robert C. Voigt Penn State University Vaughan R. Voller University of Minnesota

Rohit Trivedi AMES Laboratory Paul Trojan University of Michigan-Dearborn John Troxler Rolls Royce

Tom Waring ME Elecmetal Sufei Wei Centrifugal Casting Machine Company Incorporated David White Schaefer Furnace

Nao Tsumagari American Showa, Inc. Daniel Twarog North American Die Casting Association Derek E. Tyler Olin Metals Thin Strip (retired) Don Tyler General Aluminum Manufacturing Company Stephen Udvardy North American Die Casting Association Juan J. Valencia Concurrent Technologies Corporation

x

Lawrence Whiting CANMET Materials Technology Laboratory Richard Williams FOSECO Metallurgical Inc. Greg G.Woycik Woycik Consulting Oscar Yu RTI International Metals, Inc. Keith Zhang Teck Cominco Metals Ltd.

© 2008 ASM International. All Rights Reserved. ASM Handbook Volume 15: Casting (#05115G)

www.asminternational.org

Contents

Applications and Specification Guidelines. . . . . . . . . . . . . . . . . . . 1 Raymond W. Monroe, Steel Founders’ Society of America; Thomas E. Prucha, American Foundry Society; Daniel Twarog, North American Die Casting Association History and Trends of Metal Casting. . . . . . . . . . . . . . History of Metal Casting . . . . . . . . . . . . . . . . . . . . Metal Casting Methods. . . . . . . . . . . . . . . . . . . . . . Production Trends . . . . . . . . . . . . . . . . . . . . . . . . . Market Trends . . . . . . . . . . . . . . . . . . . . . . . . . . . . Metalcasting Technology and the Purchasing Process . . . . The Purchasing Process . . . . . . . . . . . . . . . . . . . . . Define Requirements and Develop a Purchasing Plan. Requesting and Evaluating Quotations/Bids . . . . . . . Select a Supplier and Negotiate Contract Terms . . . . Contract Fulfillment and Continuous Improvement . . Considerations in Purchasing Metal Castings . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

Principles of Liquid Metal Processing . . . . . . . . . . . . Chemical Thermodynamics and Kinetics . . . . . . . . . . . . Chemical Thermodynamics . . . . . . . . . . . . . . . . . . Chemical Kinetics . . . . . . . . . . . . . . . . . . . . . . . . Kinetics of Alloys Additions . . . . . . . . . . . . . . . . . Thermodynamic Properties of Iron-Base Alloys . . . . . . . Purification of Ferrous Melts. . . . . . . . . . . . . . . . . Thermodynamics of Ferrous Systems . . . . . . . . . . . Multicomponent Iron-Carbon Systems . . . . . . . . . . Structural Diagrams . . . . . . . . . . . . . . . . . . . . . . . Thermodynamic Properties of Aluminum-Base and Copper-Base Alloys . . . . . . . . . . . . . . . . . . . . Activities and Thermal Properties . . . . . . . . . . . . . Ineraction Coefficients . . . . . . . . . . . . . . . . . . . . . Thermal Properties for Hypothetical Standard State . Phase Diagrams . . . . . . . . . . . . . . . . . . . . . . . . . . Gases in Metals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Cast Iron. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Aluminum and Its Alloys . . . . . . . . . . . . . . . . . . . Copper and Its Alloys . . . . . . . . . . . . . . . . . . . . . Magnesium and Its Alloys . . . . . . . . . . . . . . . . . . Overcoming Gas Porosity . . . . . . . . . . . . . . . . . . . Inclusion-Forming Reactions . . . . . . . . . . . . . . . . . . . . Inclusion Types . . . . . . . . . . . . . . . . . . . . . . . . . . Physical Chemistry . . . . . . . . . . . . . . . . . . . . . . . Control of Inclusions . . . . . . . . . . . . . . . . . . . . . . Inclusions in Ferrous Alloys . . . . . . . . . . . . . . . . . Inclusions in Nonferrous Alloys . . . . . . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . .

29 31 31 35 35 41 41 45 50 51

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

56 56 56 58 58 64 64 67 70 71 73 74 74 75 76 77 81

Melting and Remelting . . . . . . . . . . Electric Arc Furnace Melting . . . . . . . Power Supply . . . . . . . . . . . . . . Arc Furnace Components . . . . . . Weighing and Chemical Analysis Scrap and Alloy Storage . . . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

85 87 87 88 91 91

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . .

Acid Melting Practice . . . . . . . . . . . . . . . . . Basic Melting Practice . . . . . . . . . . . . . . . . . Newer Technologies. . . . . . . . . . . . . . . . . . . Cupola Furnaces . . . . . . . . . . . . . . . . . . . . . . . . . Progressive Refinements in Cupola Equipment Cupola Construction and Operation . . . . . . . . Charge Materials . . . . . . . . . . . . . . . . . . . . . Cupola Control Principles . . . . . . . . . . . . . . . Control Tests and Analyses. . . . . . . . . . . . . . Specialized Cupolas . . . . . . . . . . . . . . . . . . . Induction Furnaces . . . . . . . . . . . . . . . . . . . . . . . Types of Furnaces . . . . . . . . . . . . . . . . . . . . Electromagnetic Stirring . . . . . . . . . . . . . . . . Power Supplies . . . . . . . . . . . . . . . . . . . . . . Water Cooling Systems . . . . . . . . . . . . . . . . Lining Material . . . . . . . . . . . . . . . . . . . . . . Melting Operations . . . . . . . . . . . . . . . . . . . Vacuum Induction Melting . . . . . . . . . . . . . . . . . Process Description . . . . . . . . . . . . . . . . . . . VIM Metallurgy . . . . . . . . . . . . . . . . . . . . . Production of Nonferrous Materials . . . . . . . . Electroslag Remelting . . . . . . . . . . . . . . . . . . . . . ESR Furnaces . . . . . . . . . . . . . . . . . . . . . . . Ingot Solidification. . . . . . . . . . . . . . . . . . . . Process Variations . . . . . . . . . . . . . . . . . . . . Steel ESR . . . . . . . . . . . . . . . . . . . . . . . . . . Superalloy ESR . . . . . . . . . . . . . . . . . . . . . . Vacuum Arc Remelting . . . . . . . . . . . . . . . . . . . . Process Description . . . . . . . . . . . . . . . . . . . Solidification . . . . . . . . . . . . . . . . . . . . . . . . Process Variables. . . . . . . . . . . . . . . . . . . . . Superalloy VAR . . . . . . . . . . . . . . . . . . . . . Titanium VAR . . . . . . . . . . . . . . . . . . . . . . Skull Melting . . . . . . . . . . . . . . . . . . . . . . . . . . . Vacuum Arc Skull Melting . . . . . . . . . . . . . . Furnaces . . . . . . . . . . . . . . . . . . . . . . . . . . . Induction Skull Melting . . . . . . . . . . . . . . . . Electron Beam Melting . . . . . . . . . . . . . . . . . . . . Process Description . . . . . . . . . . . . . . . . . . . Drip Melting . . . . . . . . . . . . . . . . . . . . . . . . Cold Hearth Melting . . . . . . . . . . . . . . . . . . Plasma Melting and Heating . . . . . . . . . . . . . . . . Plasma Torches . . . . . . . . . . . . . . . . . . . . . . Furnace Equipment . . . . . . . . . . . . . . . . . . . Atmosphere Control . . . . . . . . . . . . . . . . . . . Melting Processes . . . . . . . . . . . . . . . . . . . . Crucible Furnaces . . . . . . . . . . . . . . . . . . . . . . . . Applications . . . . . . . . . . . . . . . . . . . . . . . . Furnace Types . . . . . . . . . . . . . . . . . . . . . . . Design and Operation. . . . . . . . . . . . . . . . . . Reverberatory and Stack Furnaces . . . . . . . . . . . . Furnace Types . . . . . . . . . . . . . . . . . . . . . . . Reverberatory Furnace Practice . . . . . . . . . . . Molten Metal Circulation . . . . . . . . . . . . . . .

3 4 8 9 14 16 16 17 21 22 22 22

xi

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . .

91 95 98 99 99 100 103 106 106 107 108 109 109 111 111 113 116 116 119 122 124 125 128 129 129 131 132 132 134 135 137 137 139 139 139 140 142 142 143 145 149 149 149 151 151 151 155 155 156 158 160 161 163 165

© 2008 ASM International. All Rights Reserved. ASM Handbook Volume 15: Casting (#05115G)

www.asminternational.org

Molten Metal Processing and Handling . . . . . . . . . . . . . Transfer and Treatment of Molten Metal—An Introduction . Launders. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Tundishes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Molten-Metal Pumps . . . . . . . . . . . . . . . . . . . . . . . . Pouring and Dosing . . . . . . . . . . . . . . . . . . . . . . . . . Degassing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Aluminum Foundry Degassing . . . . . . . . . . . . . . . . . Degassing of Magnesium . . . . . . . . . . . . . . . . . . . . . Degassing of Copper Alloys . . . . . . . . . . . . . . . . . . . Molten-Metal Filtration . . . . . . . . . . . . . . . . . . . . . . . . . . Sources of Inclusion. . . . . . . . . . . . . . . . . . . . . . . . . Removal of Inclusions . . . . . . . . . . . . . . . . . . . . . . . Filter Characteristics . . . . . . . . . . . . . . . . . . . . . . . . Types of Molten-Metal Filters . . . . . . . . . . . . . . . . . . In-Mold Filtration . . . . . . . . . . . . . . . . . . . . . . . . . . In-Furnace Filtration of Molten Aluminum . . . . . . . . . Filtered Metal Quality . . . . . . . . . . . . . . . . . . . . . . . Steel Melt Processing . . . . . . . . . . . . . . . . . . . . . . . . . . . Furnaces and Refractories . . . . . . . . . . . . . . . . . . . . . Converter Metallurgy . . . . . . . . . . . . . . . . . . . . . . . . Ladle Metallurgy . . . . . . . . . . . . . . . . . . . . . . . . . . . Aluminum Fluxes and Fluxing Practice . . . . . . . . . . . . . . . Aluminum Fluxing. . . . . . . . . . . . . . . . . . . . . . . . . . Categories of Fluxes . . . . . . . . . . . . . . . . . . . . . . . . Foundry Practices . . . . . . . . . . . . . . . . . . . . . . . . . . Modification of Aluminum-Silicon Alloys . . . . . . . . . . . . . Heat Treated Castings . . . . . . . . . . . . . . . . . . . . . . . Porosity and Shrinkage Formation . . . . . . . . . . . . . . . Modification Mechanisms . . . . . . . . . . . . . . . . . . . . . Eutectic Nucleation . . . . . . . . . . . . . . . . . . . . . . . . . Eutectic Grain Structure . . . . . . . . . . . . . . . . . . . . . . Recommended Foundry Practices . . . . . . . . . . . . . . . Growth of Silicon Eutectic . . . . . . . . . . . . . . . . . . . . Effect of Other Elements . . . . . . . . . . . . . . . . . . . . . Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Grain Refinement of Aluminum Casting Alloys . . . . . . . . . Measurement of Grain Size. . . . . . . . . . . . . . . . . . . . Mechanisms of Grain Refinement . . . . . . . . . . . . . . . Boron as a Grain Refiner . . . . . . . . . . . . . . . . . . . . . Best Practices for Grain Refinement. . . . . . . . . . . . . . Benefits of Grain Refinement . . . . . . . . . . . . . . . . . . Concluding Remarks . . . . . . . . . . . . . . . . . . . . . . . . Refinement of the Primary Silicon Phase in Hypereutectic Aluminum-Silicon Alloys . . . . . . . . . . . . . . . . . . . . . Importance of Primary Silicon Refinement . . . . . . . . . Accomplishing Primary Silicon Refinement . . . . . . . . Principles of Solidification . . . . . . . . . . . . . . . . Diran Apelian, Worcester Polytechnic Institute Thermodynamics and Phase Diagrams . . . . . . . . Thermodynamics . . . . . . . . . . . . . . . . . . . . Phase Diagrams . . . . . . . . . . . . . . . . . . . . . Solidification . . . . . . . . . . . . . . . . . . . . . . . Nucleation Kinetics and Grain Refinement . . . . . Thermodynamics of Solidification . . . . . . . . Nucleation Phenomena . . . . . . . . . . . . . . . . Inoculation Practice . . . . . . . . . . . . . . . . . . Grain Refinement Models . . . . . . . . . . . . . . Master Alloy Processing . . . . . . . . . . . . . . . Transport Phenomena During Solidification . . . . . Fundamentals . . . . . . . . . . . . . . . . . . . . . . Transport and Microstructure . . . . . . . . . . . Summary . . . . . . . . . . . . . . . . . . . . . . . . . Plane Front Solidification . . . . . . . . . . . . . . . . . Transients and Steady State . . . . . . . . . . . . Morphological Stability . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

171 173 174 174 175 178 185 185 188 189 194 194 196 197 199 199 201 204 206 206 208 212 230 231 233 235 240 241 242 243 244 245 247 249 249 250 255 255 255 257 258 260 260

Rapid Solidification Effects. . . . . . . . . . . . . . . . . . . . . . Solidification Microstructure Selection Maps . . . . . . . . . Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Non-Plane Front Solidification . . . . . . . . . . . . . . . . . . . . . . . Cellular Microstructures . . . . . . . . . . . . . . . . . . . . . . . . Dendrite Growth . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Nonplanar Eutectic Growth. . . . . . . . . . . . . . . . . . . . . . Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Eutectic Solidification—An Introduction . . . . . . . . . . . . . . . . Binary Eutectic Phase Diagram . . . . . . . . . . . . . . . . . . . Eutectic Structures . . . . . . . . . . . . . . . . . . . . . . . . . . . . Solidification of Eutectic Alloys—Aluminum-Silicon . . . . . . . The Aluminum-Silicon Equilibrium Phase Diagram . . . . . Classification of the Aluminum-Silicon Eutectic Structure Microstructure of the Aluminum-Silicon Eutectic . . . . . . Theories of Solidification of the Aluminum-Silicon Eutectic . . . . . . . . . . . . . . . . . . Theories of Chemical Modification of the Aluminum-Silicon Eutectic . . . . . . . . . . . . . . . . . . Solidification of Eutectic Alloys—Cast Iron. . . . . . . . . . . . . . Structure of Liquid Iron-Carbon Alloys . . . . . . . . . . . . . Nucleation of Eutectic in Cast Iron . . . . . . . . . . . . . . . . Nucleation of Primary Phases in Cast Iron . . . . . . . . . . . Growth of Eutectic in Cast Iron. . . . . . . . . . . . . . . . . . . Cooling Curve Analysis . . . . . . . . . . . . . . . . . . . . . . . . Peritectic Solidification . . . . . . . . . . . . . . . . . . . . . . . . . . . . Peritectic Solidification. . . . . . . . . . . . . . . . . . . . . . . . . Peritectic Reaction . . . . . . . . . . . . . . . . . . . . . . . . . . . . Peritectic Transformation and Direct Precipitation . . . . . . Primary Metastable Precipitation of Beta . . . . . . . . . . . . Peritectic Cascades . . . . . . . . . . . . . . . . . . . . . . . . . . . Peritectic Transformations in Multicomponent Systems . . Microsegregation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Solidification Models . . . . . . . . . . . . . . . . . . . . . . . . . . Binary Isomorphous Systems, k > 1: Titanium-Molybdenum . . . . . . . . . . . . . . . . . . . . . Binary Eutectic Systems, k > 1: Hypoeutectic Aluminum-Copper Alloys . . . . . . . . . . . . . . . . . . . Binary Eutectic Systems: Hypoeutectic Aluminum-Silicon Alloys . . . . . . . . . . . . . . . . . . . Binary Peritectic Systems: Copper-Zinc a-Brass . . . . . . . Multicomponent, Multiphase Eutectic Systems: Hypoeutectic Al-Si-Cu-Mg Alloys . . . . . . . . . . . . . . . Multicomponent Multiphase Steel: Fe-C-Cr . . . . . . . . . . Multicomponent, Multiphase Nickel-Base Superalloy. . . . Macrosegregation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Macrosegregation Induced by Flow of the Liquid . . . . . . Macrosegregation Induced by Movement of the Solid . . . Interpretation and use of Coolng Curves—Thermal Analysis . . Quantitative Thermal Analysis . . . . . . . . . . . . . . . . . . . Deviation from Equilibrium . . . . . . . . . . . . . . . . . . . . . Differential Thermal Analysis . . . . . . . . . . . . . . . . . . . . Determining Liquidus and Solidus Temperature . . . . . . . X-Ray Imaging of Solidification Processes and Microstructure Evolution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . X-Ray Imaging . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . In-House Laboratory X-Ray Imaging of Solidification Processes . . . . . . . . . . . . . . . . . . . . . X-Ray Imaging Studies of Solidification Processes with Synchrotron Radiation . . . . . . . . . . . . . . . . . . . . Further Reading. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The Oxide Film Defect—The Bifilm. . . . . . . . . . . . . . . . . . . Bifilms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Entrainment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Furling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Unfurling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Reliability and Reproducibility of Mechanical Properties .

. . . . . 263 . . . . . 263 . . . . . 263

. . . . . . . . . . . . 267 . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

269 269 271 272 276 276 278 280 281 285 288 288 290 292 293 294 295

xii

. . . . . . . . . . . . . . .

. . . . . . . . . . . . . . .

. . . . . . . . . . . . . . .

296 297 297 299 300 301 304 305 307 307 307 312 312 313 313

. . . 313 . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . .

315 317 317 317 320 320 327 330 330 331 332 333 334 335 338 338

. . . 339 . . . 340 . . . 344 . . . 344 . . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

345 345 345 348 348 351 353 353 354 354 355

. . . 357 . . . 357 . . . 357 . . . . . . . .

. . . . . . . .

. . . . . . . .

358 360 362 362 363 364 364 367

© 2008 ASM International. All Rights Reserved. ASM Handbook Volume 15: Casting (#05115G)

www.asminternational.org

Central Role of Bifilm Defects in Fracture . . . . . . . . . . . Shrinkage Porosity and Gas Porosity. . . . . . . . . . . . . . . . . . . Shrinkage Porosity . . . . . . . . . . . . . . . . . . . . . . . . . . . . Gas Porosity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Factors Affecting Porosity Formation . . . . . . . . . . . . . . . Castability—Fluidity and Hot Tearing. . . . . . . . . . . . . . . . . . Fluidity. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Hot Tearing. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Semisolid Metal Processing . . . . . . . . . . . . . . . . . . . . . . . . . Flow Behavior of SSM. . . . . . . . . . . . . . . . . . . . . . . . . Semisolid Microstructural Formation . . . . . . . . . . . . . . . Semisolid Processes: Thixocasting versus Rheocasting . . . Spray Casting. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Principles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Atomization and the Droplet Spray . . . . . . . . . . . . . . . . Deposition and Grain Mulitplication . . . . . . . . . . . . . . . Microsegregation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Macrosegregation and Coarsening . . . . . . . . . . . . . . . . . Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Rapid Solidification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Microstructural Modification . . . . . . . . . . . . . . . . . . . . . Solidification During Casting of Metal-Matrix Composites . . . Incorporation of Reinforcements . . . . . . . . . . . . . . . . . . Reinforcement-Metal Wettability . . . . . . . . . . . . . . . . . . Solidification Processing of MMC . . . . . . . . . . . . . . . . . Solidification Fundamentals . . . . . . . . . . . . . . . . . . . . . Modeling of Particle-Pushing Phenomenon . . . . . . . . . . . Nanocomposites. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Conclusions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Solidification Research in Microgravity. . . . . . . . . . . . . . . . . Availability of Reduced Gavity: Past, Present, and Future Solidification Research in Microgravity . . . . . . . . . . . . . Key Solidification Experiments . . . . . . . . . . . . . . . . . . . The Future of Microgravity Solidification . . . . . . . . . . . . Homogenization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Sources of Inhomogeneity. . . . . . . . . . . . . . . . . . . . . . . Benefits of Homogenization Treatment. . . . . . . . . . . . . . Characterization Methods . . . . . . . . . . . . . . . . . . . . . . . Computational Modeling . . . . . . . . . . . . . . . . . . . . . . . Heat Treatment. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Designations of Heat Treatment . . . . . . . . . . . . . . . . . . Solution Heat Treatment. . . . . . . . . . . . . . . . . . . . . . . . Quenching . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Aging . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Emerginhg Heat Treating Technologies . . . . . . . . . . . . . Microstructural Changes due to Heat Treatment of Cast Alloys . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Hot Isostatic Pressing of Castings. . . . . . . . . . . . . . . . . . . . . History of HIP. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Reasons for Using HIP . . . . . . . . . . . . . . . . . . . . . . . . . Overview of a HIP System . . . . . . . . . . . . . . . . . . . . . . Effect of HIP on Mechanical Properties . . . . . . . . . . . . . Effect of Inclusions on As-HIPed Properties . . . . . . . . . . Effect of HIP on the Shape and Structure of Castings . . . Problems Encountered in HIP . . . . . . . . . . . . . . . . . . . . Economics of HIP . . . . . . . . . . . . . . . . . . . . . . . . . . . . Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Modeling and Analysis of Casting Processes. . . . . . . . Srinath Viswanathan, University of Alabama Numerical Methods for Casting Applications . . . . . . . . . Fundamentals of CFD . . . . . . . . . . . . . . . . . . . . . Numerical Solution of the Fluid-Flow Equations . . . Grid Generation for Complex Geometries. . . . . . . . Casting Applications . . . . . . . . . . . . . . . . . . . . . . Modeling of Transport Phenomena and Electromagnetics Conservation of Thermal Energy . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

368 370 370 372 373 375 375 376 379 379 380 381 382 382 383 384 385 385 385 386 386 390 391 391 391 392 394 395 395 398 398 399 399 400 402 402 402 402 403 404 404 404 405 405 406

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

406 408 408 409 409 411 412 413 414 415 415

. . . . . . . 417 . . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

419 419 420 420 422 425 425

Conservation of Solute . . . . . . . . . . . . . . . . . . . . . . . . . Equations Cast in Advection-Diffusion Form . . . . . . . . . Basic Equations for Mass and Momentum Conservation . Modeling the Momentum in the Solid + Liquid Mushy Region . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Modeling Turbulence . . . . . . . . . . . . . . . . . . . . . . . . . . Phenomena Restraining or Driving Flow . . . . . . . . . . . . Dealing with Fronts and Free Surfaces . . . . . . . . . . . . . . Simulation of Electromagnetic Effects in Castings. . . . . . Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Direct Modeling of Structure Formation . . . . . . . . . . . . . . . . Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Direct Microstructure Simulation Using the Phase Field Method. . . . . . . . . . . . . . . . . . . . . . . . Direct Grain Structure Simulation Using the Cellular Automation Method . . . . . . . . . . . . . . . . . Coupling of Direct Structure Simulation at Macroscopic Scale . . . . . . . . . . . . . . . . . . . . . . . . Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Modeling of Microsegregation and Macrosegregation . . . . . . . Microsegregation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Microsegregation with Diffusion in the Solid . . . . . . . . . Continuum Model of Macrosegregation . . . . . . . . . . . . . Modeling of Stress, Distortion, and Hot Tearing . . . . . . . . . . Governing Equations . . . . . . . . . . . . . . . . . . . . . . . . . . Thermomechanical Coupling . . . . . . . . . . . . . . . . . . . . . Numerical Solution . . . . . . . . . . . . . . . . . . . . . . . . . . . Model Validation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . Example Applications . . . . . . . . . . . . . . . . . . . . . . . . . Hot-Tearing Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . Conclusions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Practical Issues In Computer Simulation of Casting Processes . Setting Simulation Objectives, Modeling, and Selection of Simulation Code . . . . . . . . . . . . . . . . . . . . . . . . Modeling of Shape and Phenomena . . . . . . . . . . . . . . . . Initial and Boundary Conditions and Physical Properties . Enmeshing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Evaluation of Simulation Results . . . . . . . . . . . . . . . . . . Estimation of Structure and Properties . . . . . . . . . . . . . . Optimization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Thermophysical Properties . . . . . . . . . . . . . . . . . . . . . . . . . . Sources and Availability of Reliable Data . . . . . . . . . . . Limitations and Warning on the Use of Data . . . . . . . . . Methods to Determine Thermophysical Properties . . . . . . Specific Heat Capacity and Enthalpy of Transformation . . Enthalpy of Melting, Solidus, and Liquidus Temperatures Coefficient of Thermal Expansion . . . . . . . . . . . . . . . . . Density . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Surface Tension. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Viscosity. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Electrical and Thermal Conductivity . . . . . . . . . . . . . . . Emissivity. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Typical Themophysical Properties Ranges of Some Cast Alloys . . . . . . . . . . . . . . . . . . . . . . . . . Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Principles and Practices of Shape Casting . . . . . . . Shape Casting Processes—An Introduction . . . . . . . . Patterns and Patternmaking . . . . . . . . . . . . . . . . . . . Types of Patterns . . . . . . . . . . . . . . . . . . . . . . . Additional Features of Patterns . . . . . . . . . . . . . Pattern Allowances . . . . . . . . . . . . . . . . . . . . . Pattern Materials . . . . . . . . . . . . . . . . . . . . . . . Selection of Pattern Type . . . . . . . . . . . . . . . . . Trends in Pattern Design and Manufacture . . . . . Casting Practice—Guidelines for Effective Production of Reliable Castings . . . . . . . . . . . . . . . . . . . . .

xiii

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . 425 . . . 427 . . . 427 . . . . . . . .

. . . . . . . .

. . . . . . . .

428 428 429 429 431 432 435 435

. . . 436 . . . 438 . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . .

. . . . . . . . . . . . . . .

441 443 445 445 445 446 449 449 451 451 453 454 456 458 462

. . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . .

462 462 463 464 465 466 466 468 468 468 468 470 470 471 471 471 474 474 474

. . . 477 . . . 477 . . . . . . . . .

. . . . . . . . .

. . . . . . . . .

483 485 488 488 490 491 492 494 495

. . . . . . . . . 497

© 2008 ASM International. All Rights Reserved. ASM Handbook Volume 15: Casting (#05115G)

www.asminternational.org

Rule 1: Good-Quality Melt . . . . . . . . . . . . . . . . Rule 2: Liquid Front Damage . . . . . . . . . . . . . . Rule 3: Liquid Front Stop. . . . . . . . . . . . . . . . . Rule 4: Bubble Damage . . . . . . . . . . . . . . . . . . Rule 5: Core Blows . . . . . . . . . . . . . . . . . . . . . Rule 6: Shrinkage Damage . . . . . . . . . . . . . . . . Rule 7: Convection Damage . . . . . . . . . . . . . . . Rule 8: Segregation . . . . . . . . . . . . . . . . . . . . . Rule 9: Residual Stress. . . . . . . . . . . . . . . . . . . Rule 10: Location Points . . . . . . . . . . . . . . . . . Filling and Feeding System Concepts . . . . . . . . . . . . Entrainment Defects in Gravity-Poured Castings . Traditional Filling Designs . . . . . . . . . . . . . . . . Prepriming Techniques . . . . . . . . . . . . . . . . . . . Naturally Pressurized Filling System . . . . . . . . . Filters. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Performance of Alternative Filling Systems . . . . Recommended Foundry Practice . . . . . . . . . . . . Processing and Finishing of Castings . . . . . . . . . . . . Shakeout and Core Knockout . . . . . . . . . . . . . . Cleaning Operations. . . . . . . . . . . . . . . . . . . . . Automating Gate Removal and Grinding . . . . . . Flame Cutting . . . . . . . . . . . . . . . . . . . . . . . . . Blast Cleaning of Castings . . . . . . . . . . . . . . . . Inspection . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . .

Expandable-Mold Casting Processes with Permanent Patterns. . . . . . . . . . . . . . . . . . . John L. Jorstad, J.L.J. Technologies, Inc Introduction—Expendable Mold Processes with Permanent Patterns . . . . . . . . . . . . . Aggregates and Binders for Expendable Molds Mold Media . . . . . . . . . . . . . . . . . . . . . Foundry Sands. . . . . . . . . . . . . . . . . . . . Clays for Green Sand Molding . . . . . . . . Inorganic Binders . . . . . . . . . . . . . . . . . Organic Binders. . . . . . . . . . . . . . . . . . . Other Expendable Mold Media . . . . . . . . Green Sand Molding . . . . . . . . . . . . . . . . . . . Sand Molding . . . . . . . . . . . . . . . . . . . . System Control and Formulation . . . . . . . Influence of Molding Equipment . . . . . . . Maintaining Sand System Quality . . . . . . Molding Methods. . . . . . . . . . . . . . . . . . Green Sand Media Preparation . . . . . . . . Molding Problems . . . . . . . . . . . . . . . . . Mold Finishing . . . . . . . . . . . . . . . . . . . Shakeout. . . . . . . . . . . . . . . . . . . . . . . . Sand/Casting Recovery. . . . . . . . . . . . . . Sand Reclamation . . . . . . . . . . . . . . . . . No-Bake Sand Molding . . . . . . . . . . . . . . . . . No-Bake Sand Processing . . . . . . . . . . . . CO2 -Cured Sodium Silicate . . . . . . . . . . Air-Setting Sodium Silicates . . . . . . . . . . Air-Setting Organic Binders . . . . . . . . . . Gas-Cured Organic Binders. . . . . . . . . . . Sand Reclamation and Reuse . . . . . . . . . Coremaking . . . . . . . . . . . . . . . . . . . . . . . . . Binders for Sand Coremaking . . . . . . . . . Basic Principles of Coremaking. . . . . . . . Core Sands and Additives. . . . . . . . . . . . Compaction. . . . . . . . . . . . . . . . . . . . . . Curing of Compacted Cores . . . . . . . . . . Core Coatings (Core Washes) . . . . . . . . . Assembly and Core Setting . . . . . . . . . . . Design Considerations . . . . . . . . . . . . . . Coring of Tortuous Passages . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . .

497 497 499 500 501 501 502 503 504 505 506 506 506 507 508 511 511 512 513 513 516 517 519 519 520

. . . . . . . . . . . . . . 523 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

525 528 528 529 535 538 539 546 549 549 551 552 553 554 558 560 560 561 562 564 567 567 569 572 573 576 577 581 581 582 583 584 585 587 588 589 592

Green Sand versus Dry Sand Cores . . . . . Cores in Permanent Mold Castings . . . . . Cores in Investment Castings . . . . . . . . . Designing to Eliminate Cores . . . . . . . . . Coring versus Drilling . . . . . . . . . . . . . . Shell Molding and Shell Coremaking . . . . . . . Applicability . . . . . . . . . . . . . . . . . . . . . Properties and Selection of Shell Sands . . Resins . . . . . . . . . . . . . . . . . . . . . . . . . Preparation of Resin-Sand Mixtures . . . . . Sand Reclamation . . . . . . . . . . . . . . . . . Control Testing of Resin-Sand Properties . Patterns and Core Boxes. . . . . . . . . . . . . Production of Molds and Cores . . . . . . . . Mold Assembly . . . . . . . . . . . . . . . . . . . Casting Process . . . . . . . . . . . . . . . . . . . Dimensional Accuracy . . . . . . . . . . . . . . Causes and Prevention of Casting Defects Comparisons with Green Sand Molding . . Slurry Molding . . . . . . . . . . . . . . . . . . . . . . . Plaster Molding . . . . . . . . . . . . . . . . . . . Applications . . . . . . . . . . . . . . . . . . . Plaster Mold Compositions . . . . . . . . . Patterns and Core Boxes . . . . . . . . . . . Mold-Drying Equipment . . . . . . . . . . . Metals Cast in Plaster Molds . . . . . . . . Conventional Plaster Mold Casting. . . . Match Plate Pattern Molding . . . . . . . . The Antioch Process . . . . . . . . . . . . . . Foamed Plaster Molding Process . . . . . Ceramic Molding . . . . . . . . . . . . . . . . . . Applications . . . . . . . . . . . . . . . . . . . Shaw Process . . . . . . . . . . . . . . . . . . . Unicast Process . . . . . . . . . . . . . . . . . No-Bond Sand Molding. . . . . . . . . . . . . . . . . Magnetic Molding . . . . . . . . . . . . . . . . . Vacuum (V-Process) Molding . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Expandable-Mold Casting Processes with Expendable Patterns . . . . . . . . . . . . . . . . . . . John L. Jorstad, J.L.J. Technologies, Inc Introduction—Expendable Mold Processes with Expendable Patterns . . . . . . . . . . . . . . . . . Lost Foam Casting . . . . . . . . . . . . . . . . . . . . . Process Technique . . . . . . . . . . . . . . . . . . Processing Parameters . . . . . . . . . . . . . . . Advantages of Lost Foam Casting . . . . . . . Casting Quality . . . . . . . . . . . . . . . . . . . . Investment Casting . . . . . . . . . . . . . . . . . . . . . Pattern Materials . . . . . . . . . . . . . . . . . . . Patternmaking . . . . . . . . . . . . . . . . . . . . . Pattern and Cluster Assembly . . . . . . . . . . Ceramic Shell Mold Manufacture . . . . . . . Manufacture of Ceramic Cores . . . . . . . . . Pattern Removal . . . . . . . . . . . . . . . . . . . Mold Firing and Burnout . . . . . . . . . . . . . Melting and Casting. . . . . . . . . . . . . . . . . Postcasting Operations . . . . . . . . . . . . . . . Inspection and Testing . . . . . . . . . . . . . . . Design Advantages of Investment Castings . Design Recommendations . . . . . . . . . . . . . Applications . . . . . . . . . . . . . . . . . . . . . . Special Investment Casting Processes. . . . . Replicast Molding. . . . . . . . . . . . . . . . . . . . . . Process Details . . . . . . . . . . . . . . . . . . . . Process Capabilities . . . . . . . . . . . . . . . . .

xiv

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

593 593 594 595 596 598 599 599 600 601 603 603 604 604 607 610 611 614 615 617 617 617 617 618 618 618 619 620 620 621 622 622 622 624 628 628 628

. . . . . . . . . . . . . 635 . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . .

637 640 640 642 645 645 646 646 648 650 651 654 654 654 655 656 656 657 657 657 657 662 662 663

© 2008 ASM International. All Rights Reserved. ASM Handbook Volume 15: Casting (#05115G)

Centrifugal Casting . . . . . . . . . . . . . . . . . John L. Jorstad, J.L.J. Technologies, Inc Centrifugal Casting . . . . . . . . . . . . . . . . . . Centrifugal Casting Methods . . . . . . . . Equipment . . . . . . . . . . . . . . . . . . . . Molds. . . . . . . . . . . . . . . . . . . . . . . . Defects in Centrifugal Castings . . . . . . Applications . . . . . . . . . . . . . . . . . . . Horizontal Centrifugal Casting . . . . . . . . . . Equipment . . . . . . . . . . . . . . . . . . . . Casting Process . . . . . . . . . . . . . . . . . Applications . . . . . . . . . . . . . . . . . . . Vertical Centrifugal Casting. . . . . . . . . . . . Mold Design . . . . . . . . . . . . . . . . . . . Process Details . . . . . . . . . . . . . . . . .

www.asminternational.org

. . . . . . . . . . . . . . . . 665 . . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

Permanent Mold and Semipermanent Mold Processes . . . . . . . . . . . . . . . . . . . . . . . . John L. Jorstad, J.L.J. Technologies, Inc Permanent Mold Casting . . . . . . . . . . . . . . . . . . Gravity Casting Methods . . . . . . . . . . . . . . Advantages and Disadvantages . . . . . . . . . . Applications . . . . . . . . . . . . . . . . . . . . . . . Casting Design . . . . . . . . . . . . . . . . . . . . . Mold Design . . . . . . . . . . . . . . . . . . . . . . . Mold Materials . . . . . . . . . . . . . . . . . . . . . Mold Coatings. . . . . . . . . . . . . . . . . . . . . . Mold Life . . . . . . . . . . . . . . . . . . . . . . . . . Mold Temperature . . . . . . . . . . . . . . . . . . . Control of Mold Temperature . . . . . . . . . . . Dimensional Accuracy . . . . . . . . . . . . . . . . Surface Finish . . . . . . . . . . . . . . . . . . . . . . Low-Pressure Die Casting . . . . . . . . . . . . . . . . . Conventional Low-Pressure Casting . . . . . . . Counterpressure Casting . . . . . . . . . . . . . . . Vacuum Riserless/Pressure Riserless Casting Low Pressure Counter Gravity Casting . . . . . . . . Mold Filling . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

Die Casting Tooling . . . . . . . . . . . . . . . . . . . . . . . Die Casting Machines and Dies. . . . . . . . . . . . Product Design and the Process. . . . . . . . . . . . Tooling and Metal Flow . . . . . . . . . . . . . . . . . Heat Removal . . . . . . . . . . . . . . . . . . . . . . . . Materials Selection for Tooling . . . . . . . . . . . . Automation in High-Pressure Die Casting . . . . . . . . Die Casting Cycle . . . . . . . . . . . . . . . . . . . . . Automatic Pouring . . . . . . . . . . . . . . . . . . . . . Injection Process . . . . . . . . . . . . . . . . . . . . . . Solidification and Heat Removal . . . . . . . . . . . Casting Extraction and Insert Loading . . . . . . . Automatic Spray . . . . . . . . . . . . . . . . . . . . . . Finishing Automation . . . . . . . . . . . . . . . . . . . Die Casting Internal Quality Check Automation

667 667 669 670 670 671 674 674 675 678 680 680 684

. . . . . . . . . . . . 687 . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . .

High-Pressure Die Casting. . . . . . . . . . . . . . . . . . . . . . . . . John L. Jorstad, J.L.J. Technologies, Inc High-Pressure Die Casting . . . . . . . . . . . . . . . . . . . . . . . . . . Die Casting Alloys and Processes . . . . . . . . . . . . . . . . . Advantages of High-Pressure Die Casting. . . . . . . . . . . . Disadvantages of High-Pressure Die Casting. . . . . . . . . . Product Design for the Process . . . . . . . . . . . . . . . . . . . Metal Injection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Hot Chamber Die Casting . . . . . . . . . . . . . . . . . . . . . . . . . . Melting Process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Injection Components. . . . . . . . . . . . . . . . . . . . . . . . . . Distinctions Between Hot and Cold Chamber Processes . . Gate and Runner Design. . . . . . . . . . . . . . . . . . . . . . . . Temperature Control . . . . . . . . . . . . . . . . . . . . . . . . . . Ejection and Postprocessing . . . . . . . . . . . . . . . . . . . . . Cold Chamber Die Casting . . . . . . . . . . . . . . . . . . . . . . . . . Machine Components . . . . . . . . . . . . . . . . . . . . . . . . . . Process Parameters. . . . . . . . . . . . . . . . . . . . . . . . . . . . Squeeze Casting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Squeeze Casting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Casting and Tooling Design for Squeeze Casting . . . . . . Factors Affecting Squeeze-Cast Products . . . . . . . . . . . . Squeeze-Cast Applications . . . . . . . . . . . . . . . . . . . . . . Commonly Used Alloys, Properties, and Microstructures . Vacuum High-Pressure Die Casting . . . . . . . . . . . . . . . . . . . Conventional Die Casting . . . . . . . . . . . . . . . . . . . . . . . Vacuum Die Casting . . . . . . . . . . . . . . . . . . . . . . . . . . High-Vacuum Die Casting . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . .

689 690 690 691 691 692 693 695 697 697 698 698 699 700 700 704 706 709 709

. . . 713 . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . .

715 715 716 717 717 717 719 719 719 720 721 721 722 724 724 725 727 727 727 728 729 730 732 732 732 732

xv

. . . . . . . . . . . . . . .

. . . . . . . . . . . . . . .

Semisolid Casting . . . . . . . . . . . . . . . . . . . . . . . . . John L. Jorstad, J.L.J. Technologies, Inc Semisolid Casting — Introduction and Fundamentals . Advantages of SSM Processing . . . . . . . . . . . . . SSM Processing Routes . . . . . . . . . . . . . . . . . . Thixocasting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Bar-Making Processes . . . . . . . . . . . . . . . . . . . Billet Sawing and Volume Control . . . . . . . . . . Reheating and Billet Handling. . . . . . . . . . . . . . Rheological Tests . . . . . . . . . . . . . . . . . . . . . . Basics of Semisolid Tooling Design. . . . . . . . . . Casting Process and Parameters. . . . . . . . . . . . . Rheocasting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Thixomolding. . . . . . . . . . . . . . . . . . . . . . . . . . . . . Process Description . . . . . . . . . . . . . . . . . . . . . Emerging Developments . . . . . . . . . . . . . . . . . . Mechanical Properties and Microstructure . . . . .

. . . . . . . . . . . . . . . .

Cast Irons . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Thomas E. Prucha, American Foundry Society Introduction to Cast Irons . . . . . . . . . . . . . . . . . . . Classification of Cast Irons . . . . . . . . . . . . . . . Composition of Cast Irons . . . . . . . . . . . . . . . Solidification of Cast Irons . . . . . . . . . . . . . . . Graphite Morphologies . . . . . . . . . . . . . . . . . . Matrix Structure Development. . . . . . . . . . . . . Physical Properties of Cast Irons . . . . . . . . . . . Thermal Properties. . . . . . . . . . . . . . . . . . . . . Conductive Properties. . . . . . . . . . . . . . . . . . . Magnetic Properties . . . . . . . . . . . . . . . . . . . . Acoustic Properties . . . . . . . . . . . . . . . . . . . . Heat Treatment . . . . . . . . . . . . . . . . . . . . . . . Welding of Cast Irons . . . . . . . . . . . . . . . . . . Machining and Grinding . . . . . . . . . . . . . . . . . Coatings . . . . . . . . . . . . . . . . . . . . . . . . . . . . Materials Selection . . . . . . . . . . . . . . . . . . . . Cast Iron Foundry Practices . . . . . . . . . . . . . . . . . . Cast Iron Melting Practice . . . . . . . . . . . . . . . Gray Iron Foundry Practice. . . . . . . . . . . . . . . Ductile Iron Foundry Practice . . . . . . . . . . . . . Compacted Graphite Iron Foundry Practice. . . . Malleable Iron Foundry Practice . . . . . . . . . . . Foundry Practice for High-Nickel Ductile Irons Foundry Practice High-Silicon Ductile Irons . . . Foundry Practice for High-Silicon Gray Irons . . Foundry Practice for High-Alloy White Irons . . Gray Iron Castings . . . . . . . . . . . . . . . . . . . . . . . . Classes of Gray Iron . . . . . . . . . . . . . . . . . . . Applications . . . . . . . . . . . . . . . . . . . . . . . . . Castability. . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . .

. . . . . . . . . . . . . . .

. . . . . . . . . . . . . . .

. . . . . . . . . . . . . . .

. . . . . . . . . . . . . . .

. . . . . . . . . . . . . . .

. . . . . . . . . . . . . . .

. . . . . . . . . . . . . . .

734 734 735 737 740 742 747 747 748 749 751 753 754 755 757

. . . . . . . . . 759 . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . .

761 761 762 764 764 766 767 768 769 770 773 773 777 777 778 780

. . . . . . . . . . 783 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

785 785 788 791 792 794 794 795 797 800 801 802 805 806 809 809 812 812 815 820 826 829 830 830 830 831 835 835 835 835

© 2008 ASM International. All Rights Reserved. ASM Handbook Volume 15: Casting (#05115G)

Microstructure . . . . . . . . . . . . . . . . . . . . . . . . . . Section Sensitivity . . . . . . . . . . . . . . . . . . . . . . . Prevailing Sections. . . . . . . . . . . . . . . . . . . . . . . Test Bar Properties . . . . . . . . . . . . . . . . . . . . . . Specifications . . . . . . . . . . . . . . . . . . . . . . . . . . Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Fatigue Limit in Reversed Bending . . . . . . . . . . . Pressure Tightness . . . . . . . . . . . . . . . . . . . . . . . Impact Resistance . . . . . . . . . . . . . . . . . . . . . . . Machinability . . . . . . . . . . . . . . . . . . . . . . . . . . Wear . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Resistance to Wear . . . . . . . . . . . . . . . . . . . . . . Elevated-Temperature Properties . . . . . . . . . . . . . Dimensional Stability . . . . . . . . . . . . . . . . . . . . . Effect of Shakeout Practice. . . . . . . . . . . . . . . . . Alloying to Modify As-Cast Properties. . . . . . . . . Heat Treatment . . . . . . . . . . . . . . . . . . . . . . . . . Physical Properties. . . . . . . . . . . . . . . . . . . . . . . Ductile Iron Castings . . . . . . . . . . . . . . . . . . . . . . . . Classes and Grades of Ductile Iron . . . . . . . . . . . Factors Affecting Mechanical Properties. . . . . . . . Hardness Properties . . . . . . . . . . . . . . . . . . . . . . Tensile Properties . . . . . . . . . . . . . . . . . . . . . . . Shear and Torsional Properties . . . . . . . . . . . . . . Compressive Properties . . . . . . . . . . . . . . . . . . . Fatigue Properties . . . . . . . . . . . . . . . . . . . . . . . Fracture Toughness . . . . . . . . . . . . . . . . . . . . . . Physical Properties. . . . . . . . . . . . . . . . . . . . . . . Austempered Ducfile Iron (ADI) . . . . . . . . . . . . . Compacted Graphite Iron Castings . . . . . . . . . . . . . . . Graphite Morphology . . . . . . . . . . . . . . . . . . . . . Chemical Composition . . . . . . . . . . . . . . . . . . . . Castability. . . . . . . . . . . . . . . . . . . . . . . . . . . . . Mechanical Properties and Corrosion Resistance . . Applications . . . . . . . . . . . . . . . . . . . . . . . . . . . Malleable Iron Castings. . . . . . . . . . . . . . . . . . . . . . . Melting Practices . . . . . . . . . . . . . . . . . . . . . . . . Microstructure . . . . . . . . . . . . . . . . . . . . . . . . . . Current Production Technologies . . . . . . . . . . . . . Applications . . . . . . . . . . . . . . . . . . . . . . . . . . . Ferritic Malleable Iron . . . . . . . . . . . . . . . . . . . . Pearlitic and Martensitic Malleable Iron . . . . . . . . White Iron and High-Alloyed Iron Castings. . . . . . . . . Nickel-Chromium White Irons . . . . . . . . . . . . . . High-Chromium White Irons. . . . . . . . . . . . . . . . High-Alloy Graphitic Irons . . . . . . . . . . . . . . . . . . . . High-Silicon Irons for High-Temperature Service . Austenitic Nickel-Alloyed Gray and Ductile Irons . High-Silicon Irons for Corrosion Resistance . . . . .

www.asminternational.org

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Steel Castings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Raymond W. Monroe, Steel Founders’ Society of America; Thomas E. Prucha, American Foundry Society Steel Ingot Casting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Pouring Method. . . . . . . . . . . . . . . . . . . . . . . . . . . . . Ingot Design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Hot Tops . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Ingot Molds and Stools . . . . . . . . . . . . . . . . . . . . . . . Ingot Solidification. . . . . . . . . . . . . . . . . . . . . . . . . . . Ingot Stripping . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Solidification Simulation. . . . . . . . . . . . . . . . . . . . . . . Steel Continuous Casting. . . . . . . . . . . . . . . . . . . . . . . . . . Historical Aspects of Continuous Casting of Steel . . . . . General Description of the Process . . . . . . . . . . . . . . . Plant Layout . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Near-Net Shape Casting . . . . . . . . . . . . . . . . . . . . . . . Tundish Metallurgy . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

836 837 838 839 840 841 843 844 846 846 848 848 849 849 850 851 851 853 ... 856 857 859 859 861 861 862 865 867 869 872 872 872 873 874 879 884 884 885 887 888 889 890 896 896 899 904 904 906 907

. . . . 909 . . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

911 911 912 913 913 914 916 916 918 918 918 919 919 920

Pouring Stream Protection . . . . . . . . . . . . . . . . . . Molds for Continuous Casting of Steel. . . . . . . . . . Productivity Improvements . . . . . . . . . . . . . . . . . . Quality Improvements . . . . . . . . . . . . . . . . . . . . . Horizontal Continuous Casting . . . . . . . . . . . . . . . Future Developments . . . . . . . . . . . . . . . . . . . . . . Shape Casting of Steel . . . . . . . . . . . . . . . . . . . . . . . . Castability of Steel. . . . . . . . . . . . . . . . . . . . . . . . Melting Furnaces . . . . . . . . . . . . . . . . . . . . . . . . . Metlting Practice . . . . . . . . . . . . . . . . . . . . . . . . . Pouring Practice . . . . . . . . . . . . . . . . . . . . . . . . . Foundry Factors in Design . . . . . . . . . . . . . . . . . . Design Factors. . . . . . . . . . . . . . . . . . . . . . . . . . . Avoidance of Hot Spots . . . . . . . . . . . . . . . . . . . . Thin-Wall Steel Castings . . . . . . . . . . . . . . . . . . . Tolerances . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Size of Cores . . . . . . . . . . . . . . . . . . . . . . . . . . . Casting in Graphite Molds . . . . . . . . . . . . . . . . . . Steel Castings Properties . . . . . . . . . . . . . . . . . . . . . . . Carbon Steels . . . . . . . . . . . . . . . . . . . . . . . . . . . Low-Alloy Steels . . . . . . . . . . . . . . . . . . . . . . . . . Wear-Resistant Steels . . . . . . . . . . . . . . . . . . . . . . Corrosion-Resistant Steels. . . . . . . . . . . . . . . . . . . Heat-Resistant Steels . . . . . . . . . . . . . . . . . . . . . . Common Alloys . . . . . . . . . . . . . . . . . . . . . . . . . Carbon and Low-Alloy Mechanical Properties . . . . Mechanical Properties of Wear-Resisting Steels . . . Mechanical Properties of Corrosion-Resisting Steels Mechanical Properties of Heat-Resisting Steels . . . . Ferrite in Cast Stainless Steels . . . . . . . . . . . . . . . Carbon and Low-Alloy Heat Treatments . . . . . . . . High-Alloy Heat Treatment. . . . . . . . . . . . . . . . . . Corrosion-Resistant Applications . . . . . . . . . . . . . . Heat-Resistant Applications . . . . . . . . . . . . . . . . . Selection and Evaluation of Steel Castings . . . . . . . . . . Design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Manufacturing . . . . . . . . . . . . . . . . . . . . . . . . . . . Nondestructive Examination . . . . . . . . . . . . . . . . . Purchasing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Research and Development . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

921 922 923 924 924 924 926 926 928 930 933 934 936 938 943 946 947 948 949 949 950 950 951 952 953 953 960 961 966 968 970 971 972 973 975 975 980 982 986 986

Casting of Nonferrous Alloys . . . . . . . . . . . . . . . . . . . . . Nonferrous Casting — An Introduction . . . . . . . . . . . . . . . Nickel Alloys . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Titanium Alloys . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Lead. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Dross, Melt Loss, and Fluxing of Light Alloy Melts. . . . . . Dross Formation . . . . . . . . . . . . . . . . . . . . . . . . . . . Influence of Melter Type on Dross Generation . . . . . . Influence of Charge Materials on Melt Loss . . . . . . . . Influence of Operating Practices on Melt Loss . . . . . . Economic Implications of Dross . . . . . . . . . . . . . . . . In-Plant Enhancement or Recovery of Dross Metallics . Ways to Reduce Dross Formation . . . . . . . . . . . . . . . Aluminum Alloy Ingot Casting and Continuous Processes. . Ingot Forms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Molten-Metal Processing . . . . . . . . . . . . . . . . . . . . . Ingot Casting Processes . . . . . . . . . . . . . . . . . . . . . . Continuous Processes . . . . . . . . . . . . . . . . . . . . . . . . Solidification in the DC Process . . . . . . . . . . . . . . . . Postsolidification Processes . . . . . . . . . . . . . . . . . . . . Safety. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Aluminum Shape Casting . . . . . . . . . . . . . . . . . . . . . . . . Melting and Melt Handling . . . . . . . . . . . . . . . . . . . . Casting Process Selection . . . . . . . . . . . . . . . . . . . . . Pressure Die Casting of Aluminum . . . . . . . . . . . . . . Premium Engineered Castings . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . .

987 989 989 989 990 992 992 993 994 994 998 998 999 1001 1001 1001 1002 1004 1006 1007 1007 1009 1009 1010 1013 1016

xvi

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

© 2008 ASM International. All Rights Reserved. ASM Handbook Volume 15: Casting (#05115G)

Copper Continuous Casting . . . . . . . . . . . . . . History . . . . . . . . . . . . . . . . . . . . . . . . . Vertical Continuous Casting . . . . . . . . . . Horizontal Continuous Casting . . . . . . . . Strip Casting . . . . . . . . . . . . . . . . . . . . . Wheel Casting . . . . . . . . . . . . . . . . . . . . Ohno Continuous Casting Process . . . . . . Casting of Copper and Copper Alloys . . . . . . . Castability. . . . . . . . . . . . . . . . . . . . . . . Melting and Melt Control . . . . . . . . . . . . Fluxing of Copper Alloys . . . . . . . . . . . . Degassing of Copper Alloys . . . . . . . . . . Deoxidation of Copper Alloys . . . . . . . . . Grain Refining of Copper Alloys . . . . . . . Filtration of Copper Alloys . . . . . . . . . . . Melt Treatments for Group I to III Alloys Production of Copper Alloy Castings . . . . Casting Process Selection . . . . . . . . . . . . Gating . . . . . . . . . . . . . . . . . . . . . . . . . Feeding . . . . . . . . . . . . . . . . . . . . . . . . Casting of Zinc Alloys . . . . . . . . . . . . . . . . . Control of Alloy Composition . . . . . . . . . Melt Processing . . . . . . . . . . . . . . . . . . . Die Casting of Zinc Alloys . . . . . . . . . . . Other Casting Processes for Zinc Alloys. .

www.asminternational.org

. . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . .

Nonferrous Alloy Castings . . . . . . . . . . . . . . . . . . . . . . . Mahi Sahoo, CANMET Materials Technology Laboratory Aluminum and Aluminum Alloy Castings . . . . . . . . . . . . . Chemical Compositions . . . . . . . . . . . . . . . . . . . . . . Effects of Alloying and Impurity Elements . . . . . . . . . Structure Control . . . . . . . . . . . . . . . . . . . . . . . . . . . Modification and Refinement of Aluminum-Silicon Alloys . . . . . . . . . . . . . . . . . Foundry Alloys for Specific Casting Applications . . . . Heat Treatment . . . . . . . . . . . . . . . . . . . . . . . . . . . . Welding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Properties of Aluminum Casting Alloys . . . . . . . . . . . Copper and Copper Alloy Castings . . . . . . . . . . . . . . . . . . Alloying Additions. . . . . . . . . . . . . . . . . . . . . . . . . . Alloying Systems and Specifications . . . . . . . . . . . . . Physical Metallurgy . . . . . . . . . . . . . . . . . . . . . . . . . Types of Copper Castings. . . . . . . . . . . . . . . . . . . . . Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Heat Treatment . . . . . . . . . . . . . . . . . . . . . . . . . . . . Welding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Cupronickels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Zinc and Zinc Alloy Castings . . . . . . . . . . . . . . . . . . . . . Alloying Additions. . . . . . . . . . . . . . . . . . . . . . . . . . Physical Metallurgy . . . . . . . . . . . . . . . . . . . . . . . . . Die Casting and Aging . . . . . . . . . . . . . . . . . . . . . . . Specialty Alloys . . . . . . . . . . . . . . . . . . . . . . . . . . . Finishing of Die Castings . . . . . . . . . . . . . . . . . . . . . Metalworking and Machining . . . . . . . . . . . . . . . . . . Joining . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Applications for Zinc Die Castings . . . . . . . . . . . . . . Magnesium and Magnesium Alloys . . . . . . . . . . . . . . . . . Magnesium Alloys . . . . . . . . . . . . . . . . . . . . . . . . . . General Applications . . . . . . . . . . . . . . . . . . . . . . . . Melting Furnaces and Auxiliary Pouring Equipment . . Melting Procedures and Process Parameters . . . . . . . . Sand Casting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Investment Casting. . . . . . . . . . . . . . . . . . . . . . . . . . Permanent Mold Casting . . . . . . . . . . . . . . . . . . . . . Die Casting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . .

1019 1019 1020 1022 1023 1024 1024 1026 1026 1027 1029 1032 1035 1037 1037 1038 1039 1040 1041 1044 1049 1049 1049 1051 1054

. . . . 1057 . . . .

. . . .

. . . .

. . . .

1059 1059 1059 1063

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1066 1069 1071 1077 1077 1085 1085 1087 1088 1091 1091 1091 1091 1091 1095 1095 1096 1097 1097 1098 1098 1099 1097 1099 1100 1102 1102 1103 1106 1109 1109 1109

Direct Chill Casting . . . . . . . . . . . . . . . . . . . . . High-Temperature Alloys . . . . . . . . . . . . . . . . . Welding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Cobalt and Cobalt Alloy Castings. . . . . . . . . . . . . . . Physical Metallurgy . . . . . . . . . . . . . . . . . . . . . Foundry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Postcasting Treatment . . . . . . . . . . . . . . . . . . . Applications . . . . . . . . . . . . . . . . . . . . . . . . . . Nickel and Nickel Alloy Castings. . . . . . . . . . . . . . . Compositions . . . . . . . . . . . . . . . . . . . . . . . . . Structure and Property Correlations . . . . . . . . . . Melting Practice and Metal Treatment . . . . . . . . Foundry Practice . . . . . . . . . . . . . . . . . . . . . . . Pouring Practice . . . . . . . . . . . . . . . . . . . . . . . Gating Systems . . . . . . . . . . . . . . . . . . . . . . . . Risers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Welding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Heat Treatment . . . . . . . . . . . . . . . . . . . . . . . . Specific Applications . . . . . . . . . . . . . . . . . . . . Titanium and Titanium Alloy Castings . . . . . . . . . . . Historical Perspective of Casting Technology . . . Physical Metallurgy . . . . . . . . . . . . . . . . . . . . . Alloy Ti-6Al-4V . . . . . . . . . . . . . . . . . . . . . . . Casting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Postcasting Practice . . . . . . . . . . . . . . . . . . . . . Product Applications . . . . . . . . . . . . . . . . . . . . Zirconium and Zirconium Alloy Castings . . . . . . . . . Zirconium Reactivity Considerations . . . . . . . . . Melting Processes . . . . . . . . . . . . . . . . . . . . . . Casting Processes. . . . . . . . . . . . . . . . . . . . . . . Processing. . . . . . . . . . . . . . . . . . . . . . . . . . . . Physical Properties of Zirconium Castings . . . . . Impact Properties of Zirconium Castings . . . . . . Synthesis and Processing of Cast Metal-Matrix Composites and Their Applications . . . . . . . . . . Classification of MMCs . . . . . . . . . . . . . . . . . . Solidification Processing of MMCs and Possible Effects of Reinforcement on Solidification. . Property Motivation for Using MMCs . . . . . . . . Components of MMCs Currently in Use. . . . . . . Development of Select Cast MMCs . . . . . . . . . . Research Imperatives . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1112 1112 1112 1114 1114 1116 1117 1117 1119 1119 1121 1123 1124 1124 1124 1125 1125 1125 1126 1128 1129 1129 1132 1134 1137 1138 1143 1143 1144 1144 1145 1147 1147

. . . . . . . . 1149 . . . . . . . . 1149 . . . . .

. . . . .

. . . . .

Quality Assurance, Nondestructive Evaluation, and Failure Analysis. . . . . . . . . . . . . . . . . . . . . . . Babu DasGupta, National Science Foundation Approaches to Measurement of Metal Quality. . . . . . Metal Cleanliness Assessment Techniques . . . . . . . . Nondestructive Testing of Components . . . . . . . . . . . . . . Liquid Penetrant Inspection. . . . . . . . . . . . . . . . . . . Radiographic Inspection . . . . . . . . . . . . . . . . . . . . . Fluoroscopic Inspection and Automated Defect Recognition . . . . . . . . . . . . . . . . . . . . . Ultrasonic Inspection . . . . . . . . . . . . . . . . . . . . . . . Eddy Current Inspection . . . . . . . . . . . . . . . . . . . . . Process-Controlled Resonant Testing (Contributed by Quasar International, Inc.) . . . . . . . . . . . . . . . . Leak Test . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Electrical Conductivity Measurement . . . . . . . . . . . . Casting Fracture Characteristics . . . . . . . . . . . . . . . . . . . General Fracture Examination . . . . . . . . . . . . . . . . . Fracture Features . . . . . . . . . . . . . . . . . . . . . . . . . . Casting Failure Analysis Techniques and Case Studies . . . Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Background Information . . . . . . . . . . . . . . . . . . . . .

xvii

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

1150 1155 1156 1160 1162

. . . . . 1165 . . . . .

. . . . .

. . . . .

. . . . .

. . . . .

1167 1167 1174 1174 1174

. . . . . 1174 . . . . . 1174 . . . . . 1175 . . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

1175 1177 1177 1178 1178 1179 1183 1183 1183

© 2008 ASM International. All Rights Reserved. ASM Handbook Volume 15: Casting (#05115G)

Planning and Sample Selection . . . . . . . . . . . . . . . . Preliminary Examination . . . . . . . . . . . . . . . . . . . . Material Evaluation . . . . . . . . . . . . . . . . . . . . . . . . Data Analysis and Report Preparation . . . . . . . . . . . Case Studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Common Defects in Various Casting Processes . . . . . . . . Case Studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Repair Welding of Castings . . . . . . . . . . . . . . . . . . . . . . Repair Welding of Ferrous Castings. . . . . . . . . . . . . Repair Welding of Nonferrous Castings . . . . . . . . . . Quality Planning Tools and Procedures . . . . . . . . . . . . . . Advanced Product Quality Planning (APQP). . . . . . . Maintenance, Repair, Alterations, and Storage of Patterns and Tooling. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

www.asminternational.org

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

1184 1184 1184 1187 1187 1192 1199 1203 1203 1205 1207 1207

. . . . . 1213

Ownership . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Maintenance Responsibility. . . . . . . . . . . . . . . . . . . . . Common Failure Mechanisms . . . . . . . . . . . . . . . . . . . Maintenance Interval . . . . . . . . . . . . . . . . . . . . . . . . . Maintenance Records . . . . . . . . . . . . . . . . . . . . . . . . . Preventive Maintenance Program. . . . . . . . . . . . . . . . . Repair . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Alterations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Storage—Physical Protection, Cataloging, and Tracking. Reference Information. . . Metric Conversion Guide. . Abbrieviations & Symbols . Index. . . . . . . . . . . . . . . .

xviii

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

1213 1213 1213 1213 1214 1214 1214 1214 1215

. . . .

. . . .

. . . .

1217 1219 1222 1225

ASM Handbook, Volume 15: Casting ASM Handbook Committee, p 3-15 DOI: 10.1361/asmhba0005186

Copyright © 2008 ASM International® All rights reserved. www.asminternational.org

History and Trends of Metal Casting Thomas E. Prucha, American Foundry Society Daniel Twarog, North American Die Casting Association Raymond W. Monroe, Steel Founders’ Society of America

CASTING is one of the oldest manufacturing methods known to humankind and a very direct method of producing metal parts. It involves pouring molten metal into a cavity that is close to the final dimensions of the finished form, and it is used to manufacture many types of complex components of any metal, ranging in weight from less than an ounce to single parts weighing several hundred tons. Typical uses of cast components include infrastructure and structural components, water distribution systems (pipes, pumps, and valves), automotive Table 1

components (engine blocks, brakes, steering and suspension components, etc.), prosthetics, jewelry, and gas turbine engine hardware. In one study (Ref 1), castings were estimated to be used in 90% or more of all manufactured goods and in all capital goods machinery used in manufacturing. Some of the major market areas of castings are listed in Table 1 (Ref 2). Economically, casting processes are capable of producing highly reliable, cost-effective components ranging from low-volume, singlepart prototype production runs to economies

of scale for millions of parts. In terms of component design, casting also permits the formation of streamlined, intricate, integral parts with varied shape and size capabilities for designers. Metal casting is very flexible in terms of configuration design, and if a pattern can be made for a part, it can be cast. For example, the flexibility of metal casting, particularly sand molding, may permit the use of difficult design techniques, such as undercuts and curved or reflex contours that are not possible with other high-production processes. Tapered

      

 Other major markets include machine tools, construction

Major markets for metal castings

Ferrous castings Gray iron  Ingot molds  Construction castings  Motor vehicles  Farm equipment  Engines  Refrigeration and heating  Construction machinery  Valves  Soil pipe  Pumps and compressors  Pressure pipe  Other major markets include machine tools, mechanical power transmission equipment, hardware, home appliances, mining machinery, and oil and natural gas pumping and processing equipment(a) Malleable iron  Motor vehicles  Valves and fittings  Construction machinery  Railroad equipment  Engines  Mining equipment  Hardware  Other major markets include heating and refrigeration, motors and generators, fasteners, ordnance, chains, machine tools, and general industrial machinery(a) Ductile iron  Pressure pipe  Motor vehicles

Farm machinery Engines Pumps and compressors Valves and fittings Metalworking machinery Construction machinery Other major markets include textile machinery, woodworking and paper machinery, mechanical power and transmission equipment, motors and generators, refrigeration and heating equipment, and air conditioning(a)

Steel  Railroad equipment  Construction equipment  Mining machinery  Valves and fittings  General and special industrial machinery  Motor vehicles  Metalworking machinery  Other major markets include steel manufacturing, spring goods, heating and air conditioning, recreation equipment, industrial materials handling equipment, ships and boats, and aircraft and aerospace(a) Nonferrous castings Aluminum  Auto and light truck  Aircraft and aerospace  Other transportation  Engines  Household appliances  Office machinery  Power tools  Refrigeration, heating, and air conditioning

equipment, mining equipment, farm machinery, electronic and communication equipment, power systems, and motors and generators(a) Magnesium Power tools Sporting goods Anodes Automotive Other major markets include office machinery, health care, and aircraft and aerospace(a)

    

Copper base Valves and fittings Plumbing brass goods Electrical equipment Pumps and compressors Power transmission equipment General machinery Transportation equipment Other major markets include chemical processing, utilities, desalination, and petroleum refining(a)

       

Zinc Automotive Building hardware Electrical components Machinery Household appliances Other major markets include scientific instruments, medical equipment, radio and television equipment, audio components, toys, and sporting goods(a)

     

(a)Ranked in approximate order of tonnage shipped, but in some cases, the total of other major markets is larger as a whole than the individual markets listed. Source: Ref 2

4 / Applications and Specification Guidelines sections with thickened areas for bosses and generous fillets are routine. The inherent design freedom of metal casting also allows the designer to combine what would otherwise be several parts of a fabrication into a single piece. This is significant when exact alignment must be held, as in high-speed machinery, machine tool parts, or engine end plates and housings that carry shafts. Various surface finishes on a part are also possible, ranging from a rougher as-cast surface of sand-molded casting to the smoother surfaces obtained through shell molding, investment casting, or other casting methods. In general, casting can provide various functional advantages for component design that include:  Design both internal and external contours

independently to almost any requirement

 Place metal in exact locations where it is

needed for rigidity, wear, corrosion, or maximum endurance under dynamic stress  Produce a complex part as a single, dependable unit  Readily achieve an attractive appearance This article briefly introduces the history of metal casting, general capabilities and methods of casting, and industry trends. In terms of value and volume, metal casting ranks second

only to steel rolling in the metal-producing industry. Metal casting is one of the ten largest industries when rated on a value-added basis, and worldwide casting shipments have continued to increase in recent years. Production statistics for 2005 are summarized in Table 2 (Ref 3), with comparison to 2004 production statistics from Modern Casting’s “Annual Census of World Casting Production.”

History of Metal Casting The casting of metal is a prehistoric technology. Exactly when the casting of metals began is not known, but it appears relatively late in the archaeological record. Archaeologists give the name Chalcolithic to the period in which metals were first being mastered, and they date this period, which immediately preceded the Bronze Age, very approximately to between 5000 and 3000 B.C. (Table 3, Ref 4). There were many earlier fire-using technologies, collectively called pyrotechnology, that provided a basis for the development of metal casting. Among these were the heat treatment of stone to make it more workable, the burning of lime to make plaster, and the firing of clay to produce ceramics. At first, it did not include

smelting, because the metal of the earliest castings appears to have been native. Cast iron appeared in China in approximately 600 B.C. Its use was not limited to strictly practical applications, and there are many examples of Chinese cast iron statuary. Most Chinese cast irons were unusually high in phosphorus and, because coal was often used in smelting, high in sulfur as well. These irons therefore have melting points that are similar to those of bronze, and when molten are unusually fluid. The iron castings, like the Chinese cast bronzes, are often remarked upon for the thinness of their wall sections. From these early beginnings, development and application of metal casting continued (Table 4, Ref 5). During the wars of medieval times, metalcasters produced cannons. In peace, they recast the metal into bells. In the Middle Ages, church bells were cast by priests, abbots, or bishops who also were trained metal founders. As the metal was melted, the brethren stood around the furnace intoning psalms and prayers. The molten metal was then blessed, and divine protection was asked for the bell, which usually bore the name of a saint (Ref 5). During the 1800s, there were significant advances in technology. Metallurgy also began

Table 2 World casting production for 2004–2005 40th census of world casting production for 2005, metric tons Country

Austria Belgium Brazil Canada China Croatia Czech Republic Denmark Finland France Germany Great Britain Hungary India Italy Japan Korea Lithuania Mexico Netherlands(a) Norway Poland Portugal Romania Russia Slovenia(a) Slovakia South Africa Spain Sweden Switzerland Taiwan Thailand Turkey Ukraine (i) United States Totals

Gray iron

47,500 70,700(a) 2,460,000(b)(d) 483,000(a)(b)(d) 12,303,963 25,306 269,493 46,932 47,817 896,500 2,469,058 531,000 35,307 4,116,000 923,700 2,782,509 960,100 14,300 600,000 63,100 16,348 420,700 27,285 81,578 3,480,000 75,900 41,520 160,000 491,600 171,700(b) 26,400 827,932 170,000 567,000 626,610 4,457,905 40,788,763

Ductile iron

131,000 15,600(a)(b) ... ... 5,838,753 16,252 49,524 40,333 60,123 957,700 1,487,234 362,000 32,200 618,000 512,200 1,919,435 565,200 200 270,000 78,241 49,207 112,500 69,792 10,500 720,000 27,800(b) 8,760 65,000 551,600 83,300 40,400 220,413 70,000 327,000 40,000 4,241,088 19,591,355

Malleable iron

17,500 ... ... ... 514,129 50 5,435 0 0 1,400 52,713 12,000 10 56,000 8,400 57,851 46,500 ... 2,000 6,209 0 21,740 0 926 280,000 ... ... 10,000 17,400 ... ... ... 30,000 8,000 10,000 75,296 1,233,559

Steel

18,000 42,100 293,891 117,600(a) 3,224,374 2,219 109,680 0 18,532 113,700 200,409 112,000 5,385 805,000 73,324 276,589 149,600 30 75,000 438 3,723 59,560 12,096 37,900 1,200,000 22,700 4,300 136,000 83,200 23,400 ... 73,619 30,000 125,000 266,060 1,287,295 9,002,724

Copper base

5,459 ... ... 18,585(a) 416,097 1,120 1,724 1,126 3,999 26,264 84,386 15,000 2,110 0 16,800 97,794 23,200 5 180,000 ... 4,667 6,300 8,700 2,700 160,000 6,727(f) 2,160 14,900 7,886 11,200 2,689 41,959 28,600 16,000 11,000 292,113 1,511,270

Aluminum

103,347 19,428(c) 240,457(e) 324,973 1,886,000(c) 15,038 80,252(c) 4,000 6,565 318,445(c) 727,139 206,000(c) 75,720 516,000 857,500 1,477,349(c) 144,300 56 660,000 ... 22,926 174,300 20,250 20,000 920,000 30,183 26,260 56,000 139,312(c) 44,400 17,785 250,366 100,000 66,500 20,500 2,080,174 11,718,025

Magnesium

6,580 ... ... ... ... 0 ... 0 0 ... 27,283 ... 3,050 0 12,000 ... 10,300(f) ... 500 ... 0 30 150 10 70,000 ... ... 0 ... 1,700 ... 6,427 ... 500 ... 100,697 239,227

Zinc

12,997 ... ... ... 237,889 502 2,656 1 434 25,066 56,439 21,000 3,050 0 70,000 36,216 ... ... 350 ... 0 8,600 900 700 20,000 ... 1,800 2,600 16,051 5,900 1,616 69,316 16,900 12,700 ... 312,978 936,661

Other nonferrous

... ... ... ... ... 1,300 279 5.022 0 3,461 3,346 2,000 0 0 67,600 8,102 0 ... ... ... ... 770 120 5 50,000 ... 5 0 433 ... ... 4,418 ... ... ... 48,987 195,848

Total

342,383 124,507(a) 2,968,600 863,373(a) 24,421,205 61,787 519,043 97,414 137,470 2,342,536 5,108,007 1,261,000 156,832 6,111,000 2,541,524 6,655,845 1,899,200 14,591 1,787,850 147,988(g) 96,871 804,500 139,293 154,319 7,620,000(h) 163,310 84,805 444,500 1,307,482 341,600 88,890 1,494,450 445,500 1,122,700 974,170 12,896,533 85,741,078

Totals for 2004

325,205 124,507 2,829,916 863,373 22,420,452 56,449 522,400 95,949 133,325 2,465,617 4,984,473 1,273,000 164,892 4,623,000 2,441,282 6,386,449 1,857,300 18,715 2,185,200 147,988 86,757 804,500 122,750 207,023 6,300,000 163,310 4,800 444,500 1,309,249 321,100 104,077 1,451,768 235,850 982,000 974,170 12,314,121 79,745,467

Note: “. . .” indicates unreported data. (a) 2004 tonnage. (b) Includes malleable iron. (c) Includes magnesium. (d) Includes ductile iron. (e) Includes all nonferrous. (f) Includes zinc. (g) Includes only ferrous. (h) Includes 720,000 tons of special iron. (i) 2002 tonnage. Source: Modern Casting

History and Trends of Metal Casting / 5 Table 3

Chronological list of developments in the use of materials

Date

Development

9000 B.C. 6500 B.C. 5000–3000 B.C.

3000–1500 B.C.

3000–2500 B.C. 2500 B.C. 2400–2200 B.C. 2000 B.C. 1500 B.C.

Earliest metal objects of wrought native copper Earliest life-size statues, of plaster Chalcolithic period: melting of copper; experimentation with smelting Bronze Age: arsenical copper and tin bronze alloys Lost-wax casting of small objects Granulation of gold and silver and their alloys Copper statue of Pharoah Pepi I Bronze Age Iron Age (wrought iron)

Location

Near East Jordan Near East

Near East

Near East Near East Egypt Far East Near East

Date

700–600 B.C. 600 B.C. 224 B.C.

Development

Etruscan dust granulation Cast iron Colossus of Rhodes destroyed 200–300 A.D. Use of mercury in gilding (amalgam gilding) 1200–1450 A.D. Introduction of cast iron (exact date and place unknown) Circa 1122 A.D. Theophilus’s On Divers Arts, the first monograph on metalworking written by a craftsman 1252 A.D. Diabutsu (Great Buddha) cast at Kamakura Circa 1400 A.D. Great Bell of Beijing cast 16th century Sand introduced as mold material

Location

Italy China Greece

Date

Development

1709

Cast iron produced with coke as fuel, Coalbrookdale Boring mill or cannon developed Great Bell of the Kremlin cast Cast steel developed by Benjamin Huntsman Cast iron used as architectural material, Ironbridge Gorge Zinc statuary Electrodeposition of copper

1715 Roman world Europe

1735 1740

Germany

1779

Japan

1826 1838

China France

1884

Electrolytic refining of aluminum

Location

England

Switzerland Russia England England

France Russia, England United States, France

Source: Ref 4

Table 4

Timeline of casting developments to 1899

B.C. 9000 B.C.—Earliest metal objects of wrought native copper are produced in the Near East. 3200 B.C.—The oldest casting in existence, a copper frog, is cast in Mesopotamia. 3000 B.C.—Early metalcasters cast bronze tools and weapons in permanent stone molds. 3000–2500 B.C.—Small objects are cast via lost-wax (investment casting) process in the Near East. 1500 B.C.—Wrought iron is discovered in the Near East. 600 B.C.—The first cast iron object, a 270 kg (600 lb) tripod, is cast by the Chinese. 233 B.C—Iron plowshares are cast. 200 B.C.—Oldest iron castings still in existence are produced during the Han Dynasty. A.D. 500—Cast crucible steel is produced in India. 1200s—Loam or sweep molding is used by European metalcasters to cast bells for cathedrals. 1252—The colossal statue, the Great Buddha at Kamakura, Japan, is cast in high-lead tin bronze. The project began in the 700s, and its head alone weighed 140 tons. 1313—The first cannon is cast in bronze by a monk in the city of Ghent. 1400s—During the siege of Constantinople, heavy guns are cast from bronze “on the spot,” virtually under the walls of the besieged city. Movable, cast lead type for printing presses revolutionizes the world’s methods of communication. 1480—Vannoccio Biringuccio (1480–1539), the first true metalcaster and the “father of the metal casting industry,” is born. The founder of the Vatican, his De La Pirotechnia is the first written account of proper metal casting practice. 1500s—Sand is introduced as a molding material in France. 1600s 1612—Mined from under the sea, seacoal is mentioned for the first time by German metalcaster and inventor Simon Sturtevant. 1619—North America’s first iron furnace is built at Falling Creek, Va., on a branch of the James River, 100 km (60 miles) from the Jamestown colony. Three years later, Native Americans destroy it during a raid. 1645—Earliest recorded use of the term foundry appears in the Oxford English Dictionary in its variant “founderie.” 1646—America’s first iron metal casting facility (and second industrial plant), Saugus Iron Works, near Boston, pours the first American metal casting, the Saugus pot. The Saugus River site was selected by Richard Leader and was built to produce iron products for Massachusetts and England. 1661—First U.S. copper deposits are discovered by Gov. Winthrop in Middletown, Conn. Source: Ref 5; see also Selected References

1700s 1709—Two important developments by Abraham Darby from Coalbrookdale, England, improve casting methods. He developed the first true metal casting facility flask to modernize molding practices (which had been carried out in pits on the floor by use of pattern boards tied together or in crude box frames). He would later initiate the use of coke as a furnace fuel for iron production. 1722—A.F. de Reamur, recognized as the world’s first metallurgical chemist, develops malleable iron, known today as European whiteheart malleable. 1750—Benjamin Huntsman reinvents the cast crucible steel process in England, a process that disappeared after first being developed in India. The English parliament prohibits the refining of pig iron or the casting of iron in the American colonies, contributing to the start of the American Revolution. 1756—Coalbrookdale’s Richard Reynolds oversees the invention of cast iron tram-road rails, replacing wooden rails. 1775—Revolutionary patriot Paul Revere, who operated a bell-and-fittings foundry in Boston, rides from Boston to Lexington warning colonists of the British invasion. 1776—Metalcasters Charles Carroll, James Smith, George Taylor, James Wilson, George Ross, Philip Livingston, and Stephen Hopkins sign the American Declaration of Independence. 1779—First iron bridge ever erected (above England’s Severn River) was cast and constructed at Coalbrookdale Works. 1794—John Wilkinson of England invents the first metalclad cupola furnace, using a steam engine to provide the air blast. 1797—First cast plow in the United States is invented by Charles Newbold of Sauk, N.J. 1800s 1809—Centrifugal casting is developed by A.G. Eckhardt of Soho, England. 1815—The cupola furnace is introduced to the United States (Baltimore). 1817—The first iron water line in the United States, 120 m (400 ft) long, is laid in Philadelphia. 1818—The first U.S. cast steel is produced by the crucible process at historic Valley Forge Foundry. 1825—Aluminum, the most abundant metal in the Earth’s crust, is isolated from aluminum chloride by Denmark’s Hans Oerstad. 1830s—Seth Boyden of Newark, N.J., produces the first blackheart malleable iron in the United States. 1832—Nickel-bronze is produced commercially in England. 1837—The first reliable molding machine on the market is made and used by the S. Jarvis Adams Co., Pittsburgh. 1847—Cast steel guns are made by Krupp Works in Germany. Asa Whitney, Philadelphia, obtains a patent on

a process for annealing chilled-iron car wheels cast with chilled tread and flange. 1847—John Deere commissions Jones and Quiggs Steel Works, Pittsburgh, to cast and roll a steel plow, which it accomplishes at one-half the previous cost of the product. 1849—A manually operated die casting machine is patented to supply rapidly cast lead type for newspapers. 1850—Drop-bottom cupola is developed. 1863—Metallography is developed by Henry C. Sorby of Sheffield, England, enabling metalcasters to polish, etch, and microscopically examine metal surfaces for physical analysis. 1867—James Nasmythe, inventor of the steam hammer, develops a gear-tilted safety ladle to prevent pouring accidents. 1870—Sandblasting is developed for large castings by R.E. Tilghman of Philadelphia. 1874—The Colliau cupola, the first commercially made cupola in America, is introduced. 1876—The first authenticated aluminum castings were produced by Col. William Frishmuth at his Philadelphia foundry. Assembled to produce an engineer’s transit, these castings were made from $1/oz chemically produced aluminum. Manganese-bronze patent is granted to Parsons in England. Tobin Bronze begins developing manganesebronze in the United States. 1880–1887—W.W. Sly, Cleveland, develops the first casting cleaning mill, greatly reducing hand-chipping and grinding to allow a custom-finished product. 1884—The first architectural application of aluminum, a cast aluminum pyramid that was produced by Frishmuth, is mounted on the tip of the Washington Monument. 1886—Charles M. Hall, a 22 year old student at Oberlin College, discovers a process of aluminum reduction through electrolysis. The invention replaced chemical reduction and lowered the metal cost (from $15/lb in 1884 to $0.50/lb in 1890), spurring a new industry of aluminum applications. 1887—Eli Millett invents a core oven for drying small cores in individual drawers. 1890—The first motor-driven mold conveyor is installed, integrating molding, pouring, and cooling operations. 1897—Iowa dentist B.F. Philbrook adapts the lost-wax investment casting process for producing dental inlays, the first nonart application of the process in the modern metal casting age. 1898—Poulson and Hargraves (U.K.) produce the first sand molds bonded with sodium silicate. Germany’s Imperial Navy recommends copper-nickel alloys containing 4–45% Ni for saltwater piping system. 1899—Electric arc furnace, developed by France’s Paul Heroult, begins commercial production.

6 / Applications and Specification Guidelines to achieve prominence in 1889 when nickel was alloyed to make steel stronger. In terms of casting, centrifugal casting was developed by A.G. Eckhardt of Soho, England, in 1809. This method involves the pouring of molten metal into a rapidly rotating metallic mold. The method was soon adopted by the pipe foundries and was first used in Baltimore, Maryland, in 1848. Sir Henry Bessemer, famed for his converter, used centrifugal casting to remove gases and was the first to pour two or more metals into a single rotating mold. The centrifugal casting of steel was first attempted in 1898 at the plant of the American Steel Foundries in St. Louis, Missouri. Railroad car wheels were spun cast in 1901 at a rotation speed of 620 rpm. Following the early development of the centrifugal method, a permanent mold method known as slush casting was introduced. Slush casting is a process in which molten metal is poured into a split metal mold (generally made of bronze) until the mold is filled; then, the mold is immediately inverted, and the metal that is still liquid is allowed to run out. The time required for this casting operation is sufficient to freeze a metal shell in the mold, corresponding to the shape of the cavity wall. The thickness of the wall of the casting depends on the time interval between the filling and the inverting of the mold, as well as on the chemical and physical properties of the alloy and the temperature and composition of the mold. Usually, lead and zinc alloy castings are

Table 5

produced by slush casting. The process is limited to the production of hollow castings and was used to produce lamp bases (Ref 6). Manually operated die casting machines were patented as early as 1849 (Sturgiss) and 1852 (Barr) in an effort to satisfy the insatiable demands of a growing reading public by way of rapidly cast lead type. These early inventions led to Ottmar Mergenthaler’s Linotype, an automatic casting machine in which molten lead is forced by piston stroke into a metal mold. The first die casting machine bearing the Linotype name was patented in 1905 by H.H. Doehler. Two years later came E.B. Wagner’s casting machine, a prototype of the now familiar hot chamber die casting machine. It was first used on a large scale during World War I for binocular and gas mask parts. Zinc alloys were used for die casting as early as 1907 but were not competitive until Price & Anderson developed the Zamak die casting alloy in 1929. In 1851, James Bogardus’s factory in Chicago, which was constructed with cast iron supports, opened the way for what many art historians considered to be America’s only original contribution to the arts of the world— the skyscraper. One of the oldest casting techniques, investment (lost-wax) casting, was also rediscovered in 1897 by B.F. Philbrook of Iowa, who used it to cast dental inlays. Industry paid little attention to this sophisticated process until the urgent military demands of World War I overtaxed the machine tool industry. Shortcuts were then needed to provide finished tools and

precision parts, avoiding time-consuming machining, welding, and assembly. Other efforts at this time included the first fully automated plant in the United States (one of the first in the world), which was a Rockford, Illinois, foundry that cast hand grenades for the U.S. Army in 1918. The first separation of aluminum occurred in 1825, but important engineering applications did not occur until the cost of aluminum steadily declined as a result of smelting process improvements and scale of operations through the end of the 19th century. Aside from additions for aluminum-bronze, zinc-aluminum alloys, and the deoxidation of steel, castings were the first important commercial market for aluminum. The process was simple, low cost, and addressed the limited range of product interests. At first, applications were limited to curiosities such as house numbers, hand mirrors, combs, brushes, tie clasps, cuff links, hat pins, and decorative lamp housings that emphasized the low density, silvery finish, and novelty of the new metal. Lightweight, corrosion-resistant and heat-conductive aluminum cookware could be produced using the same patterns used to cast iron and brass pots, pans, and kettles (Ref 7).

The 20th Century Casting technology during the 20th century advanced in many ways (Table 5), with “more casting progress since World War II than in the previous 3000 years” (Ref 8). For more than

Timeline of casting developments in the 20th century

1900s 1900—Brinnell hardness test machines introduced. Aluminum-bronze in regular production in the United States. Early 1900s—First patent for low-pressure permanent mold casting process issued to England’s E.H. Lake. 1901—American Steel Foundries (St. Louis) produces the first centrifugal cast rail wheels. 1903—The Wright Brothers’ first successful machinepowered aircraft contains a cast aluminum block and crankcase (together weighing 69 kg, or 152 lb), produced either at Miami Brass Foundry or the Buckeye Iron and Brass Works. 1905—H.H. Doehler patents the die casting machine. 1906—The first electric arc furnace is installed in the United States at Halcomb Steel Co., Syracuse, N.Y. First lowfrequency induction furnace is installed at Henry Diston & Sons, Tacony, Pa. 1907—Alfred Wilm discovers that the properties of cast aluminum alloys can be enhanced through heat treating and artificial aging. 1908—Stockham Homogenous Sand Mixer Co., Piqua and Newark, Ohio, releases the sand cutter. 1910s 1910—Matchplates are developed, fostering the viability of jolt-squeeze machines. 1911—Metallurgical microscope is obtainable. First electric arc furnace for metal casting is installed at Treadwell Engineering Co., Easton, Pa. 1912—The first muller with individually mounted revolving mullers of varying weights is marketed by Peter L. Simpson. Sand slinger is invented by E.O. Beardsley & W.F. Piper, Oregon Works.

1915—Experimentation begins with bentonite, a colloidal clay of unusually high green and dry strength. Ajax Metal Co., Philadelphia, installs first low-frequency induction furnace for nonferrous melting. 1916—Dr. Edwin Northrup, Princeton University, invents the coreless induction furnace. 1917—Alcoa completes a great deal of early development work in aluminum as World War I generates a great demand for high-integrity castings for aircraft engines. 1918—The first fully automated metal casting facility in Rockford, Ill., casts hand grenade hulks for the U.S. Army. 1920s 1921—Modification of the silicon structure in aluminum begins as Pacz discovers that adding metallic sodium to molten aluminum just prior to pouring greatly improves ductility. Copper-silicon alloys are prepared in Germany as a substitute for tin-bronzes. 1924—Henry Ford sets a production record of 1 million autos in 132 working days. Automotive manufacturing will grow to consume one-third of casting demand in the United States. 1925—X-ray radiography is established as a tool for checking casting quality. By 1940, all military aircraft castings require x-ray inspection prior to acceptance. American Brass, Waterbury, Conn., installs first mediumfrequency induction furnace in the United States. 1928—Alcoa develops the first aluminum vehicle wheel, a sand-cast 355 alloy designed for truck trailers. 1930s 1930—First high-frequency coreless electric induction furnace is installed at Lebanon Steel Foundry, Lebanon, Pa. Spectrography is pioneered by University of Michigan (continued)

Source: Ref 5; see also Selected References

professors for metal analysis. Davenport and Bain develop the austempering process for iron castings. 1937—Applied Research Laboratories’ founder Maurice Hasler produces the first grating spectrograph for the Geological Survey of California. Spectrometers begin finding their way into foundries by the late 1940s, replacing the previous practice of metallurgists estimating chemical compositions with a spectroscope and welder’s arc. The austempered microstructure in cast iron is recognized. 1940s 1940—Chvorinov develops the relationship between solidification time and casting geometry. Early 1940s—Statistical process control is first employed as a quality control tool in U.S. machine shops, principally to control dimensional tolerances. Inoculation of gray iron becomes common as high-quality cast irons replace scarce steel. 1941—U.S. Lt. Col. W.C. Bliss tells the American Foundrymen’s Association St. Louis Chapter that “the side which maintains the larger production of war goods is going to win the war.” The War Production Board reports later that each U.S. soldier requires 2200 kg (4900 lb) of steel compared to 40 kg (90 lb) in World War I. 1942—The use of synthetic sands increases as a replacement for many war materials. 1943—Keith Millis, a 28 year old metallurgist working at the International Nickel Company searching for a replacement to chrome due to interrupted supply, discovers that magnesium alloy in molten iron produces a spheroidal graphite structure. In 1949, he, Albert Gagnebin, and Norman Pilling would receive a U.S. patent on ductile iron production via magnesium treatment.

History and Trends of Metal Casting / 7 Table 5 (continued) 1940s 1944—The first heat-reactive, chemically cured binder is developed by Germany’s Johannes Croning to rapidly produce mortar and artillery shells for Axis troops during World War II. Two years after the war, his shell process is discovered among other inventions in the German patent office and made public. Croning is recognized with an American Foundrymen’s Society (AFS) Gold Medal in 1957 for his invention. 1946—Allied investigators uncover German foundry research on high-temperature alloys. Having “heard” of the Croning discovery (prior to the release of the report), Ford’s Ed Ensign and E.I. Valyi, Navy Bureau of Ships, attempted to replicate the process and produced shellmolded castings at the Midwest Foundry, Coldwater, Mich. Late 1940s—Thermal sand reclamation is applied to core sands and, to a limited degree, clay-bonded sands. 1948—The first nonlaboratory ductile iron casting is produced at Jamestown Malleable Iron Co., Jamestown, N.Y., as a 168 cm (66 in.) test bar is poured. Industry’s first ductile iron pipe is cast at Lynchburg Foundry, Lynchburg, Va. 1949—Keel blocks, diesel engine parts, a pressure cylinder, an 20 cm (8 in.) cube, and two cylinder liners become the first commercial castings of ductile iron at CooperBessemer, Grove City, Pa. Development of magnesiumferrosilicon makes ductile iron treatment far easier. Buffalo Pipe & Foundry Co., Tonawanda, N.Y., is the first U.S. operation to pour castings using Croning’s shell process. 1950s Early 1950s—Experimentation in high-pressure molding begins as metalcasters begin to increase the air pressure in air squeeze-molding machines to increase mold hardness (density). Fast-drying core oils are introduced. The pneumatic scrubber is developed to reclaim clay-bonded sands. Several wet reclamation systems also are in operation. 1951—Ford Motor Co., Dearborn, Mich., converts 100% of its crankshaft production to ductile iron. 1952—D-process is developed for making shell molds with fine sand and fast-dry oil by Harry Dietert. Sodiumsilicate/CO2 system is introduced. 1953—Hotbox system of making and curing cores in one operation is developed, eliminating the need for dielectric drying ovens. 1954—The CO2 process, a novel mold and coremaking process, is introduced from Germany. Working closely with General Motors, B&P develops a method for coating individual sand particles with resin binder. It also introduces a coreblower capable of producing cores with resin-coated shell sand—a modification to the Croning process. 1955—Ductile iron pipe is introduced to the marketplace by People’s Gas of Chicago, the first to install ductile iron gas mains. Mid-1950s—The squeeze-casting process originates in Russia. 1956—The first Betatron is installed in a U.S. foundry at ESCO Corp., Portland, Ore., for radiography of heavy steel castings. 1957—The vertically parted flaskless green sand molding production machine is invented by Vagn Aage Jeppesen, a 40 year old professor at the Technical University of Denmark. He was granted a patent in 1959, which was purchased by Dansk Industri Syndikat in 1961. 1958—Harold F. Shroyer obtains a patent for the full mold process, a process developed by artists in which simple patterns and gating systems are carved from expanded polystyrene and placed into a green sand mold. The process, known today as lost-foam casting (using loose, unbonded sand), is patented a short time later. Phenolic and furan acid-catalyzed no-bake binder systems are introduced. Ductile iron desulfurization via shaking ladles is developed in Sweden. 1959—General Electric uses the transient heat-transfer digital computer program and successfully applies the finite difference method to heavy steel casting production.

1960s 1960—Furan hotbox binders are developed for core production. Deep bed filters are used commercially for aluminum casting at Alcoa and British Aluminum in the United Kingdom. Compactibility and methylene blue clay tests are developed for green sand control. 1961—Alcohol-borne shell coating process is introduced (warm-coated). 1962—New CO2 sand testing method is introduced for sands bonded with sodium silicate and cured with CO2. Beardsley & Piper’s Al Hunter, Bob Lund, and Angello Bisinello develop the first automated green sand molding machine. In their design, the cope and drag are side-blown simultaneously and then hydraulically squeezed. The birth of automated matchplate molding reportedly improved metal casting productivity by levels as much as 60% in a short amount of time. Phenolic hotbox binders are introduced. 1963—Shell flake resin is introduced, eliminating the need for solvents. 1964—Dell & Christ’s paper on mold inoculation spurs the development of many of today’s forms of mold and late stream inoculation. The first vertically parted green sand machine (max. 240 molds/h) is delivered to United Danish Iron Foundries in Frederiksvaerk, Denmark. Early adopters report man-hour per ton improvements on the order of 50%. 1965—Oil urethane no-bake binder systems are used for cores and molds. General Electric’s Jim Henzel and Jack Keverian predict freezing patterns in large steel castings via computer. Cast metal-matrix composites are first poured at International Nickel Co., Sterling Forest, N.Y. 1968—The coldbox process is introduced by Larry Toriello and Janis Robins and introduced to the metal casting industry by Ashland Chemical Co. for high-production coremaking. Germany’s Daimler-Benz foundry in Mannheim is the first to run the process for automotive parts. John Deere Silvas Foundry, Moline, Ill., is the first to use the process for mass production in North America. 1969—The Chevrolet Vega is introduced by General Motors, featuring the first all-aluminum block with no cast iron cylinder liners. A total of 2.5 million blocks were produced during the vehicle life cycle. Late-1960s—Scanning electron microscope is invented in England. Thermal analysis begins to be used in iron foundries for the rapid determination of carbon equivalent and phosphorus contents, making it possible to study the transformation of an alloy during cooling. Manganese Bronze & Brass Co. and J. Stone & Co. join to promote nickel-aluminum bronze propellers. 1970s 1970—The sodium-silicate/ester catalyzed no-bake binder system is introduced for cores and molds. Safety-critical ductile iron steering knuckles are introduced on Chevrolet’s Cadillac. A new phenolic urethane no-bake process is introduced by Ashland Chemical, replacing oil sand dump box operations and significantly reducing energy requirements for core/mold production throughout the 1970s. Diran Apelian’s doctoral work at the Massachusetts Institute of Technology (MIT) leads to the development of foam filters for metal casting by Olin Metals. Commercial ceramic foam filters will be in use in metal casting facilities by 1974. 1971—The vacuum-forming or V-process molding method of using unbonded sand with a vacuum is developed in Japan. MIT doctoral candidate David Spencer performs experiments leading to semisolid molding, a process in which a partially frozen metal (fluidity similar to machine oil) can be poured into a die cavity. Following advancement in metal slurry consistencies with Professor Mert Flemings, Newton Diecasting Co., New Haven, Conn., produces the first semisolid castings. 1972—A 0.5 kg (1 lb) crankshaft for a refrigerator compressor produced at Wagner Castings (designed and engineered by Tecumseh) becomes the first productionvolume austempered ductile iron (ADI) component. Wagner also produces the first as-cast ductile iron ductile rods for passenger cars. The Canada Center for Mineral and Energy Technology (CANMET) uses real-time (continued)

Source: Ref 5; see also Selected References

1970s radiography to study the flow of steel in molds. Hitchiner Manufacturing, Milford, N.H., patents countergravity (vacuum) casting process. 1973—The first U.S. foundry argon oxygen decarburization unit is installed at ESCO Corp. 1974—Fiat introduces the in-mold process for ductile iron treatment. The phenolic urethane no-bake binder system is introduced for mold production. Mid-1970s—Alcoa and Union Carbide commence rotary degassing for wrought aluminum. Reading Foundry Products would apply this technology to aluminum foundries in the mid-1980s. Digital codes are developed to simulate solidification and fluid flow analysis. Ultrasonic verification of ductile iron nodularity is developed. Metal casting facilities examine new beneficial reuse routes for spent foundry sand, leading to applications such as cement and paving products, bricks, and flowable fill. 1976—Foote Mineral Co. and the British Cast Iron Research Association (U.K.) develop compacted graphite iron. Acidslag cupola practices plus external desulfurization with CaC2 begin to replace basic slag cupolas. 1977—General Motors installs ADI rear differential sets in passenger cars. The alumina phosphate no-bake binder system, an inorganic nonsilicate binder, is introduced for mold production. 1978—The furan/SO2 binder system is developed for core and mold making. Polyurethane no-bake binder system is introduced for aluminum applications. 1980s Early 1980s—Tundish ladle is embraced by industry as favored practice of nodularizing ductile iron. 3-D relational parameters are developed for Computer-aided design solid models. 1981—High-production lost-foam casting begins at General Motors’ Massena, N.Y., plant for aluminum cylinder heads. 1982—Warmbox binder system is introduced. 1983—Air impulse molding process is developed. Free radical cure/SO2 binder system is introduced. 1984—Charles Hull applies for a patent on stereolithography process. Other rapid prototyping techniques emerge shortly after. Phenolic ester no-bake binder is introduced. Thermal analysis makes breakthrough in molten aluminum processing for determination of grain refinement and silicon modification. 1985—Phenolic ester coldbox binder is developed. New automaker Saturn makes a strategic decision to select lostfoam process for its aluminum cylinder blocks and heads and ductile iron crankshafts and differential cases. Mid-1980s—Computer solidification software is commercialized. Amine recycling is introduced to enhance the environmental benefits of the amine-cured coldbox process. Ube Machinery introduces first squeeze-casting equipment. Aikoh’s (Japan) flux injection technology is initiated into U.S. aluminum foundry market. Late 1980s—3-D visualization techniques are developed. CaO/CaF2 desulfurization of cupola-melted ductile base iron begins to replace CaC2 method. Lanxide, Newark, Del., develops pressureless metal infiltration process for particulate-reinforced metal bodies. Magnesium wire injection method for ductile iron treatment is first tested. Casting solidification modeling software gains acceptance, allowing metal casting facilities to optimize quality, production, and cost prior to actually pouring a casting. 1988—Rapid prototyping and Computer-aided design/ computer-aided manufacturing technologies combine in a breakthrough to shorten tooling development time. Ford adapts Cosworth process precision sand casting process for high production. Metaullics Systems combines flux injection/rotary degassing technologies for aluminum processing. 1989—IMI (Yorkshire, U.K.) begins experimenting with bismuth as a lead substitute in copper alloys. 1990s 1990—Equipment for semisolid casting is introduced by Alumax Engineered Materials and Buhler, Inc. Foseco patents a direct-pour system that permits casting production without conventional gating/risering. Major automotive application comes in 1995 with CMI International’s upper intake

8 / Applications and Specification Guidelines Table 5

(continued)

manifold. Precision sand casting and casting quality for engine blocks are improved in mass production by major automotive companies with the Cosworth and Zeus processes for aluminum and the Loramendi Key Core process for precision sand-cast iron applications. 1991—“Dry ice” CO2 process is developed for cleaning coreboxes and foundry tooling. A noncontact gage for accurate dimensional analysis of lost-foam patterns and sand cores is developed through the AFS Lost-Foam Consortium. Eight years later, the consortium develops an instrument to measure the gas permeability of lost-foam pattern coatings (which controls flow of metal and has a dominant effect on casting quality). 1993—First foundry application of a plasma ladle refiner (melting and refining in one vessel) occurs at Maynard Steel Casting Co., Milwaukee. 1994—Use of low-expansion synthetic mullite sand for lost foam is patented by Brunswick Corp., Lake Forest, Ill., to enable precision casting of large components. Delphi Chassis Systems and Casting Technology Co. begin a program on squeeze-cast aluminum front knuckle, the

first high-volume production (1.5 million cars) two-cavity squeeze-cast aluminum conversion of its kind. Mid-1990s—Microstructure simulation is developed, contributing to better understanding of metallurgy effects and the prediction and control of mechanical properties in castings. Semisolid casting makes commercial inroads and market penetration. 1996—Cast metal-matrix composites (brake rotors) are used for the first time in production model automobiles, the Lotus Elise. Environmentally friendly (fluorine-free) fluxes are developed at the Worcester Polytechnic Institute and are commercialized. General Motors Corp. introduces GMBond, a water-soluble biopolymer-based core sand binder that is nontoxic and recyclable. 1997—AFS Consortium research at CANMET, Ottawa, Canada, results in the development and commercialization of lead-free copper alloys using bismuth and selenium. Following tests at Germany’s Aachen University, American Cast Iron Pipe Co., Birmingham, Ala., builds a continuously operated electric arc furnace for cast iron production.

Late 1990s—Stress and distortion simulation introduces the benefits of controlling casting distortion, reducing residual stresses, eliminating hot tears and cracking, minimizing mold distortion, and increasing mold life. 2000s 2001—The National Aeronautics and Space Administration (NASA) and the Dept. of Energy/Office of Industrial Technologies released a physics-based software tool to accurately predict the filling of expanded polystyrene patterns and sand cores as numerous variables are changed. Mercury Marine installs North America’s first pressurized lost-foam casting line at its new facility in Fond du Lac, Wis. 2002—Math-based modeling is used to develop a new sand core blowing simulation capability for process modeling software. 2003—The AFS Magnesium Division completes a research project proving that magnesium could be cast via the lost-foam casting process. Similar trials confirmed the castability of magnesium via the V-process.

phenolics led the way to urea and the dielectric process and then to furans and urea-free resins. The continued development of binders for the production of chemically bonded cores and molds is being directed toward increasing productivity as well as achieving the dimensional repeatability necessary to meet the new challenges of net shape and near-net shape casting requirements (Table 6). Many patterns were made of epoxy resins and polyurethane and other expendables such as polystyrene. Current understanding of sand casting now addresses the fundamentals of clay mineralogy, sand preparation, sand compaction, and the physical properties of molding and core sands. A better understanding of molding sand technology has resulted in sands of a higher degree of uniformity being prepared for the repetitive green sand (clay-bonded) molding sand. This high degree of achievement could only be possible with the great advances in sand testing produced by the foremost researchers and developers of sand testing instrumentation. Permanent-mold methods, which preceded the loam mold and the sand mold by centuries, also progressed for die casting (high pressure, low pressure, and gravity), centrifugal casting (vertical and horizontal), and hybrid processes such as squeeze casting and semisolid metal casting. Other contributions to the continued advancement of metal casting include:

 Cast metal-matrix composites  Argon oxygen decarburization for steel refining

Source: Ref 5; see also Selected References

Table 6 Development of core and mold processes Process

Core oil Shell: liquid and flake Silicate/carbon dioxide Airset oils Phenolic acid-catalyzed no-bake Furan acid-catalyzed no-bake Furan hot box Phenolic hot box Oil urethane no-bake Phenolic/urethane/amine cold box Silicate ester-catalyzed no-bake Phenolic urethane no-bake Alumina phosphate no-bake Furan sulfur dioxide Polyol urethane no-bake Warm box Free radical cure sulfur dioxide Phenolic ester no-bake Phenolic ester cold box

Approximate time of introduction

1950 1950 1952 1953 1958 1958 1960 1962 1965 1968 1970 1974 1977 1978 1978 1982 1983 1984 1985

Source: Ref 9

400 years, foundry processes and materials often relied on the methods developed by Vannoccio Biringuccio, the 16th century “father of the foundry industry,” who recommended using the dregs of beer vats and human urine as binders for molding sand, both of which were in use well into the 20th century. Until the 1920s, sand testing consisted of squeezing a handful of sand to judge its ability to compact and stick together. Early in that period, a sand research committee of the American Foundrymen’s Society began to develop sand test methods. By 1924, standards were established that covered the various properties of molding sands. Since World War II, experimentation also accelerated in organic and chemical sand binders for the thermosetting of molds and cores. Beginning with the Croning process (shell process),

 Foundry automation  A scientific approach to the gating and riser   

ing of castings using computer simulation of solidification Evaporative (lost)-foam casting and semisolid casting Computer-aided design and manufacture of castings Integrated foundry systems The melting of metals using the plasma arc cupola

Metal Casting Methods Metal casting is unique among metal forming processes, with a variety of casting processes (Table 7) (with Fig. 1, Ref 10). Very broadly, casting processes fall into two groups (Fig. 2): expendable-mold and permanent-mold processes. Each manufacturing process has the capability of producing a part with a certain range of tolerance, surface finish, and complexity. Manufacturing costs depend on the required dimensional tolerance, surface detail, and the quantity of parts required. For each process, there is a minimum batch size below which it is not economical to go because of costs of tooling, fixtures, and equipment. Also related to part cost is the production rate or the cycle time, the time required to produce one part. The most commonly used casting processes are evaluated with respect to these characteristics in Table 8 (Ref 11). Virtually any metal that can be melted can and is being cast. Choice of process and mold material (Table 9) is greatly influenced by the melting point of the alloy. A vital issue is avoidance of voids (due to insufficient feeding of melt) and porosity (due to dendritic solidification and the presence of gases). Success depends, to a large extent, on the skill of the mold designer, but there are very strong influences of the metal (alloy) and mold material. Even with the best design, solidification on the mold walls closes passage of melt at a certain distance, and this leads to a minimum allowable section thickness. There are large differences in the fluidity of alloys, and therefore, minimum allowable section thickness depends on the alloy being cast. Lower minimum thickness can be allowed with zinc, aluminum, and cast

History and Trends of Metal Casting / 9 Table 7

General characteristics of casting processes

See Fig. 1 for shape descriptions Casting process Characteristic

Part Material (casting) Porosity and voids(a) Shape(b) See Fig. 1 for shapes

Green sand

Resin-bonded sand

Plaster

Lost foam

All C–E All

All D–E All

Zn to Cu D–E All

Al to cast iron C–E All

All E All

Minimum section, mm (in.) Minimum core diameter, mm (in.) Surface detail(a)

0.01–300,000 (0.02–660,000) 3–6 (0.12–2.4) 4–6 (0.16–0.24) C

0.01–100 (0.02–220) 2–4 (0.08–0.16) 3–6 (0.12–0.24) B

0.01–1000 (0.02–2200) 1 (0.04) 10 (0.4) A

0.01–100 (0.02–220) 2–4 (0.08–0.16) 4–6 (0.16–0.24) C

Cost Equipment(a) Die (or pattern)(a) Labor(a) Finishing(a)

C–E C–E A–C A–C

C B–C C B–D

C–E C–E A–B C–D

Production Operator skill(a) Lead time Rates (piece/h
mold) Minimum quantity

A–C Days 1–20 1–100

C Weeks 5–50 100

A–B Days 1–10 10

Size, kg (lb)

Permanent mold

Investment

Die

Zn to Cu A–C Not T3, 5, F5

0.01–100 (0.02–220) 1 (0.04) 0.5–1 (0.02–0.04) A

Zn to cast iron B–C Not T3, 5, F5 with solid core 0.1–100 (0.2–220) 2–4 (0.08–0.16) 4–6 (0.16–0.24) B–C

B–C B–C C C–D

C–E B–C A–B C–D

B B C B–D

A A E C–E

C Weeks–months 1–20 500

A–B Hours–weeks 1–1000 10–1000

C Weeks 5–50 1000

C–D Weeks–months 20–200 100,000

250 >63 and 100 >10 and 5000 >100 and 0.5 >0.02 and 5 19 1.8 m, or 72 in.) 50 50 200 150

500

500

Materials

Hardwood patterns and core boxes Hardwood patterns and core boxes, wearing surfaces faced with metal Aluminum patterns and core boxes; plastic match plate patterns; urethane patterns or urethane-lined magnesiumframed core boxes Cast iron patterns and core boxes Hardwood patterns and core boxes Hardwood patterns and core boxes, wearing surfaces faced with metal Aluminum patterns and core boxes; urethane inserts in wear areas Softwood patterns and core boxes Softwood patterns with exposed projections metal faced; softwood core boxes, metal faced Hardwood patterns reinforced with metal; hardwood, metal-faced core boxes

Source: Ref 7

Fig. 8

Generalized casting costs versus production quantity for four pattern materials

dimensionally adjusted based on the results of prototype castings. These adjustments may involve refitting and remachining.

Trends in Pattern Design and Manufacture The continued development of computeraided design and manufacturing (CAD/CAM) technology has had a continuing effect on both pattern design and patternmaking methods (Ref 20–23). The influence of CAD/CAM technology and CNC technology will continue to grow as they become more affordable for the small

foundry and the pattern shop. Maximum benefit can be achieved with CAD/CAM technologies if pattern design, inspection, and manufacture are fully integrated. This integrated patternmaking approach begins with a geometric data base of the part to be cast. The data base may already exist or may have to be constructed from drawings. The pattern designer adds a parting line and draft and machining allowances at the CAD workstation. The part dimensions can be automatically scaled up to the designed pattern dimensions that included the shrinkage allowance by inputting the appropriate scale factor. At the current stage of development of many CAD software systems, necessary pattern features such as the parting line and draft must be added interactively by the pattern designer. To complete the pattern design, appropriate gate and riser sizes will be determined and added, using one of a number of stand-alone software packages. Solidification modeling software is often used to model the solidification of the final resultant casting from geometry data alone before even prototype patterns are made (see the articles in the Section “Modeling and Analysis of Casting Processes” in this Volume). Pattern dimensions, gating, and riser design can all be evaluated and readily modified by adjusting the pattern geometry data base. The use of solidification modeling programs can dramatically reduce or eliminate pattern prototyping stages. Casting design changes and/or modifications are common even after the pattern is built, but these can also be easily performed by adjusting the pattern data base. Direct CNC manufacture of patterns offers many advantages. Close-tolerance patterns can be readily produced using the pattern geometry data base. The CNC part programmer can

retrieve the pattern geometry model and use that model to construct the appropriate cutter paths for pattern machining. Graphical simulation of the machining operation on the CAD terminal can be used to verify the correctness of the CNC part program before actual pattern machining takes place. Any future modifications to the pattern geometry can be readily incorporated. Changes are made to the pattern data base, and the modified CNC output is used to modify the appropriate pattern dimensions. Multiple-impression tooling can be made with the assurance that each impression is dimensionally accurate. An example of the benefits that can be obtained using CAD/CAM patternmaking techniques compared to conventional patternmaking techniques is shown in Table 6. A 17% reduction in the time necessary to produce a final pattern is shown. Complete CNC machining of patterns will continue to expand as the benefits of CAD/ CAM technology continue to be exploited. The close dimensional tolerances and complex geometries required on many patterns are more suited to CNC pattern machining than to manual construction by patternmakers and machinists. Verification of pattern and casting dimensions is also a tedious process when traditional manual layout techniques are used. An automated coordinate-measuring machine (CMM) is an extremely accurate inspection tool that can dramatically improve the speed and reliability of pattern inspection (Ref 21). Such machines provide absolute measurement capabilities in three dimensions simultaneously; this eliminates the need to make tedious comparisons between the pattern and gage blocks or other length standards. With portable robotic arm CMMs, even large patterns or complex patterns with undercuts can be easily inspected. Data storage and printout capabilities allow easy measurement of many important pattern features. For example, cope and drag pattern plate alignment can be readily determined from comparisons of the locating pin and bushing positions for each pattern plate. Pattern dimensions or pattern inserts can also be inspected at specific intervals during pattern use with little downtime for monitoring pattern wear.

REFERENCES 1. Patternmaker’s Manual, American Foundry Society, 2006 2. E. Hamilton, AFS Patternmaker’s Guide, American Foundry Society, 1976 3. W.A. Blower, Pattern Design for High Pressure Molding, Trans. AFS, Vol 78, 1980, p 313–316 4. Patterns for High Pressure Molding, Foundry, Vol 98 (No. 10), Oct 1971, p 83–84 5. Metalcasting and Molding Processes, American Foundry Society, 1987

496 / Principles and Practices of Shape Casting Table 6 Comparison of traditional and CAD/CAM patternmaking times for an aluminum investment die Technique

Time

Traditional methods (accuracy of pattern: þ0.01 in.)  1) Design 50 h 2) Wooden model 50 h 3) Duplicating 40 h 4) Other machining 85 h 5) Polishing and finishing 40 h Total

265 h

CAD/CAM methods (accuracy of pattern: þ0.002 in.)  1) CAD/CAM design 40 h 2) CNC cutter path development 40 h 3) CNC machining 25 h 4) Other machining 85 h 5) Polishing and finishing 30 h Total

220 h

9. 10.

11. 12.

13. 14.

6. D.R. Dreger, Smooth No-Draft Castings, Mach. Des., May 25, 1978, p 63–65 7. M. Blair and T.L. Stevens, Ed., Steel Castings Handbook, 6th ed., and ASM International, 1995 8. J.L. Gaindhar, C.K. Jain, and K. Subbarathnamatah, Prediction of Pattern Dimensions

15.

16.

for V-Process Precision Castings Through Response Surface Methodology, Trans. AFS, Vol 94, 1986, p 343–349 R. Brown, Plastic Patterns for High Pressure Molding, Mod. Cast., Vol 73, Nov 1983, p 41–43 J.W. Francis, Practical Patternmaking Techniques for the Foundry Industry, Br. Foundryman, Vol 73 (No. 9), 1980, p 258–264 J.W. Francis, Practical Patternmaking Techniques for the Foundry Industry, FWP J., Vol 24 (No. 6), 1984, p 29–44 J.D. Pollard, Materials for Today’s Patternshop—A Personal Choice, Foundry Trade J., Vol 158 (No. 3306), May 23, 1985, p 415–419 M.J. Sneden, Who’s Afraid of Cast-to-Size Tooling, Trans. AFS, Vol 85, 1977, p 9–14 R.L. Allen, Cold Box Design Specializing in Urethane Lined Tooling, Trans. AFS, Vol 85, 1977, p 323–326 J. Sheeham and D. Richardson, The HMP Process—A New Method for Producing Plastic Patterns, Trans. AFS, Vol 92, 1984, p 203–208 G. Anderson and T.J. Crowley, Precision Foundry Tooling Utilizing CAD/CAM and

17. 18. 19. 20.

21. 22. 23.

the HMP Process, Trans. AFS, Vol 93, 1985, p 895–900 M.K. Young, Sprayed Metal Foundry Patterns for Short Run and Production Equipment, Trans. AFS, Vol 88, 1980, p 217–223 J.R. Henry, Metallic Coatings for Patterns and Coreboxes, Mod. Cast., April 1983, p 22–24 R.D. Maier and J.F. Wallace, Pattern Material Wear or Erosion Studies, Trans. AFS, Vol 84, 1977, p 161–166 D.R. Westlund, G.R. Anderson, and A.G. Anderson, Applying CAD/CAM to Foundry Tooling, Mod. Cast., Jan 1984, p 20–24 J.D. Taylor, Coordinate Measuring Machine Application in the Pattern Shop, Trans. AFS, Vol 88, 1980, p 195–198 J.R. Woods, NC-CNC and CAM in the Pattern Shop, Trans. AFS, Vol 91, 1983, p 743–746 D.B. Welbourn, CAD/CAM Plays a Major Role in Foundry Economics, Mod. Cast., Vol 77 (No. 9), Sept 1987, p 40–42

ASM Handbook, Volume 15: Casting ASM Handbook Committee, p 497-505 DOI: 10.1361/asmhba0005277

Copyright © 2008 ASM International® All rights reserved. www.asminternational.org

Casting Practice—Guidelines for Effective Production of Reliable Castings Jonh Campbell, The University of Birmingham, England

AS OUR UNDERSTANDING of the casting process is improved and refined, guidelines for effective production of reliable castings continue to evolve. In recent years, there has been a significant increase and improvement in our understanding of the casting process. This has led to new insights and criteria, which the author has steadily added to the list of requirements as they have become known. From an initial list of four rules, now ten rules have been identified that incorporate the latest technology for the production of reliable castings. These are just the start. It is already known that additional rules exist, but these remain to be further researched and clarified. The ten rules of casting, as originally described in Ref 1 and further developed in Ref 2, are proposed as being necessary but not sufficient for the manufacture of reliable castings. It is proposed that they are used in addition to existing conventional technical specifications, such as alloy type, strength, and traceability in accordance with ISO 9000 and so on, and other well-known and wellunderstood conventional foundry controls, such as casting temperature and other parameters. There are also basic reasons for believing that the ten rules have general validity, even though they have not been tested on all cast materials. Some rules have obvious implications and are applicable to all types of metals and alloys, including those based on aluminum, zinc, magnesium, cast irons, steels, air- and vacuum-cast nickel and cobalt, and titanium. Nevertheless, although all materials will probably benefit from the application of the rules; some may benefit more whereas others will be less affected. The rules as outlined here can be viewed as a draft process specification that a buyer of castings may use to purchase the best possible casting quality. Conversely, adherence to these rules is intended to assist the casting industry by helping to ensure the quality and reliability of the castings in a cost-effective way (as opposed to more expensive quality control on the finished product). The rules can help speed up the process of producing the casting right the first time and should contribute in a major way to the reduction of scrap when the casting goes into production. The ten

rules may also help raise the reliability of castings to be equal to that of forgings that can be offered with confidence.

Rule 1: Good-Quality Melt Immediately prior to casting, the melt shall be prepared and treated, if necessary, using the best current practice. Naturally, it is of no use to incorporate the best designs of filling and feeding systems if the original melt is of such poor quality, or perhaps already damaged, that a good casting cannot be made from it. It is a requirement that either the process for the production and treatment of the melt has been shown to produce good-quality liquid, or the melt should be demonstrated to be of good quality. A goodquality liquid is one that is substantially free from nonmetallic inclusions. It should be noted that such melts are not to be assumed and, without proper treatment, are probably rare. (Additional requirements, not part of this specification, may also be placed on the melt, for example, low gas content or low values of particular solute impurities.) An interesting possibility for future specifications for aluminum alloy castings (where, fortunately, residual gas in supersaturated solution does not appear to be harmful) is that a double requirement may be made for the content of dissolved gas in the melt to be high but the percentage of gas porosity in the casting to be low. The meeting of this double requirement will ensure the customer that double-oxide films are not present. This is because these damaging but undetectable defects, if present, will effectively be labeled and made plainly visible on x-ray radiographs and polished sections by the precipitation of dissolved gas.

Maximum meniscus velocity is in the range of 0.4 to 0.6 m/s (1.3 to 2.0 ft/s) for most liquid metals. This maximum velocity may be raised in constrained running systems or thin-section castings. This requirement also implies that the liquid metal must not be allowed to fall more than the critical height (which corresponds to the height of a sessile drop of the liquid metal). Maximum Velocity Requirement. Recent research has demonstrated that if the liquid velocity exceeds a critical velocity, there is a danger that the surface of the liquid metal may be folded over by surface turbulence (Fig. 1). Therefore, there is a chance that the surface oxide film may be folded into the bulk of the liquid if its speed of advance of the liquid front exceeds this critical velocity. The foldedin films constitute initiation sites for gas precipitation, shrinkage cavities, and hot tears, and after being frozen into the casting become effective cracks, lowering the strength and fatigue resistance. The folded films may also create leak paths, causing leakage failures. Thus, the entrainment of the liquid surface by surface turbulence leads to a catalog of problems that beset the foundry engineer. The folding-in of the oxide is a random process, leading to scatter and unreliability in the properties and performance of the casting on a casting-to-casting, day-to-day, and monthto-month basis during a production run.

Rule 2: Liquid Front Damage This is the requirement that the liquid metal front (the meniscus) should not go too fast.

Fig. 1

Effect of surface turbulence in entraining double (folded-over) oxide films, which form cracks and pores in the liquid

498 / Principles and Practices of Shape Casting The critical velocity is close to 0.4 m/s (1.3 ft/s) for dense alloys such as irons, steels, and bronzes; it is 0.5 m/s (1.6 ft/s) for liquid aluminum alloys, and 0.6 m/s (2.0 ft/s) for liquid magnesium alloys. The maximum velocity condition effectively forbids top-gating of castings. This is because liquid aluminum reaches its critical velocity of approximately 0.5 m/s (1.6 ft/s) after falling only 12.5 mm (0.5 in.) under gravity. The critical velocity of liquid iron or steel is exceeded after a fall of only approximately 8 mm (0.3 in.). Such short fall distances are always exceeded in practice in top-gated castings, leading to much incorporation of the surface oxide films and consequent leakage and crack defects. Castings that are made in which velocities everywhere in the mold never exceed the critical velocity are consistently strong, with high fatigue resistance, and are leaktight (if properly fed so as to be free from shrinkage porosity). Experiments on the casting of aluminum have demonstrated that the strength of castings may be reduced by as much as 90% if the critical velocity is exceeded. The corresponding defects in the castings are not always detected by conventional nondestructive testing such as x-ray radiography or dye penetrant, since, despite their large area, the folded oxide films are too thin to detect. The speed requirement automatically excludes conventional pressure die castings, since the filling speeds are over an order of magnitude in excess of the critical velocity. Some recent special developments of highpressure technology are capable of meeting this requirement, however. These include the vertical injection squeeze casting machine by UBE Machinery Corp. and the shot control technique by Buhler, both of which can, in principle, be operated to fill the cavity through large gates at low speeds and without the entrainment of air into the liquid metal. It is necessary to saw, rather than break, such castings from their filling systems. Other uphill filling techniques, such as the Cosworth process and low-pressure systems, are capable of meeting rule 2. Even so, it is regrettable that many low-pressure die casting machines are so poorly controlled on flow rate that the speed of entry into the die greatly exceeds the critical velocity, thus negating one of the most important potential benefits of the low-pressure system. In addition, the critical velocity is practically always exceeded during the filling of the low-pressure furnace itself because of the severe fall of the metal as it is transferred into the pressure vessel, so that the metal is damaged even prior to casting. Low-pressure machines with interchangeable crucibles can avoid this problem. Metals that behave in the same way, suffering damage from the entrainment of their surface films, are suggested to be the zincaluminum alloys and ductile irons. Carbon steels and some stainless steels are thought to be similar, although in some of these systems,

the entrained oxides agglomerate because they are partially molten and thus float to form surface imperfections in the form of slag macroinclusions on the surface. Duplex stainless steels exhibit a tough surface oxide film on the melt that gives special problems if surface entrainment is permitted. For a few materials, particularly alloys based on the Cu10Al-type bronzes, the critical velocities were thought to be much lower, in the region of 0.075 m/s (0.25 ft/s). However, from recent work at the University of Birmingham, this seems to be a mistake, probably resulting from the confusion caused by bubbles entrained in the early part of the filling system. With well-designed filling systems, the aluminum bronzes fulfill the theoretically predicted 0.4 m/s (1.3 ft/s) value for a critical ingate velocity. No-Fall Requirement. It is quickly shown that if liquid aluminum is allowed to fall more than 12.5 mm (0.5 in.), it exceeds the critical 0.5 m/s (1.6 ft/s). Similar critical velocities and critical fall heights can be defined for other liquid metals. The critical fall heights for all liquid metals are in the range of 3 to 15 mm (0.12 to 0.6 in.). Thus, any form of pouring, such as in transfer between ladles (Fig. 2) or top-gating of castings, almost without exception will lead to a violation of the critical velocity requirement. Many forms of gating that enter the mold cavity at the mold joint will also violate this requirement if any significant part of the cavity is below the joint. In fact, for conventional sand and gravity die (permanent mold) casting, the requirement effectively dictates bottom-gating into the mold cavity and excludes any other form of gating at any other height. Also excluded are any filling methods that cause waterfall effects in the mold cavity. Avoiding waterfalls dictates the siting of a separate ingate at every isolated low point on the casting. The initial fall down the sprue in gravityfilled systems does introduce some oxide damage into the metal, especially if using an oversized sprue where the metal is not constrained (Fig. 3). However, for some alloys, much of the oxide introduced in this way is attached to the walls of the sprue and does not appear to find its way through to the mold cavity. This surprising effect is clearly seen in many topgated castings, where most of the oxide damage (and particularly any random leakage problem) is confined to the area of the casting under the point of pouring, where the metal falls. Although extensive damage does not seem to extend into those regions of the casting where the speed of the metal front decreases and where the front travels uphill, there appears to be a carryover of some concentration of defects. Thus, the provision of a filter immediately after the completion of the fall down the sprue is valuable. However, some damage is still expected to pass through the filter. The requirement for no fall of the liquid within the running system (after the metal

Fig. 2

Effect of increasing height on a falling stream of liquid. (a) Oxide film remains intact. (b) Oxide film detaches and accumulates to form a dross ring. (c) Oxide film and air are entrained in the bulk melt.

leaves the base of the sprue) and mold cavity implies that metal must be designed to go only uphill after it leaves the base of the sprue. Thus, the runner must be in the drag and the gates must be in the cope for a horizontal singlejointed mold. (If the runner is in the cope, there is the danger that the gates will fill prematurely, before the runner itself is filled; thus, air bubbles are likely to enter the gates.) The no-fall requirement may also exclude some of those filling methods in which the metal slides down a face inside the mold cavity, such as Durville casting-type processes. During the slide down a slope, the critical velocity is easily exceeded if the total loss of height exceeds the critical fall height. This effect is discussed in more detail in the following section. It is noteworthy that these precautions to avoid the entrainment of oxide films also apply

Casting Practice—Guidelines for Effective Production of Reliable Castings / 499

Fig. 3

Oversized sprue that has suffered severe erosion damage because of nonfilling during the pour. A correctly sized sprue shows a bright surface, free from damage.

application of pressure to force it into such narrow gaps. This well-known effect is known as capillary repulsion. The liquid surface is now so constrained that it is not easy to fold over the surface, that is, there is no room for splashing or droplet formation. Thus, the critical velocity is higher, and metal speeds can be raised somewhat without danger of exceeding this higher limit. In fact, tight curvature of the meniscus becomes so important in very thinwalled castings, with walls less than 2 mm (0.08 mm) thickness, that filling can sometimes be without regard to gravity (i.e., can be uphill or downhill), since the effect of gravity is swamped by the effect of surface tension. The dominance of surface tension also makes the uphill filling of such thin sections problematic, because the effective surface tension exceeds the effect of gravity, so that instabilities can occur, whereby the moving parts of the meniscus continue to move because of the reduced thickness of the oxide skin at the advancing location. Other parts of the meniscus that drag back are further suppressed in their advance by the thickening oxide, so that a runaway instability condition occurs. This dendritic advance of the liquid front is no longer controlled by gravity in very thin castings, making the filling of extensive horizontal sections a major filling problem. The problem is relieved by the presence of regularly spaced ribs or other geometrical features that assist in organizing the distribution of liquid and thus avoid the starving of areas that the flow happens to avoid because of random meandering. Such chance avoidance, if prolonged, leads to the development of strong oxide films or even freezing of the liquid front. Thus, the final advance of the liquid to fill such regions is hindered or prevented altogether.

Rule 3: Liquid Front Stop Fig. 4

Unstable advance of a film-forming alloy, showing the formation of laps as the interface intermittently stops and restarts by bursting through and flooding over the surface film

to casting in (so-called) inert gas or even in a (so-called) vacuum. This is because the oxides of aluminum and magnesium (as in aluminum alloys, ductile irons, or high-temperature nickel-base alloys, for example) form so readily that they effectively “getter” the residual oxygen in any contaminated gas or conventional industrial vacuum and form strong films on the surface of the liquid. Rule 2 applies to “normal” castings with walls of thickness over 3 or 4 mm (0.12 or 0.16 in.). For very thin-walled castings, of section thickness less than 2 mm (0.08 in.), the effect of surface tension in controlling filling becomes predominant, the meniscus being confined between walls that are so much closer than the equilibrium curvature that the meniscus is effectively compressed and requires the

This is the requirement that the liquid metal front should not go too slowly and, more exactly, should not stop at any point on the front. The advancing liquid metal meniscus must be kept “alive” (i.e., moving) and therefore free from thick oxide film that may be incorporated into the casting. This is achieved by designing the liquid front to progress only uphill in a continuous uninterrupted upward advance (in the case of gravity-poured casting processes, from the base of the sprue onward). This rule implies:
Only bottom-gating is permissible. No stop-

ping due to arrest of pouring. No downhill flow (either falling or sliding)
No horizontal flow (of any significant extent) The liquid metal front must continue to advance at all points on its surface. If it stops, a thick surface film has the opportunity to form and thus resist the restarting of the flow at some later moment. The incorporation of this film in

the casting as metal flows over it from other regions forms a major defect, often spanning the casting from wall to wall (Fig. 4). This problem can occur in several ways, as detailed subsequently and discussed further in the article “Filling and Feeding System Concepts” in this Volume. If the liquid metal is allowed to either fall or spread horizontally, the unstable propagation of the front in the form of jets or streams bounded by an oxide flow tube leads to a situation where the oxide flow tube grows to a considerable thickness and is finally sealed into the casting as the liquid metal envelops it. A double-oxide film defect is created, with a highly asymmetric double layer, in which one oxide is thick and the other thin (Fig. 4). However, since dry side is arranged against dry side as usual, the defect is once again a crack. Such cracks are extensive and serious. They are different from the fragmented and chaotic double films introduced by surface turbulence. They have predictable geometric shapes, such as planes, cylinders, or riverine forms. In common with the irregular and unpredictable forms produced by surface turbulence, they are normally invisible to x-rays and to dye penetrant testing. They are only revealed by gross mechanical deformation that opens the crack. The deleterious flow tube structure that forms when filling downward or horizontally is usually eliminated when filling vertically. However, even in this favorable mode of filling, a related defect can still occur if the advance of the meniscus is stopped at any time. This effect is discussed as follows. Continuous Uphill Advance of the Meniscus. The uphill motion of the liquid front, if it can be arranged in a mold cavity, assists in keeping all parts of the front “alive;” that is, the meniscus does not become pinned in place with oxide and thus become “dead.” As explained earlier, the front becomes pinned if the rate of advance of the metal front is too slow or if it stops. If this happens, the surface oxide has the opportunity to thicken, becoming so strong that there is the danger that any further advance of the front will be prevented. The breaking through of the meniscus at a weak point will then flood over the fixed, thickened oxide, sealing it into the casting as a major oxide lap defect (Fig. 4). One of the reasons to prevent any metal fall in the mold is not only to avoid exceeding the critical velocity and thus impairing the metal that falls but also to avoid the period when the rise of the metal is interrupted, causing an oxide to form on the stationary front (while the metal level is fixed during the period of the overflow) and on the tube of flow surrounding the fall itself, which is also effectively stationary. The no-fall requirement may also exclude some of those filling methods in which the metal slides down a face inside the mold cavity, such as Durville casting-type processes. These processes can be used to good advantage if used

500 / Principles and Practices of Shape Casting correctly. In such cases, the liquid metal front progresses steadily at all points, expanding and breaking its surface oxide and rolling it out against the mold wall as the front progresses. In this way, the oxide never enters the bulk of the metal; it only continues to augment the oxide already present on the surface of the casting. The production of ingots by the controlled tilting of the mold can maintain the “live” nature of the front at all times as the flow expands into the ingot mold. When a Durville-type pour is carried out badly, the metal can flow down into the mold as a stream. Its boundaries (that is, its outer surface is stationary) are formed as a tube of oxide through which the metal flows while it is running down one of the faces of the mold. This oxide flow tube becomes sealed into the finished casting, creating a curious but serious tubular crack defect. The problem is analogous to the oxide tube defect that forms around the falling metal during any waterfall effect during the filling of the casting. These flow tubes are a common defect seen in a wide variety of castings that exhibit either extensive horizontal or downhill sections. The requirement that the meniscus travels only uphill is sacred. Similar defects that have the appearance of laps also occur on the sidewalls of flows as they meander aimlessly across large horizontal parts of the mold. The rattail defect is a familiar reminder of such wanderings of the metal stream. (In this case, the rattail line is usually formed by the failure of the mold. However, a corresponding line of oxide defect in the metal casting may also be present as an additional problem.) The avoidance of extensive horizontal mold sections is therefore essential for reproducible and defect-free castings. Any horizontal sections should be avoided by the designer or by the caster by tilting the mold. (This is easily provided by some casting techniques such as the Cosworth process, where the mold is held in a rotatable fixture during casting.) The flow across such inclined planes is therefore progressive, if slow, but the continuous advance of the front at all points assists in the goal of keeping the meniscus “alive.”

Rule 4: Bubble Damage No bubbles of air entrained by the filling system should pass through the liquid metal in the mold cavity. This may be achieved by:
Use of an offset step pouring basin design of

sufficient size to facilitate detrainment of entrained air and oxides; preferred use of a stopper to aid the filling of the basin; the use of sprue and runner designed to fill in one pass; avoidance of the use of wells or other volume-increasing features of filling systems; possible use of ceramic foam filter close to sprue/runner junction; possible use of bubble traps
No interruptions to pouring

The passage of a single bubble through an oxidizable melt results in the creation of a bubble trail as a double-oxide crack in the liquid. Thus, even though the bubble may be lost by bursting at the liquid surface, the trail remains as permanent damage in the casting. Poor filling system designs can result in the entrainment of much air into the liquid stream during its traverse through the filling basin, its fall down the sprue, its travel along the runner, and its entry into the mold cavity, if the melt velocity is still too high at that point, or if the melt is allowed to fall inside the mold cavity. The repeated passage of many bubbles through the liquid metal leads to an accumulation of a tangle of bubble trails and, if the density of trails is sufficiently great, residual fragments of entrapped bubbles wedged in among the trails. This mixture is collectively christened as bubble damage (Fig. 5). Bubble damage, as a mass of oxides and fragments of porosity, is probably the most common defect in castings but is almost universally unrecognized. The problem is commonly observed just inside and above the first ingate from the runner and is often mistaken for shrinkage porosity as a result of its irregular form. However, in an aluminum alloy casting, it can be recognized on a fracture surface; instead of shiny dendrite tips characteristic of shrinkage porosity, a series of dark, nonreflective oxidized surfaces interleaved like crumpled pages of sepia-colored paper will be found. (Some associated shrinkage porosity with its dendrites will almost always be present but should not cause the major nonreflective regions to be overlooked. It is easy to fall into this trap because under the microscope, in contrast with the eye-catching regular crystallographic order of the dendrites, the oxidized regions do not attract attention.) In some stainless steels, bubble damage is seen under the microscope as a mixture of bubbles and cracks (a remarkable combination that would normally be difficult to explain as a metallurgical phenomenon), because the high cooling strain leads to high stresses in such a strong material and thus opens up the double-oxide bubble trails. Gravity-Filled Running Systems. In gravityfilled running systems, the requirement to reduce bubbles in the liquid stream during the filling of the casting calls for an offset stepped basin or other advanced basin design. The conventional conical pouring basin cannot be permitted. The cone shape usually introduces approximately 50% air, introducing massive amounts of damage. The requirement also demands properly engineered and manufactured sprues. The sprue must have a calculated taper. Preferably, the taper should not be a straight taper but a hyperbolic shape to follow the shape of the falling stream, maintaining the pressurization of the walls of the sprue at all points during the fall. Parallel or reversed-taper sprues are not permitted without some feature to reduce the damage that they cause. Such damage-limiting

features include a filter at the entrance to the runner, or inverted well, a pork-pie-shaped obstruction designed to aid the backfilling of the sprue. It is mandatory that the taper of the sprue contains no perturbations to upset the smooth fall of the liquid metal. Thus, it must fit well with the pouring basin or other mold or die joints; no steps, ledges, or abrupt changes in direction are permissible. Also, no branching or joining of other ducts, runners, gates, or sprues is allowed. All such features (especially common in investment castings) have the potential for the introduction of air into the stream or the uncontrolled escape of liquid into other parts of the mold cavity. At one time, it was mandatory that each sprue have a sprue well at its base. The well was thought to facilitate the turn of the metal through the right-angle bend into the runner with minimum turbulence. Previous work on well development had been based on water models. However, more recent work at the University of Birmingham using liquid metals has demonstrated that, at best, the well is no better than no well at all and, at worst, causes considerable extra turbulence. Thus, the new filling system designs incorporate no well at the base of the sprue. It is mandatory that no interruption to the pour occurs that leads to uncovering the entrance to the sprue, so that air enters the running system. A provision must be made for the foundry to automatically reject any castings that have suffered such an interrupted pour. Pumped and Low-Pressure Filling Systems. Pumped systems such as the Cosworth process or low-pressure casting systems into sand molds or metal dies are highly favored as having the potential to avoid the entrainment

Fig. 5

Emergence of liquid metal and air bubbles through an ingate, showing the creation of bubble trails and residual bubbles

Casting Practice—Guidelines for Effective Production of Reliable Castings / 501 of bubbles if the processes are carried out under proper control. (Good control of a potentially good process should not be assumed; it must be demonstrated. In addition, bubbles can be released erratically from the interior wall of a tube launder system as well as from the underside of a badly designed distribution plate or from a poorly maintained and cleaned launder tube, if one is used.) Although low-pressure filling systems can, in principle, satisfy the requirement for the complete avoidance of bubbles in the metal, a leaking riser tube in a low-pressure casting machine can lead to a serious violation. The stream of bubbles from a leak in the riser tube will rise up the tube and directly enter the casting. Thus, regular checks of such leakage, and the rejection of castings subjected to such consequent bubble damage, will be required. The other major problem with conventional low-pressure delivery systems where the melt is contained within a pressurized vessel is that the quality of the melt is usually damaged by the filling of the pressure vessel. This is an uncontrolled fall from the foundry transfer ladle, then down a chute, and finally into the melt. This unsatisfactory transfer process introduces much oxide into the metal, only part of which is subsequently removed by a filter at the entrance to the mold cavity. As mentioned previously, this damaging filling procedure is avoided by those low-pressure machine designs that incorporate multiple exchangeable crucibles.

Rule 5: Core Blows

occurs because the clay contains water bonded into its crystallographic structure but is impermeable. When heated by the arrival of the melt, the water vapor from the clay is prevented from escaping into the core or mold, thus forcing the escape path of the steam through the melt. The use of core and mold repair pastes is to be avoided (unless they can be demonstrated to avoid the generation of blows in the melt). To demonstrate that a core, or assembly of cores, does not produce blows may require a procedure such as the removal of all or part of the cope or overlying cores and taking a video recording of the filling of the mold. If there are any such problems, the eruption of core gases will be clearly observable, resulting in the creation of a froth of surface dross (which would normally be entrapped inside the upper walls of the casting). A series of video recordings may be necessary, showing the steady development of solutions to a core-blowing problem and recording how individual remedies resulted in progressive elimination of the problem. The video recording should be retained by the foundry for inspection by the customer for the life of the component. Any change to the filling rate of the casting, core design, core repair procedure, or change of binder will necessitate a repeat of this exercise. For castings with a vertical joint, where a cope cannot be conveniently lifted clear to provide such a view, a special sand mold may be required to carry out the demonstration that the core assembly does not cause blows from the cores at any point. This must be constructed as part of the tooling to commission the casting and justified as an investment in quality assurance.

Fig. 6

Detachment of a bubble from the top of a core, leaving a bubble trail as a permanent legacy of its journey. This bubble may be early enough to escape at the free surface of the rising metal.


No bubbles from the outgassing of cores

should pass through the liquid metal in the mold cavity. Cores must be of sufficiently low gas content and/or adequately vented to prevent bubbles from core blows.
No use of clay-based core or mold repair paste (unless demonstrated to be free from the risk of creating bubbles) Serious oxide films, which can cause leaks and mechanical cracking, can be caused by the outgassing of sand cores through the melt, causing bubbles to rise through the metal and leaving an oxidized bubble trail in its wake (Fig. 6). A succession of such bubbles from a core is very damaging to the upper parts of castings (Fig. 7). The bubble from the core contains a variety of gases, including water vapor, that are highly oxidizing to metals such as aluminum and higher-melting-point metals. Bubble trails from core blows are usually particularly noteworthy for their characteristically thick and leathery double-oxide skin, which is probably why core blows result in such efficient leak defects through the upper sections of castings. Real-time x-ray radiography of solidifying castings has revealed considerable volumes of water vapor driven into the melt from claybased core repair and mold repair pastes. This

Rule 6: Shrinkage Damage
No feeding uphill because of unreliable

pressure gradient and complications introduced by convection
Demonstrate good feeding design following all seven feeding rules, by an approved computer solidification model and test castings
Control of the level of flash at mold and core joints, the mold coat thickness, and the temperatures of metal and mold If computer modeling is not carried out, the observance of the author’s seven feeding rules is strongly recommended. Even when computer simulation is available, the seven rules are good guidelines. These rules list the requirements for good feeding under gravity. For example, Fig. 8 shows the penalty of large amounts of porosity if the feeder is too small, and even some penalty if too large. In addition, all five mechanisms for feeding (as opposed to only liquid feeding) should also be used to advantage. For example, solid feeding (a kind of self-feeding by the solidified casting) is especially useful when attempting to achieve soundness in an isolated boss or heavy

Fig. 7

(a) Core blow—a trapped bubble containing core gases. (b) Bubble trail, ending in an exfoliated dross defect as the result of the passage of copious volumes of core gas prior to freezing.

section, where the provision of feed metal by conventional techniques is impossible. Feeding with Gravity. As opposed to filling uphill (which is quite correct), feeding should only be carried out downhill (using the assistance of gravity). Attempts to feed uphill, although possible in principle, can be unreliable in practice and may lead to randomly occurring defects that

502 / Principles and Practices of Shape Casting

Fig. 8

Effect of increasing feeder solidification time on the soundness of a plate casting in Al-12Si alloy

have all the appearance of extensive shrinkage porosity. These occur because of the difficulties caused by two main problems: adverse pressure gradient and adverse density gradient leading to convection. Adverse pressure gradient is dealt with subsequently; the second problem is dealt with in “Rule 7: Convection Damage” in this article. The atmosphere is capable of holding up several meters of the head of a metal. For liquid mercury, the height is approximately 760 mm (30 in.) (the height of the barometer column). There are similar equivalent heights for other liquid metals, allowing for the density difference; thus, the atmosphere will hold up approximately 4 m (13 ft) of liquid aluminum, for example. (While no pore exists, the tensile strength of the liquid will allow the metal, in principle, to greatly exceed this height, since, in the absence of defects, it can withstand tensile stresses of thousands of atmospheres. However, the random initiation of a single minute pore will instantly cause the liquid to “break,” causing such feeding to stop and go into reverse. The casting will then empty down to the level at which atmospheric pressure can support the liquid.) However, if there is a leak path to atmosphere, allowing atmospheric pressure to be applied in the liquid metal inside the mold cavity, the melt will then fall further, attempting to equalize levels in the mold and feeder. Thus, the residual liquid will drain from the casting if the feeder is sited below the casting. This is an efficient way to cast porous castings and sound feeders. Clearly, the initiation of a leak path to atmosphere (via a double-oxide film or via a liquid region in contact with the surface at a hot spot) is rather easy in many castings, making the

whole principle of uphill feeding so risky that it should not be attempted where porosity cannot be tolerated. The recent practice of active feeding, whereby small feeders placed low on the casting are pressurized to cause the casting to feed uphill, is almost certainly allowable in small castings. For larger castings, the technique is subject to the problems of convective flow and is expected to become unreliable. Computer Modeling of Feeding. Computer modeling has demonstrated its usefulness in predicting shrinkage porosity with accuracy. It should be specified as a prior requirement before work is started on making the tooling. This minor delay will have considerable benefits in shortening the overall development time of a new casting and will greatly increase the chance of being “right the first time.” It must be recognized, however, that the computer will not necessarily be capable of any design contribution. Thus, filling and feeding will be required as inputs. Additionally, many computer simulations do not simulate the important effects due to convection, nor do some include conduction. Some neglect, or only crudely allow for, the effect of mold heating by the flow of metal during filling. The belief in the results from such models must therefore be tempered with caution if not skepticism. A further problem surrounds the use of computer predictions of shrinkage porosity when comparing results with real castings. It is common for the castings to exhibit areas of oxide film that closely resemble areas of shrinkage on a radiograph. The misidentification of these regions as shrinkage porosity has led to huge confusion and unjust maligning of computer predictions. It is important when viewing radiographs not to assume that such areas are shrinkage porosity but to report these areas as “appearing to be shrinkage.” Most often, they are not shrinkage porosity. Random Perturbations to Feeding Patterns from Casting to Casting. Flash approximately 1 mm (0.04 in.) thick and only 10 mm (0.4 in.) wide has been demonstrated to have a powerful effect on the cooling of local thin sections up to 10 mm (0.4 in.) thick, speeding up local solidification rates by up to 10 times. Thus, flash must be controlled, or used deliberately, since it has the potential to cut off feeding to more distant sections. The erratic appearance of flash in a production run may introduce uncertainty in the reproducibility of feeding and the consequent reproducibility of the soundness of the casting. Moderate flash on thicker sections is usually less serious, because convection in the solidifying casting conveys the local cooling away, effectively spreading the cooling effect over other parts of the casting, so that an averaging effect over large areas of the casting is created. In general, however, it is desirable that these uncertainties be reduced by good control over mold and core dimensions.

The other known major variable affecting casting soundness in sand and investment castings is the ability of the mold to resist deformation. This effect is well established in the case of cast irons, where high mold hardness leading to good mold rigidity is needed for soundness. However, there is evidence that such a problem exists in castings of copper-base alloys and steels. A standard system, such as statistical process control (SPC) or other technique, must be in place to monitor and facilitate control of such changes. (Permanent molds, such as metal or graphite dies, are relatively free from such problems.) The solidification pattern of castings produced from permanent molds such as gravity dies (permanent molds) and low-pressure dies may be considerably affected by the thickness and type of die coat that is applied. A system to monitor and control such thickness on an SPC system must be in place. Alternatively, for each casting on a case-by-case basis, the thickness of the die coat must be demonstrated to be of no consequence. For some permanent molds, pressure die, and some types of squeeze casting, the feeding pattern is particularly sensitive to mold cooling. Changes to cooling channels in the die or to the cooling spray during die opening must be checked to ensure that corresponding deleterious changes have not been imposed on the casting.

Rule 7: Convection Damage Assess the freezing time in relation to the time for convection to cause damage. Thinand thick-section castings automatically avoid convection problems. For intermediate sections, either reduce the problem by avoiding convective loops in the geometry of the casting and rigging, or eliminate convection by rollover after filling. Convection enhances the difficulty of uphill feeding in medium-section castings, making them extremely resistant to solution by increasing the (uphill) feeding. In fact, increasing the amount of feeding, for example, by increasing the diameter of the feeder neck for uphill feeding, makes the problem worse by increasing the driving force for convection. Many of the current problems of low-pressure casting systems derive from this source. In contrast, feeders with hot metal at the top of the casting and feeding downward under gravity are completely stable and predictable and give reliable results. Convection Damage and Casting-Section Thickness. Damage to the micro- and macrostructure of the casting can occur if the solidification time of the casting is commensurate with the time taken for convection to become established, since extensive remelting can occur. This incubation time appears to be in the region of 2 min for many castings of approximately 10 to 100 kg (22 to 220 lb). In 3 min or more, convection can cause extensive remelting, the

Casting Practice—Guidelines for Effective Production of Reliable Castings / 503 development of flow channels, and the redistribution of heat in castings. Castings that freeze either quickly or slowly are free from convection problems, as indicated subsequently. Thin-section castings can be fed uphill simply because the thin section gives a viscous constraint, which reduces flow, and more rapid freezing, thus allowing convection less time to develop and wreak damage in the casting. Therefore, instability is both suppressed and given insufficient time, so that satisfactory castings can be made. Thick-section castings are also relatively free from convection problems, because the long time available before freezing allows the metal time to convect, reorganizing itself so that the hot metal floats gently into the feeders at the top of the casting, and the cold metal slips to the bottom. Once again, the system reaches a stable condition before any substantial freezing has occurred, and castings are predictable. The convection problem arises in the wide range of intermediate-section castings, such as automotive cylinder heads and wheels, and the larger investment-cast turbine blades in nickelbase alloys and so on. This is an important class of castings. Convection can explain many of the current problems with difficult and apparently intractable feeding problems in such common products. The convective flow takes approximately 1 or 2 min to gather pace and organize itself into rapidly flowing plumes. This occurs at the same time that the casting is attempting to solidify. Thus, the action of the convecting streams will remelt channels through the newly solidified material. These channels will contain a coarse microstructure because of their greatly delayed solidification and may contain shrinkage porosity if unconnected to feed metal. This situation is likely if the feeders solidify before the channels, as undoubtedly happens on occasion. Conventional gravity castings that require a significant amount of feed metal, such as cylinder heads and blocks, and are bottom gated but top fed will dictate large top feeders. This follows because of their inefficiency as a result of being furthest from the ingates and thus containing cold metal, in contrast to the ingate sections at the base of the casting that will be nicely preheated. This unfavorable temperature regime is unstable because of the inverted density gradient in the liquid and thus leads to convective flow and consequent poor predictability of the final temperature distribution and effectiveness of feeding. The upwardly convecting liquid within the flow channels usually has a freezing time close to that of the preheated section beneath that is providing the heat to drive the flow. In many low-pressure systems, artificially heated metal supply systems lead to a constant heat input, so that the convecting streams never solidify. Thus, long after the casting should have completely solidified, the channels empty or partially empty, pouring out liquid metal when the casting is lifted from the casting station.

Alternatively, a technology where the mold is inverted after casting effectively converts the preheated bottom ingate system into a topfeeding system (Fig. 9). Furthermore, the inversion of the system to take the hot metal to the top and the cold at the bottom confers stability on the thermal regime. Convection is eliminated. This is a powerful and reliable system used by such operations as Cosworth and an increasing number of others at the present time (2008). It is expected that techniques involving rollover immediately after casting will become the norm for many castings in the future. The rollover is only reliably carried out while the pressure of filling is maintained. This is because the hydrostatic head of metal above the cores is greatly reduced during this action, creating the danger of core blows. If a bubble does form, it may travel up or down, or both, during the inversion, significantly increasing damage and confusing any identification of its source. Durville-type casting processes (where the rollover is used during casting—actually to effect the filling process) also satisfy the topfeeding requirement. However, in practice, many geometries are accompanied by waterfall effects, if only by the action of the metal sliding in the form of a stable, narrow stream down one side of the mold. Thus, meniscus control is unfortunately often poor. Where the control of the meniscus can be improved to eliminate surface film problems, the Durville-type technique is valuable. An alternative kind of convection, that driven by density differences due to segregation, may lead to other problems, as outlined subsequently.

castings and the concentration of heavy elements, such as tungsten and molybdenum, at the base of large tool steel castings. Strong concentrations of segregated solutes and inclusions are found in channel segregates, which are a feature of large, slowly cooled castings. When extensive and/or intensive, such changes in composition of the casting may cause the alloy of the casting to be locally out of specification. If this is a serious deviation, the coincidence of local brittleness in a highly stressed region of the casting may threaten the serviceability of the product. The possibility of such regions must be assessed prior to casting, if possible, and demonstrated to be within acceptable limits in the cast product.

Rule 8: Segregation Predict segregation to be within the specification limits or agree upon out-of-specification compositional regions with the customer. Avoid channel segregation formation if possible. At regions in which the local cooling rate of the casting changes, such as at a change of section or at a chill or feeder, it is expected that a change in composition of the casting will occur. One of the most well-understood segregations of this type is inverse segregation, which the author prefers to call simply dendritic segregation. In this case, the partitioned solute is segregated preferentially to the face of the mold, especially if this is a chill mold. A similar effect will occur at the junction with a thinner section, which will act as a cooling fin (Fig. 10). However, in a complex thermal field and where the geometry of the casting requires a complex distribution of residual liquid to feed shrinkage, these chemical variations can be complex in distribution and not always easily predicted, except perhaps by a sophisticated computer simulation. Other segregations are driven by gravity and account for the concentration of carbon and other light elements at the top of large ferrous

Fig. 9

Cylinder casting arranged for inversion casting

Fig. 10

Schematic illustration of the effect of chills and unequal section junctions, causing the composition of parts of the casting to be above and below specification, leading to local brittleness and weakness, respectively

504 / Principles and Practices of Shape Casting

Rule 9: Residual Stress No quenching of certain light-alloy castings into water following solution treatment. (Boiling water is also not permitted for these castings, but polymer quenchant or forced-air quench may be acceptable if casting stress is shown to be negligible.) If, when quenching certain castings following high-temperature heat treatment, the time for cooling of the casting outer sections is shorter than the time required for heat to diffuse out from the interior sections (Fig. 11a), then the internal sections will cool and contract after the outer parts of the casting form an effectively rigid frame. Thus, the interior sections go into tension. It is a fact that quenching into water causes high residual stresses in large and complex castings. The most affected products are hollow castings with small openings to the outside world, and with interior walls. Such castings include cylinder heads and some compressor housings. The stresses will be tensile in some regions, mainly in the center of the casting volume, and compressive in others, mainly the outer walls. Small and simple-shaped castings are, in general, not seriously affected. The remainder of this section deals specifically with those castings that are particularly susceptible to the dangers of high quench stress. The use of a boiling water quench has been demonstrated to be of insignificant assistance in reducing the stresses introduced by water quenching. Furthermore, the stresses are not significantly reduced by the subsequent aging treatment. Immediately following the quench, the residual stress in susceptible aluminum alloy castings solution treated and quenched into water are well above the yield point of the alloy. Even after strengthening during the aging treatment, the stress usually remains at approximately 50 þ  20% of the yield stress. Thus, the useful strength of the alloy is reduced from its unstressed state of 100% to 70, 50, or even 30% (Fig. 11b). This massive loss of effective strength, usually affecting the interior sections in the casting, makes it inevitable that residual tensile stresses are a significant cause of casting failure in service. It is a scandal that many national standards for heat treatment specify water quenching, regardless of the size and complexity of the component. This disgraceful situation must be remedied. In the meantime, such dangerous standards should be overridden by agreement with the customer. The reduced mechanical strength when using polymer or forced-air quenching (Fig. 11c) is more than compensated by the benefit of increased reliability from putting unstressed castings into service. Thus, somewhat reduced mechanical strength requirements should be specified by the casting designer and/or customer. The reductions are expected to be in the range of 5 to 10% for strength and hardness.

Fig. 11

(a) Threshold at which aluminum (or steel) castings will suffer residual quench stress as a function of cooling rate between 500 and 200  C (930 and 390  F), and the average distance over which heat will be required to diffuse during the quench. (b) Logic behind the creation of quench stresses revealed as a problem of quench strain of approximately 1% applied to the internal features of the casting. (c) Spectrum of cooling rates commonly available in a 20 mm (0.8 in.) diameter bar of aluminum

Casting Practice—Guidelines for Effective Production of Reliable Castings / 505 features but which may be formed by a difficult-to-place core or a part of the casting that requires extensive dressing by hand. The result is a casting that does not clean up on machining and is thus, perhaps somewhat unjustly, declared to be dimensionally inaccurate. This rule is designed to ensure that all castings are picked up accurately, so that unnecessary scrap is avoided. Different arrangements of location points are required for different geometries of casting. Some of the most important types are listed as follows:
Six points, as illustrated in Fig. 12, are

Fig. 12

Classic six-point location system showing various degrees of sophistication, and the jig or fixture that would be used for dimensional checking or for machining. (a) Basic location system. (b) Halving of length errors. (c) Use of tooling lugs for clamping. (d) Jig or fixture attached to mark-out table or machine tool

Although the strength of the material will be lowered by the slower quench, the strength of the casting acting as a load-bearing component in service will effectively be increased.

Rule 10: Location Points All castings are to be provided with agreedupon location points for dimensional reference and for pickup for machining. Location points have a variety of other names, such as tooling points, pickup points, and so on. It is proposed that the term location points describes their function most accurately. It is essential that every casting has defined locations that will be agreed upon with the machinist and all other parties who must pick up the casting accurately. Otherwise, it is common for an accurate casting to be picked up by the machinist using what appear to be useful

required to define the position of a component with orthogonal datum planes that is designed for essentially rectilinear machining, such as for an automotive cylinder head. (Any fewer points is insufficient, and any more will ensure that some points are potentially in conflict.) Three points define plane A, two define the orthogonal plane B, and one defines the remaining mutually orthogonal plane C.
For cylindrical parts to be picked up in a three-jaw chuck, one axis must be defined as a datum reference, together with four additional tooling points, three of which define the plane orthogonal to the axis, and the last tooling point defines a “clock” position with respect to the axis.
For prismatic shapes, comprising hollow, boxlike parts such as sumps (oil pans), the pickup may be made by averaging locations defined on opposite internal or external walls. This is a lengthier and more expensive system of location often tackled by a sensitive probe on the machine tool, which then calculates the averaged datum planes of the component and orients the cutter paths accordingly. For maximum internal consistency between the tooling points, all should be arranged to be in one-half of the mold, usually the fixed or lower half. However, it is sometimes convenient and correct to have all tooling points in one-half of a core (defined from one-half of the core box) if the machining of the part must be defined in terms of the core. Datums located at the end of a long component, or at the opposite end to a critically located feature, give rise to casting rejections or unnecessary manufacturing difficulties. It is helpful if the part can have its datums for dimensions arranged to go through or near to the center of the part. The slight variations in size of the part, from one part to the next, then remain more easily within tolerance, since any variations in length are halved. Alternatively, the part can be datumed through the feature that is most critical to have in a fixed location, such

as the dipstick boss in the oil pan casting. The problem of variability of other dimensions is then greatly reduced. The tooling points should be defined on the drawing of the part and agreed upon by the manufacturer of the tooling, the caster, and the machinist that all parties will work from only these points when checking dimensions and when picking up the part for machining. The use of the same agreed-upon points by toolmaker, caster, and machinist leads to an integration of manufacturing between these parties. Disputes about dimensions then rarely occur, or if they occur, are easily settled. Casting scrap apparently due to dimensioning faults or faulty pickup for machining usually disappears. ACKNOWLEDGMENT Adapted from J. Campbell, The Ten Casting Rules: Guidelines for the Reliable Production of Reliable Castings—A Draft Process Specification, Proceedings, Materials Solutions Conference ’98 on Aluminum Casting Technology, ASM International, 1998 REFERENCES 1. J. Campbell, The Ten Casting Rules: Guidelines for the Reliable Production of Reliable Castings—A Draft Process Specification, Proceedings, Materials Solutions Conference ’98 on Aluminum Casting Technology, ASM International, 1998 2. J. Campbell, Casting Practice: The 10 Rules of Casting, Butterworth-Heinemann, 2004 SELECTED REFERENCES
J. Campbell, Castings, 2nd ed., Butterworth-

Heinemann, 2003


J. Campbell, Invisible Macrodefects in Cast-

ings, J. Phys. (France) IV, Colloque C7, supplement to J. Phys. (France) III, Vol 3, 1993, p 861–872
N.R. Green and J. Campbell, Statistical Distributions of Fracture Strengths of Cast Al-7Si-Mg Alloy, Mater. Sci. Eng. A, Vol 173, 1993, p 261–266
N.R. Green and J. Campbell, Influence of Oxide Film Filling Defects on the Strength of Al-7Si-Mg Alloy Castings, Trans. AFS, Vol 102, 1994, p 341–347
J. Runyoro, S.M.A. Boutorabi, and J. Campbell, Critical Gate Velocities for Film-Forming Casting Alloys, Trans. AFS, Vol 100, 1992, p 225–234

ASM Handbook, Volume 15: Casting ASM Handbook Committee, p 506-512 DOI: 10.1361/asmhba0005221

Copyright © 2008 ASM International® All rights reserved. www.asminternational.org

Filling and Feeding System Concepts John Campbell, University of Birmingham, England

FILLING the casting cavity with molten metal can be accomplished in various ways. Filling methods differ among the various casting processes, and techniques depend on the geometry of the cast part and the castability of the alloy. In the past, these systems have evolved largely from the experience and capabilities of individual foundry people. However, new research tools and computer simulation have assisted in the design of improved filling systems. This article introduces filling and feeding concepts from the general perspective of what constitutes good casting practice (see the preceding article, “Casting Practice: Guidelines for Effective Production of Reliable Castings,” in this Volume). This short review supplements a more detailed account (Ref 1) and is an attempt to provide a critical overview and update of concepts (Ref 1–4) in filling-system design. The casting industry has never systematically researched its filling-system designs for castings; consequently, current knowledge and practice is poor. This missed opportunity has cost the casting industry and its customers dearly. In particular, conventional unpressurized and pressurized filling systems (designed to fill molds under gravity) perform poorly, creating generous quantities of entrainment defects (bubbles, bifilms, and sand inclusions) and thereby efficiently degrading castings and their properties. Two recent developments that can help to avoid these problems include:
Preprimed filling systems
Naturally pressurized gravity filling systems

Prepriming (or the prefilling) of a filling system, using some variety of two-stage pour, is a powerful technique that deserves wider use. In the absence of prepriming, the new concept of the naturally pressurized gravity filling system is also described in this article in terms of its successes and limitations. It seems to the author that contact pouring in conjunction with either prepriming or a naturally pressurized system may be the only good solution at the present time to the gravity pouring of heavy steel castings using a bottom-teemed ladle. Perhaps the pouring of melts at any stage during the melting or casting process may ultimately be avoided in

new designs of future foundries. Thus, new melt-handling technology would be developed for future foundries whereby countergravity filling of molds would be recommended as capable of routinely producing substantially defect-free castings (see the article “Low-Pressure Countergravity Casting” in this Volume). As noted, this article briefly reviews concepts that may help clarify and quantify objectives for more effective mold-filling designs. The design of improved filling systems is now significantly facilitated by research tools such as the video x-ray radiography technique, in which the flow behavior of real liquid metals can be studied in detail in real molds (Ref 5). In addition, computer simulation has now come of age in this field, and its intelligent use can be invaluable.

Entrainment Defects in GravityPoured Castings Entrainment defects are extremely difficult to eliminate from gravity-poured castings, and the task for the designer of the gravity-driven filling system amounts to a damage-limitation exercise. This can be understood in terms of the so-called critical ingate velocity at which the melt has sufficient energy to entrain its own surface in a phenomenon called surface turbulence. This concept incorporates a further concept called entrainment that is a mechanical folding action, so that entrainment defects such as air and oxides (more specifically, bubbles and bifilms) can be introduced into the melt. The major difference between a bubble and a bifilm is the amount of air that each entraps; the bifilm is a kind of flat bubble that is more like a crack. Whereas the bubble is visible, the bifilm is usually so thin that it is invisible and only detected under a good microscope. Both can be exceedingly damaging to the quality of a casting. The critical fall distance (the fall distance under gravity at which the critical velocity is exceeded for that particular metal) is only approximately 12 mm (0.5 in.) for the light liquid metals, aluminum and magnesium, and approximately 8 mm (0.3 in.) for the dense metals such as copper, nickel, and iron. Thus, the vertical height through which the melt falls

in the running systems of all gravity-poured castings exceeds this velocity by a large margin. Therefore, entrainment defects are extremely difficult to eliminate from gravity pouring.

Traditional Filling Designs The traditional methoding systems have included the so-called unpressurized filling design, in which an attempt is made to reduce the velocity of the metal prior to entering the mold cavity. This worthy aim is attempted by expanding the flow channels. For instance, each doubling of the area is expected to result in halving the speed of flow. In terms of areas of the sprue exit/runner/gate, the relative areas chosen are often 1 to 2 to 2 or 1 to 2 to 4 and so on. The assumption has been that the volume flow rate, Q (m3/s), and velocity, V (m/s), relate to the area, A (m2), of the stream by the simple equation:



Q¼A V

(Eq 1)

The problem with this approach is that Eq 1 is perfectly true for the melt stream itself but not necessarily for the mold channel. Thus, an expansion of the channel to reduce the velocity usually has no effect on velocity; the melt flow progresses at its original speed and cross-sectional area, so the channel simply runs only part full. This is a potential disaster for melt quality, since the velocity is well above the critical velocity, and air and oxides (i.e., more precisely, bubbles and bifilms) can now be entrained. Thus, unpressurized systems are typically poor at reducing velocity and generally deliver poorquality metal into the mold (Ref 1). The filling-system design widely employed in the cast iron industry is the pressurized system. Here, the so-called gating ratio is often approximately 1 to 0.9 to 0.9, allowing the channels to run full, provided air is not introduced in the basin. However, poor conical basin designs and improperly tapered down-sprues, resulting from the poor design of automated green sand molding plants, usually eliminate these advantages, with the result that air is introduced at the entrance to the filling system. Thus, although the potential of the pressurized

Filling and Feeding System Concepts / 507 system was promising, in practice the benefits were eliminated by poor front-end design. Finally, the pressurization effect resulted in very fast delivery of melt into the mold cavity, where the high velocity (once again, much higher than the critical velocity) had the potential to create significant additional damage. Also, damage generated in the mold is the worst kind of damage, since such damage necessarily is incorporated into the casting. (In contrast, much of the damage created in the filling system appears to “hang up” in the filling system, never finding its way into the casting.) This problem of good general principles but poor front ends to the system, or other features that introduce defects, illustrates a significant aspect of filling-system design: The whole of the system must work well. Any small detail that is incorrect can introduce destructive amounts of irreversible damage, a significant proportion of which is likely to arrive in the casting. Unfortunately, gravity filling systems are hypersensitive to small errors as a result of the high velocities involved. This fact has been the source of major problems to the casting industry. This situation contrasts with countergravity filling systems where the melt velocity can be controlled at all stages below the critical velocity, so that it is possible to introduce zero damage during filling, despite the presence of errors such as the ledges arising from the mismatch of channels and so on. It follows that countergravity systems, when properly operated and despite modest errors in the filling system, represent robust technology that can routinely deliver good products and have the potential to revolutionize the performance and profitability of the industry. The poor front-end designs used for many systems have effectively undermined many otherwise reasonable filling-system designs. Similarly, good front-end features, such as the offset stepped pouring basin (the only basin to have been researched and shown to be capable of avoiding air entrainment, Ref 6), have often been mistakenly criticized for poor performance because of the deleterious actions of other features in the system, thus adding to the overall confusion relating to filling-system design. A further deleterious feature that has a long pedigree is the well at the base of the sprue, often said to be placed there to cushion the fall of the metal. However, recent work has illustrated that the high velocity at this location, combined with the generous volume provided by the well, maximizes the damaging churning action of the melt. This turbulence entrains clouds of bubbles and, without doubt, large quantities of oxide. The sprue/runner junction for the naturally pressurized filling design described subsequently is a simple and basic turn, providing no extra volume in the system that can allow surface turbulence. When properly designed and made, this sprue/ runner junction does not contribute a single bubble to the downstream flow. Potential solutions to the central problem of the avoidance of entrainment defects are described in order in the following sections.

Prepriming Techniques The prefilling of the filling system prior to the filling of the mold is a powerful technique for eliminating entrainment problems. It deserves to be used far more widely. There are various partial solutions to the prepriming approach:
If an offset pouring basin is filled with metal

that is allowed to immediately fall into the sprue, the filling of the sprue is ragged and messy, incorporating considerable quantities of air. The resulting air/melt mixture is usually in the region of 50% air and 50% metal for the several seconds that it takes for the basin and sprue to prime. This is because the sprue is designed to work from a particular height of metal in the basin. Initially, the depth of metal in the basin is much less, so that for the first several seconds, the sprue entrance is insufficiently pressurized and thus runs only part full. The use of a stopper in the entrance to the sprue allows the basin to be filled to some minimum design level, prior to lifting the stopper and allowing the melt to enter the sprue (Fig. 1a). When observed by video x-ray radiography, the prefilling of the basin by the use of a stopper is seen to significantly improve those critical early seconds of sprue priming.
The use of a valve at the base of the sprue is valuable to allow the prefilling of the basin and the sprue (Fig. 1b). This technique is even better than the mere prefilling of the basin. There is an additional benefit since, on opening the valve, the melt in the sprue starts to flow from a standing start and must accelerate to its full flow speed. This speeding up from zero occupies only a second or so but is enough to reduce the initial jetting of the melt along the runner. The filling of a disc brake mold with an Al/SiC metal-matrix composite (MMC) was achieved only by such a system, as described by Cox and coworkers (Ref 7). The valve in this case consisted of a thin sheet of ceramic paper that was supported on a ceramic foam filter. After the sprue was filled, the ceramic paper was lifted clear by raising a wire hook attached to the edge of the paper. The use of a sheet steel slide gate in the runner has been reported to raise the quality of aluminum-base MMC castings. The excellent results were convincingly quantified by Weibull statistical analysis (Ref 8).

for a few seconds prior to melting and automatically releasing the flow are known and used in the industry, although little has been published. In laboratory tests, aluminum foil backed by a filter was not successful because the aluminum oxide layer on the aluminum foil was too strong and resisted breakthrough, despite the melting of the aluminum metal in the center of the foil. Thus, bursting discs with strong oxide films should be avoided. Conversely, metal foils made from zinc, copper, nickel, or iron would be expected to be useful. The thickness of the discs would have to be determined by experiment or perhaps by computer simulation. Any of these techniques can be combined with, and will improve the performance of, a naturally pressurized system design, as described in the next section. However, prefilling enhances the performance of even very poorly designed filling systems, which is why the technique is so valuable. The prefilling action is temporarily turbulent until the prefilling action is complete. At that stage, the melt in the running system is expected to contain both bifilms and bubbles. If a short dwell time of 1 or 2 s is provided, little heat is lost, but the melt has time to eliminate its bubbles by forcing them out through the permeable sand mold. (The second and third prepriming systems in the bulleted list can only be used safely with permeable, or sand, molds because the air needs to escape during the prepriming event. In permanent molds, the air cannot escape, and as it quickly heats and expands, it can reverse the melt, ejecting it from the sprue.) Most entrained oxides will already be attached at some point to the mold wall, so that gentle buoyancy or residual advection will assist in hinging the films against the surface of the runner, where, in general, they are expected to remain. Any attachment will not be particularly strong or secure, but the films will lie mainly in the region of low or zero flow rates in the region of the boundary layer against


The use of a valve at the gate into the mold

allows the whole filling system to prefill prior to the melt entering the mold (Fig. 1c). Such a system would be expected to deliver excellent castings even from what may be viewed as a lamentable running system. All of the techniques can be implemented with operator-controlled devices. However, sheet metal bursting discs that arrest the flow

Fig. 1

Prepriming techniques for (a) the basin, using a stopper, (b) the basin and sprue, and (c) the whole filling system. All of the prepriming methods can use devices such as mechanical valves or fusible bursting discs.

508 / Principles and Practices of Shape Casting the mold walls. Thus, much of the oxide entrained during the prepriming stage is likely to remain gently glued to the inside walls of the running system and is not in danger of entering the casting. The bifilms would not be expected to remain in place if air were to be entrained and flow through the system, since slugs of air create buffeting and impact forces by their negative momentum; the system only works well if it remains completely filled with liquid metal. As mentioned previously, the prefilling of any or all of the filling system is extremely valuable and strongly recommended, even if only achieved in part. However, if this cannot be accomplished, then the filling system must be designed to fill without such assistance. This is a significantly more difficult task, since it is a challenge to exchange the 100% air in the system for 100% metal prior to the melt entering the mold cavity. The challenge is even more difficult to meet if this 100% metal is to arrive in the mold cavity in a reasonably undamaged condition (i.e., reasonably free from entrainment defects such as bifilms and bubbles). This is the challenge that the naturally pressurized system attempts to meet.

Naturally Pressurized Filling System The naturally pressurized system is characterized by the narrowness of its flow channels. Its central concept, fundamental to the whole approach of the naturally pressurized system, is that of maximizing constraint on the flowing liquid from nearby walls of the channels. Thus, despite the melt having a velocity well above the critical velocity, it tends to not suffer damage because there is effectively no room for the advancing meniscus to fold over on itself to entrain a defect. The concept of a choke in the system becomes redundant, because the whole system, along its complete length, acts as a choke and is effectively a continuously choked design. The six individual parts of the filling system are:







Offset stepped pouring basin Sprue Sprue/runner junction Runner Gates Feeding via feeders (or running system, if possible)

Offset Stepped Pouring Basin. It is assumed that the offset stepped basin design is used in the naturally pressurized filling system (Fig. 2). While not being exclusive to the naturally pressurized system, it appears to be the best basin design thus far and is capable, in some cases, of delivering 100% metal (i.e., zero content of entrained air bubbles corresponding to a discharge coefficient of 1.00) (Ref 9). It seems to work adequately well for

aluminum-silicon alloys, particularly when a basin response time (i.e., its draining time if simply allowed to empty) of approximately 1 or 2 s is selected. This usually gives adequately slow response so that the pourer can keep the basin filled (Ref 1) and appears to give adequate detrainment time for most of the air entrained by the pour into the basin. A widespread error is to calculate the driving pressure for filling from the height of the basin itself. However, if the basin is only slightly underfilled at any time, the reduced pressurization of the sprue entrance will lead to air being taken down the sprue (although this problem is not apparent to the operator, who is under the impression that the system is working satisfactorily). This is the common reason why most designs of offset stepped basins fail. To avoid this, for the purposes of the design of the whole filling system, it is essential to assume a level of fill in the basin that can be reliably maintained. If the actual fill level is always maintained above this minimum, the sprue can remain properly pressurized and therefore completely full at all times. A doubling of the height of the basin above the minimum level is not too much to leave a good margin of safety. The scheme is illustrated in Fig. 2. Even so, the problem of basin design is not completely solved and, in fact, may not be solvable for some metals. Metals that generate strong oxide films, such as aluminum bronzes and some stainless steels, may overload the system with oxides that cannot be detrained. A mass of strong oxide films waving about in the flow like waterweed in a river will prevent detrainment in the basin and act to funnel bubbles down the sprue. Similarly, the pouring of steel from bottomteemed ladles also may represent an unsolvable situation. In this case, the great depth in the ladle naturally results in an extremely high exit velocity from the bottom nozzle. The high-velocity melt entering the basin creates a churning action that overwhelms the system with oxides and bubbles that have insufficient time to detrain. The use of an argon shroud around the pouring stream is an attempt to address this problem and has some limited beneficial effect. However, for a more rigorous solution, a sophisticated version of contact pour may be required in which the pouring basin is effectively exchanged for a bottom-poured ladle in which the outlet nozzle from the ladle makes direct contact with, and seals against, the entrance to the sprue (Fig. 3). The technique was demonstrated successfully for the first time for a 50,000 kg (55 tons) steel roll casting (Ref 10) and is a promising general method for steel castings. A production system using small, intermediate ladles sitting on top of the molds and able to contain the complete volume of melt required for the pour can be envisioned. The intermediate ladle is filled from the large ladle transferring metal from the melting furnace. A few seconds after the intermediate ladle is filled, its stopper is raised and the mold is filled.

A series of intermediate ladles on a series of molds may be workable, with the early ladles being transferred forward to be used a second time for later castings in the pour series. As an alternative to a stoppered ladle, a basin or ladle with a fusible bursting disc may be simple, economical, and practical.

Fig. 2

Offset stepped basin illustrating the target fill level, the minimum fill level, and other key

features

Fig. 3

Teeming of steel into an intermediate ladle, kept filled to at least a minimum depth, h

Filling and Feeding System Concepts / 509 Sprue. The basis of the sprue design is the prediction of the form of a freely falling liquid stream. This frictionless flow is the starting point for the calculation of the areas for the sprue and the remainder of the running system. Assuming Eq 1, for a volume per second flow rate (Q), the area of the cross-sectional area of the stream (A) after a fall of distance h under the acceleration due to gravity (g) from the upper surface of the melt is given by: A2 ¼ Q2 =ð2ghÞ

(Eq 2)

This area/height relation is the famous result describing a rectangular hyperbola (not a parabola, as has often been assumed in error). It follows that the really important relation giving the profile (i.e., the diameter, D) of a rotationally symmetrical stream at height h is: D4 ¼ 8Q2 =ð2 ghÞ

(Eq 3)

This relation of D versus h1/4 is a significantly more complex curve, shown schematically in Fig. 4 (with height in dimensionless units of 8Q2/(p2gh)), emphasizing the nonstraightness of its taper, particularly near the top of the sprue, and the approach to asymptotes at each limit. Whereas in the past it was common to calculate the entrance and exit sizes of the sprue and connect these with a straight taper, the author now uses the hypobolically tapered sprue for all castings above approximately 300 mm (12 in.) high. This nonlinear taper follows the form of the falling stream, pressurizing at all points of the fall, and effectively excludes air. The simple naturally pressurized concept arises as follows. The theoretical shape (i.e.,

the cross-sectional area of the stream at every point) is based on the theoretical prediction assuming a frictionless fall of the melt (Eq 3). When applying this theoretical shape in practice, a mold is effectively created around it, so as just to touch and contain the flowing stream. The action of a solid surface designed to just touch the surface of the falling melt changes everything. Friction is now introduced for the first time. The natural frictional back-pressure now builds up along the length of the sprue (and all the subsequent channels if contact with the walls is maintained by appropriately narrow sections, as illustrated in Fig. 5(b), causing the liquid to gently pressurize the walls of the channel. This natural pressurization is of enormous value in assisting to keep air out of the system and is a fundamentally important feature of the naturally pressurized design. (In contrast, the traditional oversized running systems illustrated in Fig. 5a do not benefit from the friction to reduce the velocity of flow and usually cannot be pressurized. As a consequence, the systems introduce copious quantities of air and oxides.) Sprue/Runner Junction. At bends in the system, particularly the sprue/runner junction, it is essential to provide a radius to the inside of the turn. The radius should be close to the diameter or width of the sprue exit (Ref 11). If no radius is provided, the junction entrains air to create bubbles and oxides. It has been widely assumed that significant friction is involved at any bends in the system, causing up to a 20% loss of velocity. If true, the runner cross-sectional area should be expandable by 20% to take advantage of the 20% reduction in velocity, giving a gating ratio of R = 1/1.2/1.4. A succession of right-angle bends, when used in this way, can be successful in making significant reductions in final velocity at the end of the system. There is some experimental work (Ref 12) on a series of right-angle bends in a runner that appears to give confirmation of this theoretical prediction from an observed series of speed-reduction ratios close to 1.0, 1.2, 1.4, 1.7, 2.1, 2.5, and

so on. Traditionally, the right angle was always avoided because of the turbulence that it could create. This was true for oversized sections. However, for the slim sections in the naturally pressurized system, the melt is constrained so that little room is available to allow the metal to damage itself by folding over its advancing front to entrain bifilms or bubbles. Therefore, instead of being a threat, the right angle may constitute a valuable speed-reducing feature in the new system. This L-junction has been studied in detail by Hsu and others (Ref 11). Runner. Alternatively, the runner need not be expanded by 20% and instead may have the same area as the sprue, giving R = 1/1/1. This will significantly assist to pressurize the melt in the runner and gates so that there will be a greatly reduced danger of the entrainment of air. If this option is taken, the choking action of the melt entering the runner will lead to a 20% increase in filling time and a further 20% at the gates, giving a total extension of approximately 40% to the filling time due to the losses at the L-shaped turns alone. These losses in the running system are often overlooked, perhaps at least partly explaining why many calculated filling systems appear to run more slowly than predicted. Alternatively, the full 20% increase in area for the runner need not be taken. If only 10% were selected, giving R = 1/1.1/1.2, the runner would benefit from an extra 10% pressurization at each turn that would suppress entrainment defects to some degree, but filling time would be extended by approximately 10 + 10 = 20%. After the initial increase of the area of the runner to allow for losses at the sprue/runner junction, its area can remain constant up to the first division or gate. This constant area will contribute to the gentle pressurization along the length because of the frictional loss accumulating along the length. The discussion about a 20% loss of velocity at a right-angle bend has been called into question by the more recent work of Hsu and coworkers (Ref 11). In the early results of their

Fig. 4

Shape of the falling stream. The straight taper, taken as an example from 50 to 500 mm (2 to 20 in.), is seen to be a poor approximation.

Fig. 5

(a) A poor traditional filling design contrasted with (b) an improved system

510 / Principles and Practices of Shape Casting detailed study of the L-junction, the rule of 20% loss per right-angle junction compared to assumed negligible losses along the straight sections of the system is seen to be oversimplistic. It appears that the frictional back-pressure associated with the bend is real, but the accumulated effect along the sprue and along the runner appears to exceed the frictional loss at the bend. This is probably because the bend acts over a rather short length of flow compared to the longer lengths of the sprue and runner. Hsu’s results suggest that the accumulating frictional losses along the lengths of channels would permit their gentle expansion, allowing a progressive reduction in speed without the penalty of reduced volume flow rate or the loss of pressurization in places. This situation clearly requires more research. The outcome of such targeted research may herald the next development of filling-system design. In the meantime, it seems valuable to be cautious, not expanding the area of the flow channel at any point and thereby ensuring that the system remains pressurized everywhere to ensure the exclusion of air. Gates. As is well known, the placing of a series of equal gates along the length of a uniform runner will favor delivery through the far gates. The delivery through the gates should be balanced as far as possible by systematically reducing the area of the runner. The traditional practice of providing steps, reducing the area of the runner as each gate is passed, is definitely not recommended because of the deflection of the flow and the turbulence caused at sharp steps. Far better is a linear taper, gradually reducing the area of the runner along its length. For many aluminum-silicon alloy castings weighing 10 to 50 kg (22 to 110 lb) and up to 0.5 m (1.5 ft) or so long, approximately 50% reduction in area is often closely correct. (Tapering to zero area is also definitely not recommended. This is an overcorrection that causes the early gates to be favored and exposes foundry personnel to the danger of accidents with dagger-shaped runners!) Because many castings have only a single mold joint and must be gated on the joint, the gates have to be horizontal. If the gates now enter the mold cavity at more than the critical distance above the base of the casting, allowing the melt to fall inside the mold cavity, significant damage is expected to the properties of the casting. Gating at the mold joint can therefore be a major risk to casting integrity. Bottom gating, where the gates enter at the lowest possible level of the casting, is necessary to achieve good and reliable results. However, horizontal gates often cause the metal to jet over the base of the mold. Such uncontrolled directional propagation can lead to metal damage in the form of oxide flow-tube defects or, in the case of silica green sand molds, mold expansion defects such as rattails that damage the mold surface at the edges of the jetting stream. By contrast, vertical bottom gates, in which gravity assists both the filling of the gates and the slowing of the melt as the gates prime, work

well. They work even better if expanded along their length (i.e., expanded in area with increasing height) (Ref 1). It is not so much a luxury but self-interest to have an additional joint beneath the casting, formed by a third mold part or craftily designed core (Fig. 6). Such features usually quickly recover their cost. Whenever possible, gate directly onto a core, particularly if the local section thickness is small, for instance, preferably 4 mm (0.16 in.) thickness or less. This rule will alarm traditional foundry people, because a well-known rule was “never gate onto a core.” Never gating onto a core was correct when poor running systems carried up to 50% air with the metal, because the air would oxidize away the core binder, eroding the core and filling the casting with sand inclusions. However, with a properly designed naturally pressurized system, no air is present. Although the core will clearly become hot, it remains unharmed. The benefit to the flow of metal is the redistribution of melt over a larger area prior to emerging into the free volume of the mold cavity. Thus, the thin wall defined by the core acts as a final extension to the runner system, making a valuable reduction in the final velocity of the melt. Gates require careful design. The requirements are:
The total gate area should, if possible, be

sufficient to ensure that the ingate velocity at the point of entry to the mold is less than 0.5 m/s (1.5 ft/s), although in practice, up to 1.0 m/s (3 ft/s) is often tolerable.
The total number of gates must be sufficient to ensure that the sideways (transverse) velocity inside the mold does not exceed the critical velocity (Ref 1).
It is essential to ensure that the gates do not form a hot-spot at the junction with the casting. The rules for ensuring that a hot spot is not created are well established (Ref 1). For the most susceptible geometry of casting/ ingate junction, the relative geometric modulus should be less than 0.5 or greater than 2.0. Sometimes, however, it is useful to use particularly narrow slot gates, with modulus much less than 0.5 of that of the local piece of the casting. These become hot during filling but rapidly lose their heat to act as

Fig. 6

cooling fins after the filling is complete. This variety of gate can be used to create a valuable temperature gradient for directional solidification and enhanced feeding efficiency (Ref 1). Feeding. The author is not in favor of feeding castings, unless absolutely necessary (Ref 1). Most castings seen in foundries are hugely overfed. This almost certainly is the result of a number of factors. It is understandable and commendable that the founder should exercise caution, but in practice, this caution is carried out to excess for the following reasons:
The damage to the melt by poor running sys-

tems is usually so bad that most of the metal used to prime the system is no longer fit for contributing to the casting. Thus, it is dumped into the feeder, making the feeder oversized for its purpose of feeding. In this case, the primary role of the feeder is merely to act as an overflow. The founder justifies the size of the feeder on the apparently reasonable grounds that if the feeder size is reduced, the casting exhibits unacceptable shrinkage porosity. This is true. However, the real reason is that the melt in the mold cavity contains a major population of large bifilms created by the poorly designed filling system, so that porosity resembling shrinkage porosity is rife. Only by increasing the feeder size (i.e., getting rid of a high proportion of the damaged metal) is the problem apparently alleviated. However, the lesson is clear: Instead of increasing the feeder size, it would be more economic and significantly more effective to improve the filling system to avoid damaging the melt in the first place.
The computer simulation packages that check for solidification shrinkage in the casting are usually programmed with significantly inflated shrinkage factors, resulting in oversized feeders. Because of the first feature in this list, the oversized feeders are currently required and thought to be necessary for feeding because of a lack of awareness of the overflow dump role of the feeder. The problem of the overestimation of feeding percentages has remained unsuspected and become the norm.

Bottom-gated systems achieved by (a) a three-part mold with accurately molded running system, (b) a two-part mold making use of a core, and (c) a two-part mold using standard preformed sections

Filling and Feeding System Concepts / 511 When good filling systems are exchanged for poor originals, the oversized feeders can usually be greatly reduced, substantially increasing the casting yield. It follows that for good metal and a good filling system, simulation packages must reduce their shrinkage factors to more realistic values. The reduction in feeder size is the result of two benefits: the elimination of the role of the feeder as an overflow dump, and the reduction in the sensitivity of the melt to feeding problems because of the reduction in bifilm content. With excellent-quality melt and an excellent filling system, the casting may have such a low bifilm content that porosity cannot be initiated, and the resulting surface sinks (the corresponding collapse of the casting to feed any residual internal shrinkage deficit) may be negligible (i.e., not below the dimensional or machining allowance). In this case, the casting effectively self-feeds without the necessity for additional melt from a feeder. This ideal situation is worth targeting but may not be completely achieved, so some residual feeding may be necessary. When reducing the feeder volume, any reduction in its height should be made with caution, if at all. The metallostatic pressure due to height is valuable to pressurize the melt in the freezing casting and thus suppress the opening of any residual bifilms. Bifilms are usually so thin as to be invisible but are so numerous that they should always be assumed to be present. If bifilms are allowed to open, properties are reduced. It is essential to realize that in the early phases of this effect, the bifilms may not have opened sufficiently to create visible porosity. Thus, the casting may still appear perfectly sound, so that the loss of properties may appear inexplicable. Only if the bifilms open by some fraction of a millimeter will they start to become visible as microporosity. In this case, properties will have fallen even further (Ref 2). If bifilms must be tolerated in melts, tolerable properties will only be maintained if the bifilms are maintained firmly closed. If closed, they can support a shear stress but not a tensile stress at right angles to their plane. Although always damaging to properties, keeping the bifilms closed limits the ensuing damage.

Filters At least one study has definitively shown that the beneficial action of a ceramic foam filter derives approximately 5% from filtration (depending on the quality of the metal) and 95% from improved flow (Ref 13). Therefore, it seems likely that the benefits of using filters in filling systems arise mainly from the reduction in speed and turbulence of the flow. In this way, the melt downstream of a filter remains cleaner not because of filtration but simply by being less turbulent and therefore creating fewer defects. To retain the benefits of the reduction in speed, the filling channels after the filter must

be adjusted in area to suit the new velocity. Thus, if the filter can reduce speed by a factor of 5, an expansion of downstream area by a factor of 4 will retain some pressurization but gain the benefit of speed reduction to one-quarter of the speed entering the filter. The correct sizing of the postfilter part of the runner is usually overlooked (but, as a kind of perverse benefit, is often not necessary because all the ductwork, both incoming and outgoing from the filter, is already grossly oversized). The filter is most valuable in those systems that entrain air in which the front-end design is poor but cannot easily be rectified. This often happens on automatic green sand lines because the sprue and basin are poorly formed. The filter assists the backfilling of the front-end of the system. It is effectively acting as a device to aid prepriming of the system. Naturally, the provision of a bubble trap to divert buoyant phases such as air (and, in the case of ferrous castings, slag) away from the main flow can also be valuable in this case (Fig. 7) (Ref 14). It is essential to control any spraying or jetting action from the exit side of the filter, especially where the incoming flow is under high pressure from a tall sprue. Such jetting creates consequent oxidation damage downstream of the filter. Figure 7(b) illustrates an arrangement in which the melt emerges against gravity, quickly covering the exit surface and therefore quickly suppressing jetting, so that the subsequent flow remains undamaged. The arrangements in Fig. 7(a) and (c) illustrate the provision of shallow (less than the critical fall height) sumps beneath the filter. These quickly fill and thus aid the effective onward transmission of clean metal. Because the most important inclusions in aluminum alloys, and many other alloys, are films (not particles), it is interesting to note that inexpensive glass cloth is often found to be as effective as the various varieties of ceramic block filters. However, the cloth is often difficult to apply in practice, explaining the prevalence of the more costly alternative.

Performance of Alternative Filling Systems With combined reductions in the volume of both filling and feeding systems, metallic yields

Fig. 7

can easily rise from 45 to 70% or more. Provided the metal that is poured is of adequate quality (which is not to be assumed), defects such as porosity and hot tears can be reduced, if not eliminated (Ref 4). In addition, the mechanical and corrosion properties of the casting can be significantly enhanced (Ref 15–17). Although the naturally pressurized filling system gives a significantly superior performance to current unpressurized and pressurized systems, the approach remains far from optimal. More research is needed to further refine the concept to take advantage of the progressive accumulation of frictional back-pressure to progressively expand sections and thus reduce speeds without loss of volume flow rate or loss of pressurization. For this reason, computer simulations of pressure distribution against the walls of the filling system, to ensure the absence of unpressurized areas, will be more important than current simulations of velocity of flow. Alternative filling systems not discussed here include the horizontal transfer of metal into a mold via the level pour system (Ref 1). This is a clever mechanized system that deserves greater use for appropriate boxlike castings and possibly other shapes. Tilt casting can also be valuable if carried out correctly. In particular, the tilt process is usually started from the horizontal level. This is a mistake. It results in a poorly controlled turbulent fill. A start angle of approximately 20 above the horizontal usually avoids this problem (Ref 1). Further problems occur if the melt suffers a waterfall effect at some stage during the tilt because of the geometry of the casting. Thus, there are a number of reasons why many tilt pouring operations are currently not free from significant levels of scrap. Direct pour, in which the melt is poured via a fiber sleeve and filter sited on top of the casting, is risky but may be worth the risk. It has some key advantages: It is extremely simple, takes up minimal space in the mold, and provides a combined filling and feeding action. It provides significant reproducibility to the pouring action and, for this reason, can deliver 100% good or 100% scrap results. Thus, some trial and error may be involved in finding a suitable location where it gives good results for a particular casting. It appears to work best for small fall distances below the filter, its effectiveness decreasing with fall

Filter arrangements combined with bubble traps. Source: Ref 15

512 / Principles and Practices of Shape Casting height up to 200 mm (8 in.), after which its benefit falls to zero (Ref 13). Also not included in this review is the problem of filling very thin sections in which flow is controlled by surface tension and which suffer the defect-generating problem of microjetting (Ref 2), a kind of naturally occurring microturbulence. Despite its potential importance, this phenomenon is not well understood, requiring further research.

Recommended Foundry Practice Every new foundry should, wherever possible and appropriate, eliminate all pouring of metal and employ a good countergravity filling system to achieve mold filling at velocities below the critical velocity, thereby achieving sound and high-performance castings on a routine basis. Processes that include unconstrained pouring by gravity (such as most low-pressure permanent mold techniques in which the pressure vessel is filled by pouring) have problems delivering products of reliable quality. If countergravity cannot be implemented and gravity must be used, then the next-best option is the prepriming of parts of the filling system, or better, if possible, the whole of the filling system if a permeable mold is used. If prepriming is not possible, then adoption of tilt casting (with suitable precautions as listed in Ref 1) or the naturally pressurized filling system is the next-best option. Key features of the naturally pressurized gravity system are
Control air entrainment at entrance to sys-

tem by offset stepped basin or, even better, by contact pour.


During the early stages of filling, target to

fill and pressurize the whole system with melt in one pass (simulate pressure distribution if possible to ensure pressurization at all locations along the length of the running system).
Control entrainment events at the entry to the mold cavity, limiting jetting by reducing velocity to less than 1 m/s (3 ft/s) by such devices as accumulated frictional loss (possibly by appropriate use of a filter), surge control (Ref 1), or some prepriming filling (Ref 1). Pressurize the casting during freezing with high top feeders to suppress the opening of bifilms. The benefits are suppression of porosity (and, to some extent, tears) and maintenance of tolerable properties (excellent properties will only result from the elimination of bifilms).

6. 7. 8.

9. 10.

ACKNOWLEDGMENT

11.

This article is based on a paper first submitted to TMS Shape Casting Symposium 2007 but is published here expanded and updated.

12.

REFERENCES 1. J. Campbell, Casting Practice, Butterworth-Heinemann, 2004 2. J. Campbell, Castings, 2nd ed., Butterworth-Heinemann, 2003 3. J. Campbell; Shape Casting: The John Campbell Symposium, M. Tiryakioglu and P.N. Crepeau, Ed., TMS, 2005, p 3–12 4. J. Campbell, Mater. Sci. Technol., Vol 22, 2006, p 127–143 and correspondence p 999–1008 5. B. Sirrell, M. Holliday, and J. Campbell, The Seventh Conf. on the Modeling of

13. 14. 15. 16. 17.

Casting, Welding and Advanced Solidification Processes (London), M. Cross and J. Campbell, Ed., 1995 X. Yang and J. Campbell, Int. J. Cast Met. Res., Vol 10, 1998, p 239–253 B.M. Cox, D. Doutre, P. Enright, and R. Provencher, Trans. AFS, Vol 102, 1994, p 687–692 M. Emamy, R. Taghiabadi, M. Mahmudi, and J. Campbell, Second Int. Al Casting Technology Symposium, Advances in Aluminum Casting Technology II, Oct 7–10, 2002 (Columbus, OH), ASM International, p 85–90 T. Isawa and J. Campbell, Trans. Jpn. Foundrymen’s Soc., Vol 13, Nov 1999, p 38–49 X. Kang, D. Li, L. Xia, J. Campbell, and Y. Li, Shape Casting: The John Campbell Symposium, M. Tiryakioglu and P.N. Crepeau, Ed., TMS, 2005, p 377–384 F.-Y. Hsu, M. Jolly, and J. Campbell, Second International Symposium, Shape Casting, P.N. Crepeau, M. Tiryakioglu, and J. Campbell Ed., TMS, 2007, p 101–108. H. Nieswaag and H.J.J. Deen, 57th World Foundry Congress (Osaka), 1990, part 10, p 2–9 T. Din, R. Kendrick, and J. Campbell, Trans. AFS, Vol 111, 2003, paper 03-017 A. Habibollah Zadeh and J. Campbell, Trans. AFS, Vol 110, 2002, p 19–35 N.R. Green and J. Campbell, Trans. AFS, Vol 102, 1994, p 341–347 C. Nyahumwa, N.R. Green, and J. Campbell, Trans. AFS, Vol 106, 1998, p 215–223 H. Sina, M. Emamy, M. Saremi, A. Keyvani, M. Mahta, and J. Campbell, The Influence of Ti and Zr on Electrochemical Properties of Aluminum Sacrificial Anodes, Mater. Sci. Eng. A, Vol 431 (No. 1–2), 2006, p 263–276

ASM Handbook, Volume 15: Casting ASM Handbook Committee, p 513-521 DOI: 10.1361/asmhba0005355

Copyright © 2008 ASM International® All rights reserved. www.asminternational.org

Processing and Finishing of Castings After solidification and cooling, further processing and finishing of the castings is required. The required operations and levels of finishing depend on the process (green sand, no-bake sand, lost foam, investment casting, etc.) and on the type of metal being cast. In general, however, the castings must first be removed from the tree assembly, consisting of the casting with additional areas of metal (sprues and risers) used to direct metal flow during pouring. Cut-off saws are usually used for this task. Risers must be removed by either cutting them off or knocking them off by force. Once the castings are separated from the trees or risers, additional finishing operations are done to remove all flash and parting line metal and to achieve the required finish requirements. These finishing operations include:

Shakeout and Core Knockout After solidification and cooling of castings in sand molds, the casting must be removed from the sand mold. In the case of green sand or resin-bonded sand molds, a considerable amount of energy is required to remove castings and break up lumps of sand still strongly bonded together. In contrast, unbonded sand systems (such as lost foam) do not require as severe mechanical agitation for shakeout. Shakeout from bonded sand molds is done by shaker pan conveyors, vibrating deck or grate, or any other type of vibratory, oscillating, or rotational mechanical action. Some castings will crack if shaken out too soon. Others require a controlled cooling rate to attain a

specific microstructure. In establishing the time for shakeout, consideration must be given to the metal, pouring temperature, casting size and design, coring, and sand mixture used. Shakeout equipment is described in the article “Green Sand Molding” in this Volume. With the exception of those castings poured from low-melting-point metals (aluminum or magnesium), the heat of the metal burns out the resin bond in the sand sufficiently for the mechanical action of the shakeout device to cause almost all of the sand to fall away from the casting. Likewise, core knockout, or decoring, is the process of removing internal resinbonded sand cores from castings. Ferrous metal castings are poured at much higher temperatures (1650  C, or 3000  F) than nonferrous

 Shotblasting  Grinding (from coarse to fine, depending on

the casting finish quality)

 Trimming  Machining (milling, drilling)  Quality testing and inspection

Depending on the alloy and requirements, the processing of castings may also include repair welding, heat treatment, and/or hot isostatic pressing. These additional operations and the machining of castings are not discussed in this article. This article briefly describes the general operations of shakeout, grinding, cleaning, and inspection of castings. Particular emphasis is also placed on automation technology, which can be an important productivity factor for foundries. Cleaning (fettling) operations are traditionally a very labor-intensive component of the cost of a casting, often exceeding 60% of the total casting cost. A very high percentage of cleaning operations may still be performed manually. However, robotic technology (Fig. 1) is becoming more prevalent. Until recently, casting variation and the inability to program robots easily in an offline environment prevented many foundries from automating. Both of these problems have been overcome, and user-friendly solutions are now more widely available. The most common application of robots in any manufacturing environment consists of pick-and-place tasks. This is also true in the foundry industry.

Fig. 1

Robotic handling in grinding. Courtesy of Vulcan Engineering Company

514 / Principles and Practices of Shape Casting castings. Consequently, more energy is required to remove core sand from aluminum castings than from gray iron castings. In gray iron, relatively low-energy shaking, vibrating, tumbling, and/or shotblasting is used to force loose core sand out through the casting openings. Core sand in ferrous castings has much less strength because the ambient heat of the metal, as it solidifies in the mold, burns the resin-bonding agents out of the sand. When the resin is burned out, the grains of sand cannot attach to one another, making high-production cleaning of ferrous castings less complex. The major focus of this section is the cleaning of aluminum castings, which are molded in a temperature range of 705 to 815  C (1300 to 1500  F), approximately one-half the temperature required to cast ferrous metals. This comparatively low molding temperature does not allow sufficient heat to transfer from the casting walls to the resin-bonded core sand. Only the cored passageway closest to the metal will show heat penetration and breakdown. Sand remaining on the surfaces of the castings can be removed by abrasive blast cleaning, as described in the section “Blast Cleaning of Castings” in this article. Core Knockout. Casting complexity, type of resin, type of sand, wall thickness, type of core (solid or hollow), metal-to-sand ratio, and gating/ risering affect the amount of energy required to remove the sand core from the casting. The most common method used to clean core sand from aluminum castings is mechanical vibration with pneumatic tools. In most instances, the core sand has less inherent strength than does the metal; therefore, it breaks down without metal damage. Many foundries use hand-held, vertical, A-frame, or fixtured air tools that operate in the 1000 to 3000 blows per minute range. The hand-held air tool method of decoring castings involves an operator placing a casting on the floor or on a table and locating the chisel of the air tool on the risering. When the operator pulls the trigger on the tool to allow air to pass through and create vibration, the casting shakes, and core sand breaks down. Each type of casting and each core sand is unique. For optimal cleaning, the proper type of vibration and/or blows per minute must be suited for each application. This may not be easily determined, so cracked castings or excessive air tool component failure may result. The advantages of hand-held air tools include:  Low cost  Flexibility (can be used on many different

castings)  Effectiveness applications

for

low-volume

cleaning

The disadvantages are:  Noise (100 to 120 dB); hearing protection

required for personnel

 Vibration, which affects operator’s hand,

arm, and body

   

Required maintenance of air tool At times, insufficient energy to clean casting Operator required to hold air tool Creation of silica dust, which affects operator  Insufficient speed for high production The vertical core knockout machine (Fig. 2) involves the use of a specially made air tool that strokes down to clamp the casting for holding and simultaneously vibrates the casting to break down the core sand. When the operator sees that the internal passages are cleared, the foot switch is depressed to stop the vibration, and the air tool is retracted from the casting. These tools are commonly referred to as C-frame models because the supporting structure resembles the letter “C.” This cleaning method uses the vibration-under-compression theory, which is generally applicable for all castings, yet is somewhat slow. The advantages of vertical core knockout tools include:  Low cost  Flexibility (can be effectively used on cast-

ings from 2 to 15 kg, or 5 to 35 lb)

 No adverse effect on operator’s hands, arms,

or body

 Several machines can be run by one operator  Applicability to low-volume production  Relatively small amount of floor space

Fig. 2

Vertical core knockout machine

Fig. 3

A-frame core knockout machine

required  Portability  Easily adjustable frequency and amplitude The disadvantages are:      

Noise (100 to 120 dB) Required maintenance of air tool parts Lower effectiveness on larger castings Insufficient speed for high production At times, inappropriate for a cleaning method Creation of silica dust

The A-frame core knockout machine involves the use of a cast iron clamp/vibrating head and cast iron clamping head mounted on a common shaft and A-frame leg stands and controls (Fig. 3). The cast iron heads are supported by the common mounting shaft and face one another. The left side contains a clamping and vibrating tool, and the right side contains a bumper or an air cushion device to soften the vibration and energy transmission. To operate the A-frame core knockout machine, an operator positions the casting between the clamp heads and depresses a foot switch to simultaneously hold and vibrate the casting to break down the core sand. This method allows the operator to rotate and probe the casting to reduce cleaning cycle time. When the core sand has been removed, the operator depresses the foot switch to stop vibration and unclamp the casting. The A-frame machines use vibration under compression, which is generally suitable for all types of castings. On this machine, both

the clamping pressure and the vibrating pressure are independently adjustable, so the vibration can be adapted to suit the casting. The advantages of A-frame machines include:  Moderate price  Flexibility (can be used on both large and     

small castings, that is, from 2 to 45 kg, or 5 to 100 lb) No harmful effect on operator’s body Several machines can be run by one operator Applicability to low- and high-volume production Rugged construction to withstand abusive foundry practices Approximately 2.5 m2 (25 ft2) of floor space required

Processing and Finishing of Castings / 515 The disadvantages are:  Noise (100 to 120 dB)  Multiple-unit facilities generate sound waves      

that can have harmful effects on internal organs. In medium- to high-production plants, excessive maintenance is required. High replacement part inventory is necessary. Castings must be transported to and from special acoustical rooms. Insufficient speed for high production Ineffectiveness on thin-walled heavily cored castings Creation of silica dust

The multiple-station air tool fixture involves the use of air tools mounted on a supporting structure with holding and locating blocks to position castings. These fixtures are dedicated to the core knockout of a specific casting. These devices either operate by means of vibration under compression, whereby the casting is clamped and vibrated, or they have a 6 to 10 mm (¼ to 3/8 in.) gap surrounding the casting, which undergoes vibration. As the casting is impacted, it moves violently within this tight space. This allows more energy to penetrate the casting from many different directions. The major problem with this method is that loading and unloading of the casting can be difficult because of the varying sizes of risers. The retractable tools employing vibration under compression allow easy loading/unloading and maintenance, yet can reduce cleaning effectiveness. With dedicated fixtures, multiple air tools can be used on a single casting, thus inducing greater amounts of energy to those areas that require it most. Multiple-station fixtures are normally built to clean two, three, or four castings simultaneously. These core knockout fixtures must be isolated from their support structures by vibration-proof mounts. Dedicated fixtures are most commonly used on high-volume-production applications, although certain low-to-medium production run castings may require dedicated tooling because of the complexity of the casting. These units generate extremely high noise levels; therefore, they are often housed in acoustical-control cabinets that are designed to be portable with interchangeable fixtures (Fig. 4). The advantages of acoustically enclosed, dedicated fixture units include:  Low noise levels  High production       

and lower per-casting cleaning cost Easier cleaning Timed cycle Interchangeable fixtures Sand containment Improved safety Portability Location possible near production line

The disadvantages are as follows:  High initial capital expenditure

Fig. 4

Acoustical-control cabinets. (a) With three-station dedicated fixture. (b) With A-frame tooling

 Acoustical enclosures are a nonincome-

producing expense  Maintenance required for air tool components  Few available suppliers  Fixturing dedicated to specific castings (not universal) A high-frequency drive machine involves the use of two motors spinning weights in opposite directions to develop a rapid push-pull movement of a load table. The casting is hydraulically retained on the mounting table. The rapid back-and-forth movement causes a floating piston to move and contact a strike plate that transmits energy to the casting. This high-frequency drive device is said to develop between 5 and 32 g (50 and 315 m s 2, or 160 and 1030 ft s 2) of force. The motors operate at between 2900 and 3600 rpm. The optimal amplitude or stroke is between 5 and 10 mm (3/16 and 3/8 in.). These units have proved to be effective in cleaning some of the most difficult castings. However, they are very loud, and the hydraulic clamping devices have failed in production settings. The advantages of high-frequency drive machines include:




   

Speed Adjustability Interchangeable fixtures Timed cycle




The disadvantages are:    

High cost Self-destruction Dedicated construction Noise (110 to 130 dB)

The high-pressure water blast device incorporates the use of a gimbal-mounted spray gun positioned in a stainless steel enclosure with a high-pressure pumping system. The successful operating pressures lie within a 17 to 105 MPa (2.5 to 15 ksi) range. A casting is located on a rotating, tiltable table, and the water is directed at the core material through the spray gun. The operator moves the casting to blast the water at the cored passageways. This system is effective in cleaning ceramic material from investment castings and complex sand-cored castings, and is also effective for removing burnt-in core sand from internal passages. The advantages of a high-pressure water blast system include:     

Adjustable water pressure Water can be focused directly at core Safety Quiet operation Dedicated fixturing can be used The disadvantages are:

 High cost  Contaminated water disposal

516 / Principles and Practices of Shape Casting     

Low operating speed Maintenance required for nozzle and pump Unreliable water recirculation Expensive spare parts Possible inability to direct water at blind or oblique passageways

Shotblast Cleaning. In the shotblast cleaning method, castings are hung on a tree and placed in an enclosure. Shot is spun off a wheel at the casting. This method is not very effective for cleaning aluminum castings because the energy is not sufficient to break down the cores, and the cycle time is longer than with other methods. Also, mixing sand with spent shot can cause excessive wear to the reclamation system. Nevertheless, some foundrymen do use this process because of the availability of a shotblast machine and/or because the casting requires shotblasting. Core Bake-Out Method. With the core bake-out method, castings are loaded into baskets and placed into the ovens. The castings are heated to the temperature that will completely burn out the resin binders without affecting the metallurgical structure of the casting. Eventually, the sand becomes loose enough to flow out of the castings when moved. This method is normally used when all other processes fail, but it is expensive, time consuming, and requires heat energy. The castings still hold the loose sand. Either the loose sand that does fall out has to be cleaned from the ovens or a sand reclamation unit must be added to the system.

machine or it can hold the grinding tools and apply them to the casting. The advantages of using a robot to manipulate a casting include:  The robot can pick up the casting from a

 





fixed point that can be magazine fed or from several predetermined positions if the castings are palletized. The robot can pick up castings within the limits of its capacity that would, in many cases, be too heavy for a manual operation. The robot can move the casting to a succession of different tools to perform the cleaning operation to the required standard in the minimum amount of time by using the most efficient tool available at each operation. These sequential operations can be carried out without having to stop the cycle to change tools. The robot can put the casting down on a fixture and regrip it to allow access to other surfaces of the casting. For example, the robot can turn the casting around or completely over. Upon completion of cleaning, the robot can place the casting in a fixture, drop it in a

bin, or place it on any predetermined position on a pallet. When the casting weight is high or when a variety of grinding wheels are necessary, it is more feasible to allow the robot to handle the tool rather than the casting. Industrial manipulators, although not programmable, are useful for this application. Tool change downtime is obviously very important with such a system. A number of problems are inherent in any system, regardless of the type used. These problems include the following:  Grinding develops varying load conditions,

which cause inaccuracies in robot positioning. This can be overcome by either an adaptive control system or by a template-driven system, such as that shown in Fig. 6.  Imprecision in casting geometries creates inaccuracies in location.  Corrections must be made for wheel wear; simple light beam sensors are often used to sense the location of the edge of the wheel.

Cleaning Operations Cleaning operations can be divided into two categories: fixed and variable. Fixed grinding is the removal of the material that is present on every casting in a fixed position (that is, feeder pads, flash, gates, and so on). For highvolume production, automatic fixed-stop machines have been specifically designed to grind castings. In contrast, variable cleaning occurs anywhere on the casting and usually results from defects in the mold or cores. Robotic technology is not sufficiently developed for cost-effective automation of variable cleaning operations. Programmable cleaning operations are normally conducted by robots. Figure 5 shows a typical robotic installation. This cell consists of a robot and four specially designed grinding wheels. The wheels are hydraulically driven from a central unit. Wheel wear compensation is automatic such that the working point of the wheel is at the exact same location from cycle to cycle. Before entering the cell, castings are introduced to a dedicated press, where the main riser is removed and specific contour locations are stamped that will serve as reference areas for positioning. The casting is picked up by the robot and moved to each grinding machine. Depending on casting size, the robot can either hold the casting and move it to the grinding

Fig. 5

Schematic of a first-generation robotic system designed to perform fettling operations

Fig. 6

Guidance template used to control the movement of the robotic arm

Processing and Finishing of Castings / 517

Automating Gate Removal and Grinding ((Ref 1) As noted, the improved ability to easily program robots in an offline environment improves automation capabilities for highly labor-intensive operations. Automating gate removal and grinding (Fig. 7) is now very feasible for short-run casting producers. Cells can be easily reconfigured for quick and easy changeover between jobs. General Factors in Automated or Manual Gate Removal. To optimize the gate removal process, there are several things that should be considered. Many of the following items apply to manual and automated gate removal. Consistent gating includes the following considerations:  The more consistent the runner and gating

system, the easier it is to remove the castings from the runner system. Providing gating that has minimal warpage and is flat across the gating attachment area will make manual removal easier but will make automation even more cost-effective.  Limiting the number of gating schemes will reduce the number of fixtures, grippers, and programs that are needed to process parts. In many cases, the program, gripping points, and infeed features can be the same for a variety of parts.  Design gating so that it is easy to automate. With a little forethought, gating can be designed so that it will allow the robot

Fig. 7

gripper to consistently locate the tree at the pickup location. By adding features to the gating that allow the robot to use the same pick points, the same robot program can be used for a large cross section of parts. This eliminates having to teach many programs. Ease of Programming Robotic programming for gate removal is normally a fairly simple task. Some things to consider when automating gate removal are:  Develop a standard set of base programs that

are easily modified for different parts.

 If done correctly, then new programs only

require changing of variables. This greatly reduces the time to generate a program and will keep the cell in production. Flexibility of System to Accommodate Multiple Parts The next area to consider is the ability of the cell to accommodate a large variety of parts. It is also very important that changeover time between parts can be accomplished very quickly. In the ideal situation, the changeover can be done during the cycle time of the robot. There are also several other factors that a properly configured cell will contain to provide flexibility:  The use of vision and/or laser measurement

systems provides an easy method to find and determine the orientation of the runner system. This information is then uploaded to the robot, which reorients the gripper to pick up the part. The use of these systems

Gate removal. Courtesy of Vulcan Engineering Company

eliminates or greatly simplifies fixturing and grippers.  Human/machine interface needs to be userfriendly to allow the operator to easily change over between parts without ever having to use the teach pendant. It should contain all the information that is required to allow the monitoring of alarms, wheel wear, and job information. All programs that are run on the cell should be stored on this computer, with quick upload of the correct programs to the cell.  If locating fixtures or different grippers are required, they should be quick-change. Depending of the length of the job runs, the decision to use manual quick-change or fully automatic tool change must be considered.  For parts that use multiple fixtures or grippers, work objects must be the same so that only one program must be taught. Proper Peripheral Equipment The cell is only as good as the peripheral equipment around the robot. When looking at the overall scope of work to be performed, the first decision that must be made is to determine if the spindle is mounted on the robot or if the part/runner system is picked up by the robot and taken to the saw. Once this is determined, the following items are very important for either configuration:  The cut-off saw must have programmable

force control. This is the most important factor to properly cut the gating system. The saw must have the force control on it because this overcomes the inability of the robot to change velocity dynamically at the rate required by the cutoff operation. Robots are very good at following a path and moving very accurately to a point, but due to scan times from an input device, they cannot react quickly enough to maintain the proper cutting pressure.  Because the abrasive wheels naturally wear down, a measurement system must be used to offset robot programs to compensate for this wear.  The system must be able to track wear to determine the life of the wheel and notify the operator when it is time to change a wheel. Force control is important to properly cut the castings from the runner system. Figure 8(a) shows what happens when the cutoff saw does not have force control. As the arc of contact of the wheel moves into the riser and the velocity or feed rate is too high, the amount of force exerted on the wheel causes an increase in temperature. This increase in temperature does two things: It will cause the wheel to break down at a rapid rate that results in poor wheel usage, and the heat generated is transferred to the casting. If the casting is a heat-sensitive alloy, this can cause heat cracking. To solve the problems created in Fig. 8(a), the programmer attempts to slow down the feed rate.

518 / Principles and Practices of Shape Casting This may seem like a good idea, but the same thing happens as mentioned previously. Because the feed rate is too slow, the wheel does not break down properly to keep the abrasive sharp. The wheel becomes loaded with material, which causes an increase in friction. As this happens, the temperature of the wheel increases, which heats up the resin that holds the abrasive together and allows it to break down prematurely. The friction also causes the part to become hot (Fig. 8b) and again can cause heat cracking in certain alloys. Figure 8(c) shows what happens in the cut when the machine is equipped with force control. The correct cutting force is constantly applied against the part, and the spindle actually moves back and forth as the arc of contact changes along the circumference of the wheel. Because the system is able to cut with the correct force, the results are a cool cut that optimizes wheel life and provides for quicker gate removal. Gate Grinding. After parts have been removed from the runner system, there is remaining gating material that must be removed to complete the process. There are several factors that must be considered to properly grind gates:  Determine the best method to grind: Does the

robot take the casting to the grinding equipment, or is a spindle mounted on the robot, and the grinding medium is taken to the casting?  Determine grinding tolerances that can be achieved. This can depend on a number of

things: part size, gripper or fixture location in relation to the area to be ground, and variation in the casting versus the final tolerance to be achieved.  If the tolerance is very tight or if the casting variation is greater than the final tolerance, what type of measurement devices are required to complete the task?

Choosing the correct method of introducing the part into the cell is very important. Several options are:  Directly mounting the part on the robot. This



Different methods of programming the cell may be considered, depending on the application. The different types of programming methods are:  The teach pendant is used to teach simple

parts that consist of curves and straight grinds.

 Offline programming using solid models

allows for quick generation of multiple paths that have many points along a profile. This method is less disruptive to production because the program is written in the office and then downloaded to the robot. This method is a requirement for parts with complex geometries.  The use of a digitizer is another way to program offline. It is typically used where no solid model of the part exists. A replica of the same diameter of the grinding wheel is mounted on the end of the digitizer. As the programmer moves the digitizer around the part, he uses a foot switch to teach the points along the grinding path. Again, once this program is taught, it is downloaded directly into the robot.







method requires the operator to unload and load the robot after every cycle, which can impact throughput. Fixture tables with part fixture plates are effective for larger-sized parts. Two to four stations can normally be arranged in front of the robot. The robot picks up the fixture plate and takes it to the grinding equipment. While the robot is working, the operator can unload and load the next part on the remaining stations. This method helps maximize the robot grinding time in the cell. Because multiple input stations are used, this frees the operator to do other tasks while the robot works on processing the casting on the fixture tables. Sliding draws are typically used for smaller parts, where several are placed on a fixture and processed one after the other. Usually, there are two draws located next to each other so that the operator can unload or load while the robot is working. A turntable with a work station on each side allows the operator to load and unload a part while the robot is working. These devices are usually configured with quick-change fixturing that allows the operator to quickly change between jobs. Vision systems allow casting to be sent into the cell on belt conveyors; the vision system then tells the robot the orientation of the part. The robot picks up the part and takes it to the grinding machines. This method eliminates the need for costly fixturing. Vision systems, along with quick-change grippers, provide a quick changeover between jobs, which makes a cell very productive.

A large area of cost savings can come from the combination of several functions in a cell. Figure 9 shows a combination cell. This cell combines both the cut-off and the grinding operations in one cell. One advantage of this type of cell is that it reduces the amount of work that would be required for a cut-off cell and a grinding cell in a process. It also makes quick work of a casting that must be cut off from the gating

Fig. 8

Force control in cutting castings from the runner system. (a) Effect when the cutoff saw does not have force control. (b) Overheating when feed rate is too slow. (c) Proper force control of cutting. Source:Ref 1

Fig. 9

Combination cell. Courtesy Engineering Company

of

Vulcan

Processing and Finishing of Castings / 519 system and then requires additional gate removal and grinding to final tolerance.

Flame Cutting For heavy-section steel castings, where risers cannot be removed economically by grinding, impact, and so on, robotic flame cutting is used for riser and gate removal. Unlike cleaning, however, the torch must be held by the robot. As in cleaning, such systems range from semiautomatic to fully programmable robotized systems. Installations can use both conventional and gantry-style robots. Steel thicknesses in excess of 610 mm (24 in.) have been successfully cut. To perform the flame cutting operation, the operator, using the control panel, moves the torch into the general position of the riser. The unit is preprogrammed for six riser contact sizes. The operator then selects the proper size, and the cut is executed. More automated systems use a six-axis robot for complete automation of the cutting operation. Specially designed locating blocks or pads are cast into each casting. Contact location principles are used to provide alignment between the robot and the casting. The robot is programmed using a teach pendant device. In addition to position and path, the operator programming the robot must also define the travel speed between each point, the preheat delay, and the cutting oxygen pressure for each cut. Results indicate a savings of 120 to 180%, depending on the casting to be cut. The following factors must be taken into account when considering an installation or application:  Fixturing: Distortion of gate position due to

rough handling, mold inaccuracy, and so on, leads to errors in position. Some means of sensing the positions of the gates and risers is necessary.  Feedback system: Because of variations in thickness and other defects, such as flash and drops on the casting, an adaptive control system that accounts for variations would be very beneficial.  Automatic torch ignition: This provides an added measure of safety and convenience.  Preheat sensing: One system is equipped with a fiber optic lens that sights through the oxygen orifice. An infrared sensing measurement is made to the computer controlling the robot. If the signal indicates kindling temperature, oxygen is employed, and the cut is continued.

Blast Cleaning of Castings Blast cleaning of castings is a process in which abrasive particles, usually steel shot or grit, are propelled at high velocity to impact the casting surface and thereby forcefully remove surface contaminants. The contaminants are usually adhering mold sand, burned-in sand, heat treat scale, and so on. For aluminum

castings, the process is often used to provide a uniform cosmetic finish in addition to merely cleaning the workpiece. This is especially true of engine components, such as heads and manifolds, that are highly visible. In the case of cast aluminum wheels, die cast transmission cases, and so on, the process prevents leakage by healing surface porosity. Additionally, in some cases, deburring and a pleasing cosmetic surface are obtained. Methods for the handling of castings during blast cleaning depend on production volume, size, shape, and mix of the parts and the ability of the parts to be tumbled. Robots and similar manipulative devices are also used in blastcleaning operations. The usual methods of imparting high velocity to abrasive particles are by the use of either centrifugal wheels or compressed air nozzles. Centrifugal wheels are the most widely used method because of their ability to efficiently propel large volumes of abrasive. For example, a 56 kW (75 hp) centrifugal wheel can accelerate steel shot to 73 m/s (240 ft/s) at 55.8 Mg/ h (123,000 lb/h) flow. To do the same with 13 mm (½ in.) direct-pressure venturi nozzles at 45 kg/min (100 lb/min) per nozzle would require approximately 20 nozzles and a total air flow of 2.45 m3/s (5200 ft3/min) at 550 kPa (80 psi). However, even though the nozzle blast is not as efficient overall as the wheel blast, in some applications it may be more efficient because the blast stream can be more efficiently applied when blasting into small holes to clean the interior areas of a casting. Other reasons for using nozzle blast include portability and suitability for very hard abrasives, such as aluminum oxide. The machine type most often used is a batch tumbling barrel when the production requirement is low to moderate for parts that can be tumbled. Continuous tumbling is used for high-volume production. When castings cannot be tumbled, the machine type most often used is the rotating-hanger-type machine (Fig. 10). The hanger and hook are moved horizontally to a position opposite the centrifugal wheels, where the castings are blasted while rotating about a vertical axis. These machines come in a number of configurations. When higher production rates are required, a turnstile-type hanger machine is sometimes used. Hangers may also be powered or controlled in various ways. When parts are too large to either tumble or hang on a hook, or when production requirements are low, the machine types most often used are either a table-type machine or a room-type machine. Small table machines are used to handle smaller castings that cannot be tumbled. They are also used when size, shape, or low production requirements rule out a hanger machine. The part or parts are placed on the table by a power lift device, and, after the doors are closed, the parts are blasted by one to three centrifugal wheels. A disadvantage of this type of machine is that usually the parts

must be turned over or repositioned, and the blast cycle must be repeated to effect complete cleaning. Room-type machines are used when the requirement is low production on a variety of medium-sized to very large parts. These machines usually consist of a room with power-operated entrance doors. A self-powered car with a rotating table carries the work to be blasted into the room to the blast position. Mounted on the room are two or more centrifugal wheels. While the rotating part is being blasted, either the car moves to three blast positions, or, in some cases, the wheels are oscillated for better blast coverage. As with a table machine, the part or parts may need to be turned over or repositioned for complete coverage. Cleaning Internal Surfaces. Many times when cleaning castings it is important to clean internal surfaces such as the water passages in engine heads and blocks or other surfaces that cannot be impacted directly because of the casting configuration. In these cases, ricocheting abrasive must do the cleaning. To clean internal surfaces, it is usually important to drain the internal passages while blasting so that pocketed abrasive does not cushion the impact. It is also important to have sufficient ricochet velocity to remove the surface contaminants effectively. In the majority of cases, the best way to drain internal passages is to rotate the part on a horizontal axis, as is done on axial-flow machines or other machines that incorporate horizontal rotating devices. Recent laboratory tests and field experience have shown that increasing the wheel speeds improves internal cleaning significantly. At times, the blast from a centrifugal wheel does not effectively reach into all of the internal cavities of a casting. Often, the solution to this situation is to resort to a “spot blast,” that is, use of a compressed-air nozzle to direct the blast into selected casting openings. Here again, high abrasive velocity can be effective. Changing from straight nozzles to long venturi nozzles can result in abrasive velocity increases

Fig. 10

Front view of a typical rotating assembly. The workpiece is rotated it is held in position on the monorail track to full coverage by the blast pattern produced centrifugal wheels.

hanger 360 as provide by the

520 / Principles and Practices of Shape Casting of 30% or more, or, if abrasive velocity is more important than coverage, the nozzle flow can be decreased and the velocity increased. Blast equipment manufacturers can advise nozzle users on velocity increases that are possible with different types of nozzle configurations and flow rate changes.

Table 1

Casting defects, descriptions, and prevention

Casting imperfections or defects

Cold shuts

Prevention

Pour as quickly as possible Design gating system to fill entire mold quickly without an interruption

Possible causes:

Preheat the mold Modify part design Avoid excessively long, thin sections

 Interruption in the pouring operation  Too slow a pouring rate  Improperly designed gating

Inspection After finishing, castings are inspected for surface quality. Inspection can be performed manually by visual checking, manually by template comparison, or by an automated inspection station. Visual inspection is simple yet informative. A visual inspection would include significant dimensional measurements as well as general appearance. Surface discontinuities often indicate the presence of internal discontinuities. Computer-assisted coordinate-measuring machines measure both pattern and casting dimensions. These machines can perform a variety of dimensional checks ranging from basic geometric measurement to parallel and plane projection. The operator simply identifies critical part locations so that the machine can establish a working plane. The coordinate-measuring machine can perform in a few minutes the tedious checks that take 2 to 4 h to be done manually. For small castings produced in reasonable volume, a destructive metallographic inspection on randomly selected samples is practical and economical. This is especially true on a new casting for which foundry practice has not been optimized and a satisfactory repeatability level has not been achieved. For castings of some of the harder and stronger alloys, a hardness test is a good means of estimating the level of mechanical properties. Hardness tests are of less value for the softer tin bronze alloys because hardness tests do not reflect casting soundness and integrity. Inspection also includes various methods of nondestructive testing (NDT) to screen castings for imperfections that may be considered to be defects. The techniques for NDT of casting are briefly summarized as follows, with further information given in the article “Nondestructive Testing of Components” in this Volume. The general types of imperfections or defects in castings are listed in Table 1. As a general rule, the method of inspection applied to some of the first castings made from a new pattern should include all those methods that provide a basis for judgment of the acceptability of the casting for the intended application. Any deficiencies or defects should be reviewed and the degree of perfection defined. This procedure can be repeated on successive production runs until repeatability has been ensured. Liquid penetrant inspection is extensively used as a visual aid for detecting surface flaws. One of the most useful applications, however, is the inspection of alloys susceptible to hot

Description

Appear as folds in the metal — occurs when two streams of cold molten metal meet and do not completely weld

Hot tears and cracks

Hot tears are cracklike defects that occur during solidification due to overstressing of the solidifying metal as thermal gradients develop. Cracks occur during the cooldown of the casting after solidification is complete due to uneven contraction.

Fill mold as quickly as possible Change gating system; e.g., use several smaller gates in place of one large gate Apply thermal management techniques within the mold (e.g., chills or exothermic material) to control solidification direction and rate Insulate the mold to reduce its cooling rate Modify casting design:

 Avoid sharp transitions between thin and thick sections

 Taper thin sections to facilitate establishment of appropriate solidification gradients

 Strengthen the weak section with additional material, ribs, etc. Inclusions

Presence of foreign material in the microstructure of the casting

Modify gating system to include a strainer core to filter out slag

Typical sources:

Avoid metal flow turbulence in the gating system that could cause erosion of the mold Improve hardness of the mold and core

 Furnace slag  Mold and core material Misruns

Incomplete filling of the mold cavity Causes:

   

Too low a pouring temperature Too slow a pouring rate Too low a mold temperature High backpressure from gases combined with low mold permeability  Inadequate gating Porosity

Holes in the cast material Causes:

Control mold and metal temperature Increase the pouring rate Increase the pouring pressure Modify gating system to direct metal to thinner and difficult-to-feed sections more quickly

Pour metal at lowest possible temperature Design gating system for rapid but uniform filling of the mold, providing an escape path for any gas that is generated Select a mold material with higher gas permeability

 Dissolved or entrained gases in the liquid metal

 Gas generation resulting from a reaction between molten metal and the mold material Microshrinkage Liquid metal does not fill all the dendrific interstices, causing the appearance of solidification microshrinkage

Control direction of solidification:

 Design gating system to fill mold cavity so that solidification begins at the extremities and progresses toward the feed gate  Lower the mold temperature and increase the pouring temperature  Add risers, use exothermic toppings to maintain temperature longer  Control cooling rate using chills, insulators, etc. in selected portions of the mold

Source:Ref 2

cracking. Such cracks are virtually undetectable by unaided visual inspection but are readily detectable by liquid penetrant inspection. Pressure testing is used for castings that must be leaktight. Cored-out passages and internal cavities are first sealed off with special fixtures having air inlets. These inlets are used to build up the air pressure on the inside of

the casting. The entire casting is then immersed in a tank of water, or it is covered by a soap solution. Bubbles will mark any point of air leakage. Radiographic inspection is a very effective means of detecting such conditions as cold shuts, internal shrinkage, porosity, core shifts, and inclusions in aluminum alloy castings.

Processing and Finishing of Castings / 521 Radiography can also be used to measure the thickness of specific sections. Aluminum alloy castings are ideally suited to examination by radiography because of their relatively low density; a given thickness of aluminum alloy can be penetrated with approximately one-third the power required for penetrating the same thickness of steel. The typical complexity of shape and varying section thicknesses of the castings may require digital radiography or computed tomography. Magnetic particle inspection can be applied to ferrous metals with excellent sensitivity, although a crack in a ferrous casting can often be seen by visual inspection. Magnetic particle inspection provides good crack delineation, but the method should not be used to detect other defects. Nonrelevant magnetic particle indications occasionally occur on ferrous castings, especially with a strong magnetic field. For example, a properly fused-in steel

chaplet can be indicated as a defect because of the difference in magnetic response between low-carbon steel and cast iron. Even the graphite in cast iron, which is nonmagnetic, can cause a nonrelevant indication. Ultrasonic inspection for both thickness and defects is practical with most ferrous castings except for the high-carbon gray iron castings, which have a high damping capacity and absorb much of the input energy. The measurement of resonant frequency is a good method of inspecting some ductile iron castings for soundness and graphite shape. Electromagnetic testing can be used to distinguish metallurgical differences between castings. Aluminum alloy castings are sometimes inspected by ultrasonic methods to evaluate internal soundness or wall thickness. The principal uses of ultrasonic inspection for aluminum alloy castings include the detection of porosity in castings and internal cracks in ingots.

ACKNOWLEDGMENTS Adapted from Ref 1 and the following:  G.J. Maurer, Jr., Shakeout and Core Knock-

out, Casting, Vol 15, ASM Handbook, ASM International, 1988, p 502–506  J.H. Carpenter, Blast Cleaning of Castings, Casting, Vol 15, ASM Handbook, ASM International, 1988, p 506–520  R.L. Lewis and Y.L. Chu, Foundry Automation, Casting, Vol 15, ASM Handbook, ASM International, 1988, p 566–573

REFERENCES 1. C. Cooper, “Automating Gate Removal and Grinding,” Vulcan Engineering Company 2. Mechanical Engineering Handbook, CRC, 1999

ASM Handbook, Volume 15: Casting ASM Handbook Committee, p 525-527 DOI: 10.1361/asmhba0005241

Copyright © 2008 ASM International® All rights reserved. www.asminternational.org

Introduction—Expendable Mold Processes with Permanent Patterns Brian V. Smith, General Motors Corporation

CASTING can be done with either expendable molds for one-time use or permanent molds for reuse many times. Both expendable and permanent molds must be separable into two or more parts in order to permit withdrawal of either the permanent pattern (from an expendable mold) or the raw casting (in the case of a permanent mold or die). In the case of sand (or other loose granular material) molding, permanent patterns are removed after the sand is bonded in a flask in cope-and-drag sections (Fig. 1). Cores are separate shapes that are placed in the mold to provide castings with contours, cavities, and passages that are not practical or obtainable with molds. Patterns may be permanent (as is typical in sand casting) or expendable (as in lost foam and investment casting). With expendable patterns, the limitation of two or more separable parts of the mold is not necessary. The molding medium must only surround the expendable pattern and maintain its shape during molten-metal pouring and solidification; after the shape has solidified, the sand or other molding medium is shaken off and out of the part. Casting of parts using expendable mold processes with either permanent or expendable patterns is a very versatile molding method that provides tremendous freedom of design in terms of size, shape, and product quality.

article, “Aggregates and Binders for Expendable Molds,” in this Volume). Manufactured ceramic molding media with specific properties are also becoming more popular in various casting processes. The choice of molding method depends on several factors, such as part size and shape, quantity, tooling, and the molten metal being poured into the mold. Table 1 lists a general comparison of casting methods. The characteristics of foundry molds must address four basic requirements:  Be formable into the desired shape around

some type of pattern

 Be able to hold that shape while the molten

metal is introduced into it

 Be able to maintain that shape while the

molten metal solidifies

 Be able to break down and become strippa-

ble after the metal solidifies

The first requirement is that the mold material must flow or be pliable enough to encapsulate the surface around the shape of the pattern. Patterns (Fig. 2) can be made from a permanent-type material (i.e., wood, steel, hard plastic), so that numerous molds can be duplicated before significant pattern wear would alter the shape of the part being cast. Once the mold material is tightly bonded around the pattern, the mold must be stripped and removed from that pattern, while not damaging either the permanent pattern material or the established mold shape. The use of tapered pattern ends permits it to be removed from the sand mold without restriction (Fig. 3). The mold shape, which is typically the reverse or mirror image of the part being cast, must be strong enough to be handled and manipulated, so that it can be combined with other mold shapes (copes, drags, cores, gatings, etc.) that

Expendable mold processes are widely used in casting, and various methods are used to fabricate expendable molds from permanent patterns. The methods include:  Molding of sand with clay, inorganic binders,

or organic resins

 Shell molding of sand with a thin resin    

bonded shell that is baked No-bond vacuum molding of sand Plaster-mold casting Ceramic-mold casting Rammed graphite molding Magnetic (no-bond) molding of ferrous shot

Sand is the most prevalent molding medium, and various types of binders are used to bond the sand into useable molds (see the next

Fig. 1

Cope-and-drag mold halves with cores in place, ready for closing. Source: Ref 1

526 / Expendable Mold Casting Processes with Permanent Patterns Table 1 General comparison of casting processes Weight, kg (lb) Process

Typical materials cast

Sand Shell Expendable pattern Plaster mold Investment Permanent mold Die Centrifugal

All All All Nonferrous (Al, Mg, Zn, Cu) All (high melting point) All Nonferrous (Al, Mg, Zn, Cu) All

Minimum

0.05 0.05 0.05 0.05

(0.11) (0.11) (0.11) (0.11)

0.005 (0.011) 0.5 (1.1) 1470  C, or 2678  F), tridymite transforms to cristobalite. Linear thermal expansion of quartz increases up to 17  106/K (perpendicular to c-axis) and 10  106/K (parallel to c-axis) at 840 K with increasing temperature. Above that temperature, it has a constant coefficient of thermal expansion up to 1000  C (1830  F). Zircon is a naturally occurring zirconium silicate (ZrO2 SiO2) mineral. It has good volume stability for extended periods of exposure to high temperatures. Its primary advantages are a very low thermal expansion, high thermal conductivity and bulk density (which gives it a chilling rate approximately four times that of quartz), and very low reactivity with molten metal. Zircon requires less binder than other sands because its grains are rounded. It is also compatible with all known binders—both organic and inorganic. The very high dimensional and thermal stabilities exhibited by zircon are the reasons it is widely used in steel



foundries and investment foundries making high-temperature alloy components. The high chilling rate is useful for controlling directional solidification by using it locally for selective chilling. Olivine is a naturally occurring magnesium silicate. Olivine minerals (so called because of their characteristic green color) are a solid solution of forsterite (Mg2SiO4) and fayalite (Fe2SiO4). Forsterite and fayalite in approximately equal proportions are the primary constituents of rock called dunite found in Washington state and North Carolina (Ref 4). Other locations of olivine sand include Scandanavia, Africa, and New Zealand (Fig. 5). The physical properties of olivine vary with their solid-solution composition of forsterite (Mg2SiO4) and fayalite (Fig. 6). Crystal structure of olivine is much more complex than other specialty sands. Inclusions of magnetite (Fe3O4), spinel (magnesium, iron, zinc, manganese, aluminates), apatite (Ca5(PO4 CO3)3), liquids, or gases are common. Movable bubbles are found at times. Olivine alters readily; in fact, it is much more commonly found altered than fresh. The most common alteration products are hydrous magnesium silicates (serpentine). The alteration develops in scales and fibers, and finally the entire olivine crystal can be transformed. Therefore, the composition must be specified to control the reproducibility of the sand mixture. Care must be taken to calcine the olivine sand before use to decompose the serpentine content, which contains water (Ref 5). Likewise, the refractory characteristics of olivine



are damaged if the system is allowed to be contaminated by the accumulation of silica from core, for example. The specific heat of olivine is similar to that of silica, but its thermal expansion is far less. Therefore, olivine is used for steel casting to control mold dimensions. Olivine is somewhat less durable than silica, and it is an angular sand. Unlike silica, olivine is chemically basic. It can be bonded with clays and most organic and inorganic binders, but it cannot be used with acid-curing binders. Chromite (FeCr2O4) is a black, angular sand. It is highly refractory and chemically unreactive, and it has good thermal stability and excellent chilling properties. However, it has twice the thermal expansion of zircon sand, and it often contains hydrous impurities that cause pinholing and gas defects in castings. It is necessary to specify the calcium oxide (CaO) and silicon dioxide (SiO2) limits in chromite sand to avoid sintering reactions and reactions with molten metal that cause burn-in (Ref 6). Aluminum Silicates include kyanite, sillimanite, and andalusite, and all have the same chemical composition (Al2SiO5). Kyanite and sillimanite have long been mined and used either raw or calcined to manufacture refractory mortars, cements, castibles, and plastic ramming mixes. Crushed kyanite and sillimanite have been tested and found unacceptable for foundry sand use because of their cleavage and irregular shapes produced by crushing. Minerals of the aluminum silicate group are similar in mineralogy, as shown in Table 4.

L

L L

L

M L

L

L

L

L

H

M

L

M

L

H H

M M

H

H

H

M M

M

H

M

M

H

M

M

H

M

M

M

H H

H H

H

M M M

H H H

Rate of gas evolution (a)

L

L

L

L

L

L

H

L

L ...

L

L

L

L

L

L

L

M

L L

L L L

Degree of thermal plasticity(a)

L

M

L

L

L

L

L

M

M ...

H

H

M

M

M

M

M

H

H M

M M M

Collapsibility speed(a)

P

G

P

P

P

P

P

P

G ...

G

G

F

G

G

G

G

G

G G

F F F

Ease of shakeout (b)

P

P

F

F

P

P

P

F

G G

F

F

G

G

G

G

G

G

F G

G G G

Moisture resistance (b)

30

45 30–60

30–60

... ... ...

...

...

30

...

H

5–60

...

...

H

H H

...

P

...

...

...

...

...

...

F

G G

2–20

...

...

... ...

2–45

...

...

G

2–45

...

...

...

...

1–45

...

H

1–45

...

P

...

...

F F

... ...

1–45

G G G

Resistance to overcure (b)

... ... ...

Strip time(c), min

...

L

H H

H H H

Curing speed (a) C

24

32

24

24

24

24

24

120

24 ...

24

27

32

27

27

27

27

205

230 230

260 260 260



F

75

90

75

75

75

75

75

250

75 ...

75

80

90

80

80

80

80

400

450 450

500 500 500



Optimum temperature

F

F

P

F

F

F

F

P

P ...

P

P

F

P

P

P

P

F

P P

F F F

Clay and fines resistance(b)

F

F

G

F

G

G

G

G

G G

G

G

F

G

G

G

G

F

G G

E E E

Flow ability (b)

H

H

L

L

H

H

M

M

M H

H

H

H

H

H

H

H

H

M M

N N N

Air drying rate(a)

N

N

L

N

N

N

N

M

M ...

M

M

H

M

M

M

M

M

M M

M M M

Pouring smoke (a)

NOTES: (a) H, high; M, medium; L, low; N, none. (b) E, excellent; G, good; F, fair; P, poor. (c) Rapid strip times require special mixing equipment. (d) Use minimum N2 levels for steel. (e) Iron oxide required for steel

Vapor-cured Inorganic Sodium silicate — CO2

Phosphate — oxide cured

Self-setting inorganics Silicates: Sodium silicate — ester cured Sodium silicate — FeSi cured Sodium silicate — 2CaO.SiO2 cured Aluminates: Cement — hydraulic cured Cement (fluid sand) — hydraulic cured Phosphates:

Thermosetting inorganic Silicate (warm box)

Vapor-cured resins (cold box) Phenolic urethane — amine Phenolic/peroxide – SO2 Solvent evaporation — air

Alkyd — organometallic Phenolic — pyridine

Phenolic — acid Urethanes:

High-nitrogen furan — acid Medium-nitrogen furan — acid Low-nitrogen furan — acid Phenolic no-bakes:

Self-setting resins Furan no-bakes:

Organic thermosetting resins Shell processes: Dry blend Warm coat (solvent) Hot coat Hot box processes: Furan Phenolic Oils: Core oil

Relative tensile strength(a)

Table 1 Comparison of mold and binder system characteristics

...

...

...

...

...

...

...

(d)

... ...

(e)

(e)

...

...

...

...

Steel

...

(d) Steel

... ... ...

Metals not recommended

P (sodium silicate)

E-G (phosphates)

G-F (cements)

P (silicates)

E-F (urethanes)

G-F (phenolic nobake)

G-F (furan no-bake)

F (shell)

Ease of Reclaimability(b)

Aggregates and Binders for Expendable Molds / 531

532 / Expendable Mold Casting Processes with Permanent Patterns

Fig. 3

Thermal expansion of silica compared with other mold aggregates

Table 2 Summary of mineralogy and physical properties of foundry sands Property

Quartz

Color

White to yellowishbrown Hexagonal None 7 2.65 1.54–1.56 None

Crystal system Cleavage Mohs hardness Specific gravity Index of refraction Alteration products Melting temperature,  C ( F) Thermal conductivity (K), Btu/hr ft  F Specific heat (C) Btu/lb  F Density (r), lb/ft3 Thermal diffusivity (K/rC) Heat diffusivity (K/rC)1/2





1713 (3115) 0.52 0.28 93 0.02 3.68

Olivine

Chromite

Zircon

Green

Black

White to pale yellowish-brown

Orthorhombic (010) (100) 6.5–7.0 3.2–4.8 1.63–1.69 Hydrous magnesium silicates Variable 1205–1900 (2200–3450) ...

Cubic None 5.5 4.4–5.2 2.00–2.12 Hydrous iron oxide

Tetragonal None 7.5 4.68 1.92–2.02 Metamict zircon

Variable 1760–1980 (3200–3600) 0.63

2200 (3990) 0.57

0.23 150 0.018 4.67

0.198 176 0.0165 4.46

... ... ... ...

All three minerals are thermally stable up to temperature, where they break down to yield 88% mullite and 12% free silica (cristobalite). The mullite is stable to 1810  C (3290  F). The mullite imparts to the products their

desirable properties of high refractoriness, low thermal expansion and resultant resistance to thermal shock, intermediate thermal conductivity, high load-bearing ability even at high temperatures, and resistance to chemical erosion.

Aluminum silicates are blended with zircon sands. Aluminum silicate-zircon blended sands have lower expansion than any of the specialty foundry sands except zircon. The mixture is characterized by relatively coarse, uniform, rounded sand grains with no fines. The grains readily accept binders and can be used with standard coremaking procedures. The mixture has good resistance to shock, high permeability, good chilling action, and refractoriness. Mullite Manufactured from Aluminum Silicates. Another type of mold aggregate is the manufacturing of mullite ceramics from aluminum silicate (Al2SiO5) minerals. Mullite is the only chemically stable intermediate phase at atmospheric pressure in the SiO2-Al2O3 system. It is rare in nature, the most important place of occurrence being the island of Mull, U.K. As a manufactured ceramic, mullite is produced from the high-temperature breakdown of aluminum silicate (Al2SiO5), which occurs in three common forms: kyanite, sillimanite, and andalusite. These forms of aluminum silicate break down at high temperatures to form mullite and silica. Fine mullite particles are produced by calcining aluminum silicates. The grains are highly angular, but additional processing of the fine grains can produce rounded spherical pellets for the base aggregate of foundry molds. Depending on the sintering cycle, the silica may be present as cristobalite or as amorphous silica. The composition may also be varied as either an intermediate-density or low-density ceramic medium (Table 5). These materials have high refractoriness, low thermal expansion, and high resistance to thermal shock. They are widely used in precision investment foundries, often in combination with zircon. Additives. Foundry sands are mixed with sands to improve properties. If done properly, these additives can alleviate the deficiencies of silica, such as its thermal expansion. In addition, silica reacts with molten iron to form a slag-type compound, which can cause burn-in, or the formation of a rough layer of sand and metal that adheres to the casting surface. Additives include carbon additions (such as seacoal), cellulose, and cereal or starches. The additives help control expansion, enhance casting peel and surface finish, reduce moisture sensitivity, and, in some cases, create a reducing atmosphere. Iron molding sands typically use seacoal and seacoal supplements. Steel sands often use cereals and starches rather than seacoal (to prevent carbon pickup). Nonferrous sands usually use cellulose and starches. If low-expansion sands such as chromite, zircon, olivine, or ceramic aggregates are used, carbons are sometimes not necessary. Some nonferrous silica sand systems operate with no carbons and without expansion defects. These organic components of the system sand are normally measured by the percent combustible (loss on ignition) test, which indicates the total amount of carbons. This test is often

Aggregates and Binders for Expendable Molds / 533 Table 3

Density, hardness, and thermal properties of quartzite (silica) compared with other ceramic materials for refractories Maximum service temperature

Melting point Density, g/cm3

Material

Alumina (Al2O3) Graphite (C) Magnesia (MgO) Mullite (3Al2O3 2SiO2) Quartzite (SiO2) Silicon carbide (SiC) Zircon (ZrO2 SiO2) Zirconia (ZrO2)







C





F

C



F

At 100  C (212  F)

At 1000  C (1830  F)

At 1200  C (2200  F)

Coefficient of linear thermal expansion, from 25 to 800  C (77 to 1470  F), 106/ C

Thermal conductivity, W/m

K

Specific heat (mean) at 25 to 1000  C (77 to 1830  F), J/kg  C

Hardness, Mohs scale



3.96 2.2 3.60 2.8

2050 3600(a) 2850 1850

3720 6510(a) 5160 3360

1950 ... 2400 1800

3540 ... 4350 3270

30 180 38 6

6 63 7 4

... ... 2.5 4

8.0 2.2 13.5 5.0

1050 1600 1170 840

9 0.5–1.0 6 7.5–9

2.65 3.2

1400 2200– 2700(b) 2500 2700

2550 3990– 4890(b) 4530 4890

1090 1400– 1700(c) 1870 2400

1990 2550– 3090(c) ... 3400

... ...

... ...

2.1 17

8.6 4.5

1170 840

7 9

... 2.0

... 2.3

4 0.9

4.5 7.5

630 590

4.5–4.7 5.5–5.8

7.5 6.5

(a) Sublimes. (b) Decomposes. (c) Oxidizes

Fig. 4

carbonaceous materials can improve surface finish to castings. Best results are achieved with such materials as seacoal and pitch, which volatilize and deposit a pyrolytic (lustrous) carbon layer on sand at the casting surface (Ref 7). Cellulose is added to control sand expansion and to broaden the allowable water content range. It is usually added in the form of wood flour, ground cereal husks, or nut shells. Cellulose reduces hot compressive strength and provides good collapsibility, thus improving shakeout. At high temperatures, it forms soot (an amorphous form of carbon), which deposits at the mold/metal interface and resists wetting by metal or slags. It also improves the flowability of the sand during molding. Excessive amounts generate smoke and fumes and can cause gas defects. In addition, if present when the clay content drops too low, defects such as cuts, washes, and mold inclusions will occur in the castings. Cereals, Which include corn flour, dextrine, and other starches, are adhesive when wetted and therefore act as a binder. They stiffen the sand and improve its ability to draw deep pockets. However, use of cereals makes shakeout more difficult, and excessive quantities make the sand tough and can cause the sand to form balls in the muller. Because cereals are volatile, they can cause gas defects in castings if used improperly.

Thermal expansion of silica minerals. Source: Ref 3

Sand Properties

Fig. 5

Locations of major deposits for chromite, olivine, and zircon

supplemented with the volatiles test, since this indicates the amount of carbons that come off quickly at low temperatures. Carbonaceous Additions. Carbon can be added in the form of seacoal (finely ground bituminous coal), asphalt, gilsonite (a naturally occurring asphaltite), or proprietary petroleum products. Seacoal changes to coke at high temperatures, expanding three times as it does so;

Fig. 6

Phase diagram of olivine solid-solution series

this action fills voids at the mold/metal interface. Too much carbon in the mold gives smoke, fumes, and gas defects, and the use of asphalt products must be controlled closely because their overuse waterproofs the sand. Carbon is added to the mold to provide a reducing atmosphere and a gas film during pouring that protects against oxidation of the metal and reduces burn-in. The addition of

The composition, size, size distribution, purity, and shape of the sand are important to the success of the moldmaking operation. The selection of sand grain structure dictates the ultimate mold permeability and density, and both of these parameters are critical to the production of quality castings. In general, it is good practice to select sand with a fineness number no greater than necessary to produce good surface finish and prevent metal penetration. Coarser grains reduce the accumulation of fines during recycling. Modern high-density (high-pressure) molding machines also can use coarser sands without compromising surface finish.

534 / Expendable Mold Casting Processes with Permanent Patterns Table 4 Mineral and thermal properties of the aluminum silicate minerals Property

Color Crystal system Cleavage Index of refraction Mohs hardness Specific gravity Mullite inversion,  C ( F)

Table 6

Kranite

Sillimanite

Andalusite

Blue to gray to white Triclinic (100) (010) (001) 1.71–1.73 5–7 3.6–3.7 1200 (2190)

White to crayish-brown Orthorhombic (010) 1.65–1.68 6–7 3.2 1600 (2910)

White to reddish Orthorhombic (110)(100) 1.63–1.69 7.5 3.2 1400 (2550)

Screen scale sieves equivalent

USA series No.

Tyler screen scale sieves, openings per lineal inch

6 8(a) 12 16(a) 20 30 40 50 70 100 140 200 270

6 80(a) 10 14(a) 20 28 35 48 65 100 150 200 270

Sieve opening mm

mm

Sieve p opening, in., ffiffiffi ratio 2, or 1.414

Permissible variation in average opening, þmm 

Wire diameter, mm

3.35 2.36 1.70 1.18 0.850 0.600 0.425 0.300 0.212 0.150 0.106 0.075 0.053

3350 2360 1700 1180 850 600 425 300 212 150 106 75 53

... 0.0937 0.0661 0.0469 0.0331 0.0234 0.0165 0.0117 0.0083 0.0059 0.0041 0.0029 0.0021

0.11 0.08 0.06 0.045 0.035 0.025 0.019 0.014 0.010 0.008 0.006 0.005 0.004

1.23 1.00 0.810 0.650 0.510 0.390 0.290 0.215 0.152 0.110 0.076 0.053 0.037

Note: A fixed pffiffiffi ratio exists between the different sizes of the screen scale. This fixed ratio between the different sizes of the screen scale has been taken as 1:414ð 2Þ. For example, using the USA series equivalent No. 200 as the starting sieve, the width of each successive opening is exactly 1.414 times the opening in the previous sieve. The area or surface of each successive opening in the scale is double that of the next finer sieve or onehalf that of the next coarser sieve. That is, the widths of the successive openings have a constant ratio of 1.414, and the areas of the successive pffiffiffi openings have a constant ratio of 1:414ð 2Þ. This fixed ratio is very convenient; by skipping every other screen, a fixed ratio of width of 2 to 1 exists. (a) These sieves are not normally used for testing foundry sands. Source: Ref 8

Table 7 Typical calculation of American Foundry Society (AFS) grain fineness number Size of sample: 50 g; AFS clay content: 5.9 g, or 11.8%; sand grains: 44.1 g, or 88.2% Amount of 50 g sample retained on sieve USA sieve series No.

6 12 20 30 40 50 70 100 140 200 270 Pan Total

g

%

None None None None 0.20 0.65 1.20 2.25 8.55 11.05 10.90 9.30 44.10

0.0 0.0 0.0 0.0 0.4 1.3 2.4 4.5 17.1 22.1 21.8 18.6 88.2

Multiplier Product

3 5 10 20 30 40 50 70 100 140 200 300 ...

0 0 0 0 12 52 120 315 1,710 3,094 4,360 5,580 15,243

Table 8 Similarity in American Foundry Society (AFS) grain fineness number of two sand samples with different grain size distributions Percentage retained USA sieve No.

Sand A

Sand B

6 12 20 30 40 50 70 100 140 200 270 Pan Total AFS grain fineness No.

0.0 0.0 0.0 1.0 24.0 22.0 16.0 17.0 14.0 4.0 1.7 0.3 100.0 60.0

0.0 0.0 0.0 0.0 1.0 24.0 41.0 24.0 7.0 2.0 0.0 1.0 100.0 60.0

Source: Ref 8

Source: Ref 8

Most mold aggregates are mixtures of new sand and reclaimed sand. The rejuvenation of reclaimed sand is the principal function of the sand preparation system. Adding new sand also helps maintain fineness and permeability by diluting out fines that build up from the breakdown of angular or weak sand grains. Reclaimed sand includes not only reclaimed molding sand but also core sands. Sand returning to the system

from cores (core sand dilution) should be considered as part of the new-sand addition. It is also beneficial to choose the same base sand for molding and coremaking to maintain consistent grain size and distribution when mold and core sands mix during reclamation. Sand Fineness. Most foundries in the United States use the American Foundry Society (AFS) grain fineness number as a general indication of sand fineness. The AFS grain fineness number

Table 5 Properties of intermediate- and low-density manufactured mullite aggregate Intermediate density

Property

Chemical properties, % Al2O3 SiO2 TiO2 Fe2O3 Other

Low density

75 11 3 9 2

48 48 2 1 1

Thermal properties Expansion, % linear change Expansion coefficient, 1  106 in. in.  C Conductivity, W/cm  C Heat capacity, W/(s/g  C) Diffusivity, cm2/s

0.65 6.00

0.61 5.59

0.0066 1.142 0.0028

0.0068 1.180 0.0033

Mineralogical properties, % Mullite Corundum Beta cristobalite Quartz silica

52 48 0 0

75 13 12 0

 

of sand is approximately the number of openings per inch of a given sieve that would just pass the sample if its grains were of uniform size, that is, the weighted average of the sizes of grains in the sample. It is approximately proportional to the surface area per unit weight of sand, exclusive of clay. The AFS grain fineness number is determined by taking the percentage of sand retained on each of a series of standard screens, multiplying each by a multiplier, adding the total, and then dividing by the total percentage of sand retained on the sieves. Table 6 lists the series of sieves used to run the AFS standard sieve analysis. A typical calculation of the AFS fineness number, which includes the multiplier factor, is given in Table 7. It is important to understand that various grain distributions and grain shape classifications can result in similar grain fineness numbers. Table 8 provides a sample sieve analysis demonstrating that two sands assigned the same AFS grain fineness number can have very different grain size distributions. Grain shape determines the surface area of sand grains. As the sand surface area increases, the amount of bonding material must increase if the sand is to be properly bonded. Thus, a change in surface area, perhaps due to a change in sand shape or the percentage of core sand being reclaimed, will result in a corresponding change in the amount of bond required. Rounded grains have a low surface-area-tovolume ratio and are therefore preferred for making cores because they require the least amount of binder. However, when they are recycled into the molding sand system, their shape can be a disadvantage if the molding system normally uses a high percentage of clay and water to facilitate rapid, automatic molding. This is because rounded grains require less binder than the rest of the system sand. Angular sands have the greatest surface area (except for sands that fracture easily and produce a large percentage of small grains

Aggregates and Binders for Expendable Molds / 535 and fines) and therefore require more mulling, bond, and moisture. The angularity of a sand increases with use because the sand is broken down by thermal and mechanical shock. The subangular-to-round classification is most commonly used, and it affords a compromise if shape becomes a factor in the sand system. However, control of grain size distribution is more important than control of grain shape. The grain size distribution, which includes the base sand size distribution plus the distribution of broken grains and fines from both molding sand and core sands, controls both the surface area and the packing density or porosity of the mold. Mold Porosity and Permeability. When molten metal is introduced into a sand mold, gases and steam are generated as a result of the thermal decomposition of the binder and other additives, cores, and contaminants that are present. If the permeability of the mold is not sufficient to allow the escape of the generated gases, mold pressures will increase, impeding the flow of molten metal and causing misruns or cold shuts, or even causing the metal to be blown from the mold. Conversely, if the porosity of the mold is too great, metal may penetrate the sand grains and cause a burn-in defect. Therefore, it is necessary to control mold permeability by proper balancing of sand distribution for adequate mold porosity.

Fig. 7

Grain size distribution controls the permeability of the mold. The highest porosity will result from grains that are all approximately the same size. As the size distribution broadens, there are more grains that are small enough to fill the spaces between large grains. As grains break down through repeated recycling, there are more and more of the smaller grains, and the porosity of the mold decreases. The size of the voids is determined by the size, size distribution, shape, and packing pattern of the grains. Figure 7 illustrates two sizes of rounded sand grains. Figure 8 shows that the voids in a mold face are large for a coarse sand and small for a fine sand, although the total void area per cubic unit of volume is almost the same for both sands. However, these distribution criteria also govern the dimensional stability of the base sand. It is also important to point out that mechanical venting of the mold is much more effective in providing adequate venting of mold gases than relying solely on sand permeability. The fact that gas is generated within the mold cavity is not always a disadvantage. Gas at the moldmetal interface can help prevent metal penetration into the sand. This minimizes burned-on sand grains and resulting problems associated with cleaning and machining the casting. Thus, a balance between mold permeability and gas generation must be maintained. For

Two sizes of rounded sand grains. Original magnification: 35

example, if mold permeability is low because of the fineness of the base sand, the sand additives should be those conducive to the production of a low-volume gas. On the other hand, if permeability is high, it is advantageous to select materials that yield higher levels of gas. Sand quality control is of fundamental importance. Inadequate control of sand is the leading cause of casting defects, scrap, and rework. Various types of sand tests are described in the articles “Green Sand Molding,” “No-Bake Sand Molding,” and “Shell Molding and Shell Coremaking” in this Volume. A basic in-house sand control program, in the long run, is the most effective way for the prompt identification and prevention of sand problems. While testing a system sand is very important, of equal importance is knowledgeable and responsible review of the data and corrective actions when necessary. A good sand control program, with knowledgeable personnel, can help to reduce material usage, scrap, and rework costs since most molding difficulties and casting defects are sand related. Considering the sand laboratory an entry-level job and rotating the people out as soon as they become proficient is not in the best interest of the product quality. Routine review of all test results should include representation from all the various disciplines within the foundry. It should be recognized that to become proficient at sand testing and interpretation of the data requires time and training. It is to the foundry’s benefit to recognize this and to provide the training necessary and realize the value of the personnel responsible for sand control. Included in the control program should be routine monitoring of the dust collection system. Particular attention should be given to the screen distribution, methylene clay content test, and the total combustible level test. Deviations from normal operating levels indicate equipment malfunctions that require corrective action. Screen distribution is important. For system sands comprised of only base sand (plus newsand additions and core sand dilution), if the grain shape remains constant, changes in screen distribution will occur slowly, a major contributor being the condition of the dust collector system. On the other hand, in those systems that have an influx of various types of base sand from varying core processes and sand sources, changes in the base sand distribution will be more dramatic and will cause dimensional problems unless preventive measures are implemented. For these systems, screen distribution should be tested more frequently.

Clays for Green Sand Molding

Fig. 8

Sizes of pores in faces of molds made from coarse sand and from fine sand. Original magnification: 35

Green sand additives can be divided into two categories: clays and carbonaceous materials. The major purpose of the clays is to function as a bonding agent to hold together the sand

536 / Expendable Mold Casting Processes with Permanent Patterns grains during the casting process. The carbonaceous materials reduce penetration and burn-in, control expansion, improve dimensional stability of the mold (seacoal only), improve surface finish, and improve cleanability of the finished casting. Clays normally used in green sand molding are of two general types:
Montmorillonite, or bentonite clays. These

are subdivided into two general types: Western, or sodium, bentonite; and Southern, or calcium, bentonite. The two clays differ in their chemical composition as well as in their physical behavior within a system sand.
Kaolinite, or fireclay as it is normally called The most significant clays used in green sand operations are the bentonites. Western and Southern bentonites differ in the types of ions adsorbed to their surface. Western bentonite has primarily sodium ions adsorbed to the ion exchange sites. Southern bentonite has primarily calcium ions adsorbed to the ion exchange sites. Due to differences in exchangeable ions, soluble salts, and sometimes other compositional differences, the physical characteristics vary also. In general, Western bentonite requires more mulling energy to develop its green strength and has higher hot strength than the

Fig. 9

Structure of montmorillonite. Large closed circles are aluminum, magnesium, sodium, or calcium. Small closed circles are silicon. Large open circles are hydroxyls. Small open circles are oxygen.

Fig. 10

same amount of Southern bentonite. Southern bentonite, at the same concentration, develops green strength with less mulling energy and produces lower hot strength. The fireclays contribute little to green property development but contribute dramatically to dry and hot property development. Bentonites. The most common clays used in bonding green sand molds are bentonites, which are forms of montmorillonite or hydrated aluminum silicate. Montmorillonite is built up of alternating tetrahedra of silicon atoms surrounded by oxygen atoms, and aluminum atoms surrounded by oxygen atoms, as shown in Fig. 9. This is a layered structure, and it produces clay particles that are flat plates. Water is adsorbed on the surfaces of these plates, and this causes bentonite to expand in the presence of water and to contract when dried. As noted, there are two forms of bentonite: Western (sodium) and Southern (calcium). Both are used in foundry sands, but they have somewhat different properties. Bentonites are not electrically neutral and can therefore attract water molecules between the clay plates. Water is also adsorbed on the quartz surfaces. Thus, there is a network of water adsorbed on sand and clay particles that is set up throughout the molding sand. If the clay covers each sand grain entirely, then clay-water bridges form between grains. In the case in which the clay coverage is nonuniform, similar bridges are formed. Western Bentonite. In Western bentonite, some of the aluminum atoms are replaced by sodium atoms. This gives the clay a net negative charge, which increases its activity and its ability to adsorb water. Western bentonite imparts high green and dry strengths to molding sand, and it has advantages for use with ferrous alloys, as follows. First, Western bentonite develops a high degree of plasticity, toughness, and deformation, along with providing good lubricity when mulled thoroughly with water. Molding sand bonded with plasticized Western bentonite squeezed uniformly around a pattern produces excellent mold strengths.

Second, because of its ability to swell with water additions to as much as 13 times its original volume, Western bentonite is an excellent agent between the sand grains after compaction in the mold. It therefore plays an important role in reducing sand expansion defects. Finally, Western bentonite has a high degree of durability. This characteristic allows it to be reused many times in a system sand with the least amount of rebonding additions. In using Western bentonite, it is important to control the clay/water ratio. Failure to do so can result in stiff, tough, difficult-to-mold sand with poor shakeout characteristics. Although these conditions can be alleviated by adding other materials to the molding sand, control of the mixture is preferable. Southern Bentonite. In Southern bentonites, some of the aluminum atoms are replaced by calcium atoms. Again, this increases the ion exchange capability of the clay. Southern bentonite is a lower-swelling clay, and it differs from Western bentonite in the following ways:
It develops a higher green compressive

strength with less mulling time.


Its dry compressive strength is approxi-

mately 30 to 40% lower.


Its hot compressive strength is lower, which

improves shakeout characteristics.


A Southern bentonite bonded sand flows

more easily than Western bentonite and can be squeezed to higher densities with less pressure; it is therefore better for use with complex patterns containing crevices and pockets. Use of Southern bentonite also requires good control of the clay-water mixture. Southern bentonite requires less water than Western bentonite and is less durable. Clay Blends. In practice, it is common to blend Western and Southern bentonites together to optimize the sand properties for the type of casting, the molding equipment, and the metal being poured. Examples of the effect of mixing bentonites on various sand properties are shown in Fig. 10. At high temperatures, bentonites lose

Effect of blending sodium and calcium bentonites on molding sand properties. (a) Dry compression strength. (b) Hot compression strength at 900  C (1650  F). (c) Green compression strength

Aggregates and Binders for Expendable Molds / 537 their adsorbed water and therefore their capacity for bonding. The superior high-temperature properties of Western bentonite are due to the fact that it retains water to higher temperatures than Southern bentonite (Ref 9). However, if the sand mix is heated to more than 600  C (1110  F), water is driven out of the clay crystal structure. This loss is irreversible, and the clay must be discarded. Fireclay consists essentially of kaolinite, a hydrous aluminum silicate that is usually combined with bentonites in molding sand. It is highly refractory but has low plasticity. It improves the hot strength of the mold and allows the water content to be varied over greater ranges. Because of its high hot strength potential, it is used for large castings. It is also used to improve sieve analysis by creating fines whenever the system does not have an optimum wide sieve distribution of the base sand. However, because of its low durability, its use is generally limited. In addition, the need for fireclay can usually be eliminated through close control of sand mixes and materials. Clay-Water Bonds. As noted, bentonites are not electrically neutral. Thus, water molecules between the clay plates provide adhesive attraction within the network of water adsorbed on sand and clay particles. The clay-water bond can also be explained in terms of the specific surface area of the clay, the type and strength of the water bond at the clay surface, and the hydration envelopes of the adsorbed cations. Clay particles hold adsorbed cations on their surfaces. The bonding of cations on clay particles is weak, and ion exchange is possible in the presence of appropriate electrolytes. Therefore, clay particles and ions are surrounded by electric force fields that direct the water dipoles (the water is polarized at the clay surface) and bind the water network. The field strength decreases with increased distance from the surface of the clay, so that the dipoles closest to the clay surface are bonded most strongly. Beyond the distance at which the force field is effective, the water behaves as a liquid and has no bonding action. There is an ideal water content at which all of the water is polarized and active in the bonding process (because the water added to activate the clay bond is called temper water, this is known as the temper point). Above this water content, some of the water will exist as liquid water, which is not involved in bonding. Below this value, there is insufficient water to develop the bond fully. At the temper point, the green strength of the sand is at its maximum, and additions of water beyond this point decrease the strength of the sand/clay/water mixture. The effect of this can be seen quite clearly in Fig. 11. Controlling Clay Properties. All clays become plastic and develop adhesive qualities when mixed with the proper amounts of water. All clays also can be dried and then be made plastic again by the addition of water, provided the drying temperature is not too high.

Fig. 11

Variation of mold properties with water content. (a) Southern bentonite. (b) Western bentonite. (c) Kaolinite

However, if the temperature becomes too high, they cannot be replasticized with water, and the clay is termed “dead clay.” It is this third condition that dictates the durability of the clay in a system sand. Southern bentonites begin to convert to dead clay once temperature exposure is in the range of approximately 315 to 480  C (600 to 900  F). Western bentonites begin to convert to dead clay in the range of 480 to 705  C (900 to 1300  F), so they are more thermally durable (Fig. 12e). Sand systems with bentonites with higher thermal durability require lower new-clay additions. All clays, regardless of type, develop both adhesive and cohesive properties when mixed with water. The amount of adhesive or cohesive property depends on the amount of water added. When the water content is low, the cohesive properties are enhanced and the clays tend to cohere, or stick to themselves. When water content is high, the clays tend to adhere, or stick to the sand grains to be bonded.

In addition to having different bonding, hot strength, and durability characteristics, the various clays have other very distinctive behavior patterns as a result of their differing physical characteristics. The flowability of sands with Western bentonite is lower than that of sands with Southern bentonite because of the greater swelling tendency of the Western bentonite and its tendency for stickiness. Therefore, the proper formulation of clay materials for a green sand system must take into consideration the flowability requirements as well as the shakeout requirements of the sand. Western bentonite causes more difficult shakeout due to its high hot strength. The ratio of clay to water is of critical importance in optimizing the properties of clays. The ratio of shear strength to green compressive strength of a clay-water mixture increases as mulling is increased. In a sand and clay mixture, water is absorbed by the clay up to its maximum capacity.

538 / Expendable Mold Casting Processes with Permanent Patterns of clay to carbons is constant, and fewer bins are needed for additives. Also, there is less fire hazard from the carbons when preblended with the clays. Some flexibility, however, is lost as the clay and carbon levels cannot be changed independently when using a preblend.

Inorganic Binders In addition to clay, sand can be bonded with other types of inorganic compounds based on silicates, aluminates (cement), or phosphates. The curing (or setting) of inorganic binders is activated by chemical or thermal means. Inorganic binders are also used in the production of ceramic shell molds for investment casting. This section discusses the methods of sand bonding with inorganic compounds. Binders for ceramic shell molds are discussed in the section “Other Expendable Mold Media” in this article. Sodium-Silicate-Bonded Sand Molds. The sodium silicate process is an inorganic system made up of a silicate polymer. This system consists of liquid sodium silicate and CO2 gas. Silicate binders are odorless, nonflammable, and suitable for all types of work (high production to large molds). They are applicable to all types of aggregates and produce no noxious gases upon mixing/molding/coring. A minimum of volatile emissions occurs during pouring/ cooling/shakeout. Carbon dioxide is used to precipitate sodium from what is essentially a silicic acid containing large quantities of colloidal sodium. The reaction of this process is:

Fig. 12

Effect of several variables on the efficiency of clay used as a bonding agent in sand molds. (a) Relationship of shear strength, as measured by pressure required for extruding a continuous worm of clay through an orifice, to water content, for three clays. (b) Effect of type and quantity of clay on erosion of sand-clay mixtures. (c) Effect of temperature on shrinkage of various types of clays. (d) Effect of mold temperature during casting on fusion of clay binder. (e) Effect of temperature on loss of combined moisture in clays

Na2 O 2SiO2 þ CO2 Ð Na2 CO3 þ 2SiO2

Continued gassing gives: Na2 O 2SiO2 þ 2CO2 þ H2 O Ð 2Na2 HCO3 þ 2SiO2

As water content increases, green strength increases and reaches a maximum. The moisture or compactability level at maximum green strength is termed the temper point. Any additional water beyond the temper point causes green and shear strengths to drop, is carried as free water in the system sand, and does not contribute to bonding. However, because excess water may contribute to gas defects such as pinholes, it should be avoided. While the ratio of clay to water in a sand mixture controls the ultimate strength of the sand mixture, the origin of the clay has a significant contribution on the strength potential. Clays from different geographic regions, even though they may be classified as being the same, have different strength curves. However, many of these differences are minimized by modern techniques used in the mining of the clays. Once the type of clay is determined for the system sand based on desired properties, economic considerations must be evaluated, because the geographic location of the foundry

will, in part, dictate the type, or the combination, of clays used in the operation. Clays are often blended to provide special properties. The proper combinations of clays allow the formulation of a system sand conducive to the production of quality castings. Of utmost importance in controlling a green sand system is the selection and consistency of the raw materials introduced into the system sand. Acquisition of the basic raw materials should be from reputable sources only, that is, those that have ongoing quality-improvement programs, including the understanding and application of statistical process control techniques. This is critical for a successful control program. Inconsistency in the raw materials used in the system results in sand variations that no amount of attention or corrective action can overcome. Preblends are in common use in the U.S. foundry industry. These are preblended mixtures of clay and carbons introduced as a single addition. Because they are added as a single addition, system control is simplified, the ratio

This shows that continued gassing dehydrates the amorphous silica gel and increases the strength of the mold. Sodium silicate molds are widely used for large cores and castings where there is a premium on mold hardness and dimensional control. The bond breaks down easily at high temperatures and therefore facilitates shakeout. The silicate-bonded sand, after pouring and shakeout, can be reclaimed by mechanical means, and up to 60% of the reclaimed sand can be reused. Wet reclamation of silicate sand systems is also possible. The amount of silicate binder used for cores and molds varies from 3 to 6%, depending on the type of sand, grain fineness, additives in the silicate binder, and degree of sand contaminants. The type of metal poured and its temperature and the amount of erosion resistance the core or mold will have to withstand are additional factors. A clean, rounded sand grain of 55 fineness requires approximately 3 to 4% of a binder. As the sand fineness increases, the amount of binder that must be used to coat each

Aggregates and Binders for Expendable Molds / 539 grain increases. Thus, a sand of 120 to 140 fineness requires from 1.5 to 3.0% more binder than a sand of 55 fineness. Either continuous-type or batch-type mixers can be used with sodium silicates. Overmixing must be avoided. Mixtures normally have a good bench life, which becomes shorter when higher-ratio silicates are used. Hoppers of mixed sand should be covered with plastic sheets or damp sacking to prevent premature hardening (crusting). The initial tensile strength of cores gassed for 5 s with CO2 varies from 255 to 310 kPa (37 to 45 psi), depending on the binder used. When the core stands, strength increases to a maximum of approximately 670 to 1380 kPa (100 to 200 psi) in 24 h. This increase is due to some dehydration of unreacted silicates and continued gelling of the silicate. Under normal conditions, cores and molds do not deteriorate and can be stored. An exception may occur during periods of high humidity when sand mixtures are unable to achieve maximum strength by dehydration during storage after gassing. Strength deterioration in high-humidity conditions is even more prevalent when the binder formulation or the sand mixture contains organic additives such as sugars, starches, or carbohydrates. Under these conditions, a short heating cycle is necessary to obtain the desired strength. When a core is hardened by carefully controlled gassing, and further hardened by dehydration and polymerization during the subsequent 24 h, good strengths are maintained over a long storage period. Washes may be necessary on cores and molds made by the silicate/CO2 process to prevent burned-on sand and metal penetration. Alcohol-based and solvent-type washes are normally used; isopropyl alcohol is preferred to methyl alcohol because it is less toxic and has a higher boiling point. Specially prepared graphite and zircon refractory pastes, which may be diluted with alcohol or solvent in the foundry, are commercially available. Washes on cores and molds promote peel, aid collapsibility, improve casting surface finish, and aid in resisting moisture absorption during periods of high humidity. Water-based washes (preferably with high solids content) can be used if care is taken to dry the core thoroughly immediately after coating. This procedure must be carried out with care due to the softening effect of the water on the silicate bond. Additives may be used in sodium-silicatebonded sand mixtures to improve shakeout or collapsibility. Sugars are commonly used for this purpose; they can be compounded with the silicate binder or added separately to a sand mixture. Small cores and molds are successfully made with core blowers or shooters combined with gassing stations operating on predetermined cycles and automatically controlled. For these applications, CO2 gas is injected through a hollow pattern or double-wall core box, or by means of a mandrel in a core, or through a hood covering the box.

Larger cores can be cured by means of lance pipes of approximately 5 mm ð 3=16 in:Þ in diameter that are open at the lower end. Using a rod, holes are made approximately 150 mm (6 in.) apart; the lance is inserted into each hole successively, and the gas is applied for 10 to 15 s at approximately 170 kPa (25 psig). The gas permeates and cures an area of sand having approximately a 75 mm (3 in.) radius around the hole. Large cores and molds may be gassed using specially designed, gasketed covers or hoods that fit over the flask or box. It is important to place vents properly to ensure the flow of gas through all parts of the mass. Core boxes and patterns may be made of wood, metal, or plastic and should be washed regularly to prevent sticking problems caused by a buildup of sodium silicate. An ideal release agent is formulated with wax or traditional silicone-based core room release agent. The silicate-bonded sand, after pouring and shakeout, may be reclaimed by mechanical means; up to 60% of the reclaimed sand can be reused. Wet reclamation of the silicate sand system is possible but requires a significant amount of water to scrub the sand. Methods of attrition reclamation combined with low-temperature thermal methods have shown some promise (see the discussion on sand reclamation in the article “Green Sand Molding” in this Volume). Phosphate-Bonded Sand Molds. This inorganic binder system, which consists of an acidic, water-soluble, liquid phosphate binder and a powdered metal oxide hardener, was designed to comply with air quality-control regulations. Because its components are inorganic, fumes, smoke, and odor are reduced at pouring and shakeout. The phosphate no-bake binder system has shakeout properties superior to those of the silicate/ester-catalyzed no-bake systems. The bonded sand can be reclaimed easily by either shot blast or dry-attrition reclamation units. Phosphoric acid bonds are used in both ferrous and nonferrous precision casting to produce monolithic molds. The reaction of this bond has the general form: ½MO þ H3 PO4 Ð MðHPO4 Þ þ H2 O

where M is an oxide frit or mixture of frits. The pH must be controlled carefully and kept acidic (Ref 6). The powdered metal oxide hardener is dry-blended with the sand, and the liquefied phosphoric acid is then incorporated. The coated sand is compacted into core or pattern boxes and allowed to harden chemically before removal. The hardener component is an odorless, freeflowing powder. It must be kept dry; in contact with water, it slowly undergoes a mildly alkaline hydration reaction that alters its chemical reactivity and physical state. Under normal ambient conditions, the material is not hydroscopic. Flow properties of the powdered hardener are good. Standard powder feeding equipment can be used to disperse the hardener into sand mixers.

Curing characteristics of the phosphate no-bake binder system depend on the ratio of hardener to binder. Varying the level of hardener can typically control strip times from 25 min to more than 1 h. Recommended phosphoric acid-base binder levels are from 2.5 to 3.0% for molds and 3.5 to 4.0% for core production. The hardener level should be kept within 18 to 33% of the binder weight for best results. Sand type also affects cure speed. Strongly alkaline sands such as olivine tend to accelerate the cure rate. Zircon forms an extremely strong and stable bond with phosphate binders, and, as a result, shakeout is more difficult. High-quality defect-free castings can be produced using the phosphate binder system for molds and cores with a variety of metals, including gray and ductile irons and various steels. Erosion resistance of both washed and unwashed molds is excellent. Veining resistance is good on unwashed surfaces and can be controlled with the proper coating selection. Aluminate (Cement)-Bonded Sand Molds. Portland cement produces sand molds with very good dry strength. When properly cured and dried, this type of mold has little gas-forming tendency. Cement molds have the advantage of relatively high chill rates but must be used carefully to avoid hot tearing. Most cement molds require drying at higher temperatures, but the method is similar to dry sand molding (see the article “Green Sand Molding” in this Volume). Sand reclamation is expensive.

Organic Binders Organic bonds are used in resin-bonded sand systems. Resin-bonded sands are widely used in metal casting, particularly for cores of all sizes and production volumes and for low-volume, high-accuracy molding. The sand is coated with two reactants that form a resin upon the application of heat or a chemical catalyst. The mixture is typically from 0.7 to 4.0 parts (usually 1 to 2 parts) of binder added to 100 parts of sand. The mixture is then compressed into the desired shape of the mold or core, and the binders are hardened, that is, cured, by chemical or thermal reactions to fixate the shapes. Resin binder processes can be classified into the following categories:
No-bake binder systems
Heat-cured binder systems
Cold box binder systems

In the no-bake and cold box processes, the binder is cured at room temperature; in the shell molding, hot box, warm box, and oven-bake processes, heat cures are applied. Selection of the process and type of binder depends on the size and number of cores or molds required, production rates, and equipment. No-bake, heat cured, and cold box binder systems are compared, respectively, in Table 9, Table 10, Table 11.

540 / Expendable Mold Casting Processes with Permanent Patterns Resin Types

Table 9 Comparison of properties of no-bake binder systems Process(a) Acid catalyzed Parameter

Furan

Ester cured

Phenolic

Relative tensile H M strength Rate of gas evolution L M Thermal plasticity L M Ease of shakeout G F Humidity resistance F F Strip time(b), min 3–45 2–45 Optimum (sand) 27 (80) 27 (80) temperature,   C ( F) Clay and fines P P resistance Flowability G F Pouring smoke M M Erosion resistance E E Metals not (d) ... recommended

Alkaline/ phenolic

Silicate

Oil urethane

Phenolic urethane

Polyolisocyanate

Alumina phosphate

L

M

H

M

M

M


Those composed of liquid polymeric binders

L M G E 3–60 27 (80)

L H P P 5–60 24 (75)

M L P G 2–180 32 (90)

H L G G 1–40 27 (80)

H L E G 2–20 27 (80)

L L G P 30–60 32 (90)

that cross link and set up in the presence of a catalyst (thus transforming from a liquid to a solid)
Those composed of two reactants that form a solid polymeric structure in the presence of a catalyst
Those that are heat activated

P

F

F

P

P

F

F L E ...

F N G ...

F H F ...

G M G (e)

G M P(c) (c)

F N G ...

(a) H, high; M, medium; L, low; N; none; E, excellent; G, good; F, fair; P, poor. (b) Rapid strip times required special mixing equipment. (c) Use with nonferrous metals. (d) Use minimum N2 levels for steel. (e) Iron oxide required for steel

Table 10

The types of resins for bonding of foundry sand fall into the three categories of process types:

Fluid-to-solid transition plastics are primarily furfuryl alcohol-based binders that are cured with acid catalysts. The polymers coat the sand when in the liquid form and are mixed with the liquid catalyst just before being placed in the core box. Alternatively, the catalyst can be delivered to the mix as a gas once the sand mix is in the core box. Reaction-based mold resins include:





Comparison of properties of the heat-cured binder systems

Phenolics (phenol/aldehyde) Oil/urethanes Phenolic/polymeric isocyanates Polyol/isocyanate systems

Process(a) Hot box Parameter

Relative tensile strength Rate of gas evolution Thermal plasticity Ease of shakeout Humidity resistance Cure speed Resistance to overcure Optimum core pattern temperature,  C ( F) Clay and fines resistance Flowability Bench life of mixed sand Pouring smoke Metals not recommended

Shell process

Furan

Phenolic

Warm box

Oven bake (core oil)

H M M F E H G 260 (500) F E E M N

H H L G F H F 260 (500) P G F M (b)

H H M F G M F 260 (500) P F F M Steel

H M L G G H F 175 (350) P G F M (b)

M M M G G L P 205 (400) F F G M (c)

(a) H, high; M, medium; L, low; N, none; E, excellent; G, good; F, fair; P, poor. (b) Use minimum N2 levels for steel. (c) Iron oxide required for steel

Table 11

Comparison of properties for cold box binder systems Process(a)

Parameter

Relative tensile strength Rate of gas evolution Thermal plasticity Ease of shakeout Moisture resistance Curing speed Resistance to overcure Optimum temperature,  C ( F) Clay and fines resistance Flowability Bench life of mixed sand Pouring smoke Erosion resistance Veining resistance Metals not recommended

Phenolic urethane

SO2 process (furan/SO2)

FRC process acrylic/epoxy(b)

Phenolic ester

Sodium silicate CO2

H H L G M H G 24 (75) P G F H G F (c)

M L N E H H G 32 (90) P G G L E F ...

H H L G M H G 24 (75) P E E M F G (d)

L L L G M M G 24 (75) P F F L E G ...

L M H P L M P 24 (75) F P F N G F ...

(a) H, high; M, medium; L, low; N, none; E, excellent; G, good; F, fair; P, poor. (b) FRC, Free radical cure. (c) Iron oxide required for steel. (d) Binder selection available for type of alloy

Curing catalysts include esters, amines, and acids, which can be delivered to the core mix either as liquids or gases. Heat-activated plastics are primarily thermoplastics or thermosetting resins such as novolacs, furans (furfuryl alcohols), phenols, and linseed oils. They are applied as dry powders to the sand, and the mix is heated, at which time the powders melt, flow over the sand, and then undergo a thermosetting reaction. Alternatively, they may consist of two liquids that react to form a solid in the presence of heat. Most binder systems are proprietary. The major ingredients are often mixed with nonreactive materials to control the reaction rate. The reactants are often dissolved in solvents to facilitate handling. Although various materials and schemes are used to form organic bonds in mold-and coremaking, the technology rests on only a few compounds. The presence of so many different systems allows casting producers to tailor the bonding system to the particular application. However, selection of the bonding system requires care. Care must also be taken in controlling process parameters because the systems are sensitive to variations in temperature and humidity. Consideration must also be given to environmental issues in the selection of the system because some organic systems emit noxious odors and fumes.

No-Bake Processes A no-bake process is based on the ambienttemperature cure of two or more binder components after they are combined on sand. Curing

Aggregates and Binders for Expendable Molds / 541 of the binder system begins immediately after all components are combined. For a period of time after initial mixing, the sand mix is workable and flowable to allow the filling of the core/mold pattern. After an additional time period, the sand mix cures to the point where it can be removed from the box. The time difference between filling and stripping of the box can range from a few minutes to several hours, depending on the binder system used, curing agent and amount, sand type, and sand temperature. In North America, phenolic-urethane no-bake molding is probably the most significant (Ref 1). Furan Acid-Catalyzed No-Bake. Furfuryl alcohol is the basic raw material used in furan no-bake binders. Furan binders can be modified with urea, formaldehyde, phenol, and a variety of other reactive or nonreactive additives. The great variety of furan binders available provides widely differing performance characteristics for use in various foundry applications. Water content may be as high as 15% and nitrogen content as high as 10% in resins modified with urea. In addition, zero-nitrogen and zerowater binders are available. The choice of a specific binder depends on the type of metal to be cast and the sand performance properties required. The amount of furan no-bake binders used ranges between 0.9 and 2.0% based on sand weight. Acid catalyst levels vary between 20 and 50% based on the weight of the binder. The speed of the curing reaction can be adjusted by changing the catalyst type or percentage, given that the sand type and temperature are constant. The furan no-bake curing mechanism is shown in Fig. 13. Furan no-bake binders provide high dimensional accuracy and a high degree of resistance to sand/metal interface casting defects, yet they decompose readily after the metal has solidified, providing excellent shakeout. Furan nobake binders also exhibit high tensile strength, along with the excellent hot strength needed for flaskless no-bake molding. They often run sand-to-metal ratios of as low as 2 to 1. Phenolic Acid-Catalyzed No-Bake. Phenolic resins are condensation reaction products of phenol(s) and aldehyde(s). Phenolic no-bake resins are those formed from phenol/formaldehyde where the phenol/formaldehyde molar ratio is less than 1. Again, as with furan no-bakes, these resins can be modified with reactive or nonreactive additives. These resins are clear to dark brown in appearance, and their viscosities range from

Fig. 13

medium to high. Sand mixes made with these resins have adequate flowability for the filling of mold patterns or core boxes. Resins of this type contain free phenol and free formaldehyde. Phenol and formaldehyde odors can be expected during sand-mixing operations. One disadvantage of acid-cured phenolic nobake resins is their relatively poor storage stability. Phenolic binders are usually not stored for more than 6 months. Phenolic resole resins contain numerous reactive methylol groups, and these are generally involved in autopolymerization reactions at ambient or slightly elevated temperatures. The storage period can be considerably longer during the winter months if the temperature of storage remains at 20  C (70  F) or lower. The viscosity advances as the binder ages. The catalyst needed for the phenolic no-bake resin is a strong sulfonic acid type. Phosphoric acids will not cure phenolic resins at the rate required for most no-bake foundry applications. The phenolic no-bake reaction mechanism is: Acid Cured Phenolic þ catalyst ! polymer þ Water resin

The catalyst initiates further condensation of the resin and advances the cross-linking reaction. The condensation reactions produce water, which results in a dilution effect on the acid catalyst that tends to slow the rate of cure. Because of this effect, it is necessary to use strong acid catalysts to ensure an acceptable rate of cure and good deep-set properties. Ester-Cured Alkaline Phenolic No-Bake. The ester-cured phenolic binder system is a two-part binder system consisting of a watersoluble alkaline phenolic resin and liquid ester co- reactants. The reaction mechanism is: Alkaline phenolic resin þ Ester co-reactant ! Suspected unstable intermediate ! Splits to form :

on the resin are used to coat washed and dried silica sand in most core and molding operations. Both the resin and co-reactant are water soluble, permitting easy cleanup. Physical strength of the cured sand is not as high as that of the acid-catalyzed and urethane no-bakes at comparable resin contents. However, with care in handling and transporting, good casting results can be obtained. The distinct advantages of the ester-cured phenolic no-bake systems are the reduction of veining defects in castings and excellent erosion resistance. Silicate/Ester-Catalyzed No-Bake. This nobake system consists of the sodium silicate binder and a liquid organic ester that functions as the hardening agent. High-ratio binders with SiO2:Na2O contents of 2.5 to 3.2:1 are employed for this process, and mixtures usually contain 3 to 4% binder. The ester hardeners are materials such as glycerol diacetate and triacetate, or ethylene glycol diacetate; they are low-viscosity liquids with either a sweet or acetic acidlike smell. The normal addition rate for the ester hardener is 10 to 15% based on the weight of sodium silicate and should be added to the sand prior to the addition of the silicate binder. The curing rate depends on the SiO2:Na2O ratio of the silicate binder and the composition of the ester hardener. Suppliers produce blends of ester hardeners giving work times that are controllable from several minutes to 1 h or longer. The hardening reaction, involving the formation of silica gel from the sodium silicate, is a cold process, and no heat or gas is produced. When added to a sand mixture containing the alkaline sodium silicate, the organic esters hydrolyze at a controlled rate, reacting with sodium silicate to form a silica gel that bonds the aggregate. A simplified version of this curing mechanism is: Sodium silicate Liquid ester Cured þ hardener ! polymer ðNa2 SiO3 Þ

Polymerized phenolic resin Alkaline salts and alcohol

A secondary reaction is thought to occur when the partially polymerized resin contacts heat during the pouring operation, yielding an extremely rigid structure. The viscosity of the ester-cured phenolic is similar to that of the acid-catalyzed phenolic no-bakes. It has a shelf life of 4 to 6 months at 20  C (70  F). Typically, 1.5 to 2.0% binder based on sand and 20 to 25% co-reactant based

The furan acid-catalyzed no-bake curing mechanism

Mixed sand must be used before hardening begins. Material that has exceeded the useful work time and feels dry or powdery should be discarded. The use of sand past the useful bench life will result in the production of weak, friable molds and cores that can result in penetration defects. Curing takes several hours to complete after stripping. Large molds may need 16 to 24 h. Strengths can be higher than those of CO2cured molds, and shelf life is better. Although shakeout is easier than with CO2-silicate systems, it is not as good as the other no-bake processes outlined in this article. Odor and gaseous emission levels are low during mixing, pouring, cooling, and shakeout but depend on the extent of organic additives in the mix. Casting defects such as veining and expansion are minimal. Burn-on and penetration are generally more severe than for the other no-bake systems and can be controlled by sand additives and a wash practice.

542 / Expendable Mold Casting Processes with Permanent Patterns Oil urethane no-bake resins (also known as oil-urethane, alkyd-urethane, alkyd-oil-urethane, or polyester-urethane) are three-component systems that consist of part A, an alkyd oil-type resin; part B, a liquid amine/metallic catalyst; and part C, a polymeric methyl di-isocyanate (MDI) (the urethane component). The three-part system uses the part B catalyst to achieve a predictable work/strip time. It can be made into a two-part system by preblending parts A and B when the amount of the part B catalyst added to the resin controls the work/ strip time. Part A can also be modified for better coating action, improved performance in temperature extremes, or better stripability. Part A is normally used at 1 to 2% of sand weight. The part B catalyst, whether added as a separate component or preblended with part A, is 2 to 10% by weight of part A. The part C isocyanate is always 18 to 20% by weight of part A. Although the oil urethane no-bake system is easy to use, the curing mechanisms are difficult to understand because there are two separate curing stages and two curing mechanisms. When the three components are mixed together on the sand, the polyisocyanate (part C) quickly begins to cross link with the alkyd oil resin (part A) at a rate controlled by the urethane catalyst component of part B, as shown in Fig. 14. This action produces a urethane coating on the sand with enough bonded sand strength to strip the pattern and handle the core or mold. The other stage of the curing reaction is similar to a paint- drying mechanism in which oxygen combines with the alkyd-oil resin component and nearly polymerizes it fully at room temperature to form a tough urethane bond. The metallic driers present in the part B catalyst accelerate the oxygenation or drying (slowly at room temperature or quickly at 150

to 205  C, or 300 to 400  F), but because the full cure is oxygen dependent, section size and shape, along with temperature, determine how long it takes to attain a complete cure. The alkyd-oil urethane mechanism is a twostage process involving: Alkyd þ NCO ðpolymeric isocyanateÞ þ Urethane ! Alkyd catalyst urethane ðpartial cross linkÞ Rigid Alkyd þ O2 þ Metallic driers ! cross-linked urethane

For maximum cure and ultimate casting properties, the mold or core should be heated to approximately 150  C (300  F) in a forced air oven for 1 h. The oil urethane no-bake system, with its unique two-stage cure, results in unmatched stripping characteristics and provides foundrymen with a good method of producing large cores and molds that require long work and strip times. The phenolic urethane no-bake (PUN) Binder system has three parts. Part I is a phenol formaldehyde resin dissolved in a special blend of solvents. Part II is a polymeric MDI-type isocyanate, again dissolved in solvents. Part III is an amine catalyst that, depending on strength and amount, regulates the speed of the reaction between parts I and II. The chemical reaction between part I and part II forms a urethane bond and no other by-products. For this reason and because air is not required for setting, the PUN system does not present the problems with through-cure or deep-set found in other no-bake systems. A simplified version of the curing mechanism for phenolic urethane no-bake systems is: Liquid Liquid phenolic resin þ polyisocyanate Part I Part II Liquid amine þ catalyst ¼ Solid resin þ heat Part III

Fig. 14

Effect of increasing oil urethane system (part B) (catalyst) on work time and strip time

Phenolic urethane no-bake binders are widely used for the production of both ferrous and nonferrous castings and can be successfully used for high-production operations or jobbing shops because of their chemical reaction time and ease of operation. Although many types of mixers can be used with PUN binders, zero-retention high-speed continuous mixers are the most widely used. Because the mixing takes place rapidly, the fast strip times (as fast as 30 s) of the PUN system can be used in practice. No mixed sand is retained in the mixer to harden after it is shut off. Further, the mixers can be coordinated with pattern movement, sand compaction, stripping operations, and mold or core finishing and storage to create a simple manual or fully automated no-bake loop.

Total binder level for the PUN system is 0.7 to 2% based on the weight of sand. It is common to offset the ratio of part I to part II at 55 to 45 or 60 to 40. The third-part catalyst level is based on the weight of part I. Depending on the catalyst type and strip time required, 0.4 to 8% catalyst (based on part I) is normally added. Compaction of the mixed sand can be accomplished by vibration, ramming, and tucking. The good flowability of PUN sand mixes allows good density with minimum effort. Because the PUN system cures very rapidly, the time required for the compacted pattern to reach rollover and strip must coincide with the setup or cure time of the sand mix. For certain ferrous applications (most commonly steels), the addition of 2 to 3% iron oxide to the sand mix can improve casting surface finish. This addition is also beneficial in reducing lustrous carbon defects by promoting a less reducing mold atmosphere. The PUN resin system contains approximately 3.0 to 3.8% N (which is approximately 0.04% based on sand). To reduce the chance of nitrogenrelated casting defects, the part I to part II ratio can be offset 60 to 40 in favor of the part I because substantially all the nitrogen is in part II. It has also been shown that as little as 0.25% red iron oxide is effective in suppressing the ferrous casting subsurface porosity associated with nitrogen in the melt and/or evolved from the PUN binder. The polyol-isocyanate system Was developed in the late 1970s for aluminum, magnesium, and other light-alloy foundries. Previously, the binder systems used in light-alloy foundries were the same as those used for the ferrous casting industry. The lower pouring temperatures associated with light alloys are not sufficient to decompose most no-bake binders, and removal of cores from castings is difficult. The polyol-isocyanate system was developed to provide improved shakeout. The nonferrous binders are similar to the PUN system described previously. Part I is a special polyol designed for good thermal breakdown dissolved in solvents. Part II is a polymeric MDI-type isocyanate, again dissolved in solvents. Part III is an amine catalyst that can be used to regulate cure speed. The chemical curing reaction of the polyolisocyanate system is as follows: Liquid Solid resin Liquid polyol þ ¼ þ polyisocyanate resin heat

In practice, polyol-isocyanate binders are used in much the same way as the PUN binders they evolved from. One difference is that the system does not require a catalyst. Several phenol formaldehyde (part I) resins are available that provide strip times from 8 min to over 1 h. For maximum control, however, an amine (Part III) catalyst can be used to regulate strip times to as fast as 3 min.

Aggregates and Binders for Expendable Molds / 543 For light-alloy applications, binder levels range from 0.7 to 1.5% based on sand. Part I and part II should be used at a 50 to 50 ratio for best results. Reactivity, strengths, and work-time-to-strip-time ratio are affected by the same variables as the PUN binders. Because of the fast thermal breakdown of the binder (Fig. 15), the polyol-urethane system is not recommended for ferrous castings.

Shell Process In the shell process, also referred to as the Croning process, the sand grains are coated with phenolic novolac resins and hexamethylenetetramine. In warm coating, dissolved or liquid resins are used, but in hot coating, solid novolac resins are used. The coated, dry, freeflowing sand is compressed and cured in a heated mold at 150 to 280  C (300 to 535  F) for 10 to 30 s. Sands prepared by warm coating cure fast and exhibit excellent properties. Hotcoated sands are generally more free flowing with less tendency toward caking or blocking in storage. See also the article “Shell Molding and Shell Coremaking” in this Volume. Novolac Shell-Molding Binders. Novolacs are thermoplastic, brittle, solid phenolic resins that do not cross link without the help of a cross-linking agent. Novolac compositions can, however, be cured to insoluble cross-linked products by using hexamethylenetetramine or a resole phenolic resin as a hardener. A simplified version of the novolac curing mechanism is: Novolac þ

HexamethHeat ylenetetramine !

Cured polymer þ ammonia

Phenol-formaldehyde novolac resins are the primary resins used for precoating shell process sand. These resins are available as powder, flakes, or granules or as solvent- or waterborne solutions. A lubricant, usually calcium stearate (4 to 6% of resin weight), is added during the resin production or the coating process to improve flowability and release properties. Hexamethylenetetramine, 10 to 17% based on resin weight, is used as a cross-linking agent. Producing Cores and Molds. The shellresin curing mechanism involves the transition

Fig. 15

Collapsibility of polyol-urethane compared to that of phenolic urethane

from one type of solid plastic to another—thermoplastic to thermosetting. This physical conversion must be completed during a brief period of the shell cycle before the heat (necessary to cure the resin) begins to decompose the binder. Pattern temperatures are typically 205 to 315  C (400 to 600  F). Operating within the ideal temperature range provides a good shell thickness, optimum resin flow, and minimal surface decomposition. Higher pattern temperatures of 275 to 315  C (525 to 600  F) are often successfully used to make small cores, because the shell cycle is short enough that little surface definition is lost by decomposition of the resin at the pattern interface during the relatively brief cure cycle generally needed. Various additives are used during the coating operation for specific purposes. They include iron oxide to prevent thermal cracking, to provide chill, and to minimize gas-related defects. The shell process has some advantages over other processes. The better blowability and superior flowability of the lubricant- containing shell sand permit intricate cores to be blown and offer excellent surface reproduction in shell molding. Because the bench life of the coated shell sand is indefinite, machines do not require the removal of process sand at the end of a production period. Storage life of cured cores or molds is excellent. A variety of sands are usable with the process, and nearly all metals and alloys have been successfully cast using shell sand for cores and molds. See the article “Shell molding and shell coremaking” in this Volume.

Hot Box and Warm Box Processes In the hot box and warm box processes, the binder-sand mixture is wet. A liquid thermosetting binder and a latent acid catalyst are mixed with dry sand and blown into a heated core box. The curing temperature depends on the process. Upon heating, the catalyst releases acid, which induces rapid cure; therefore, the core can be removed within 10 to 30 s. After the cores are removed from the pattern, the cure is complete as a result of the exothermic chemical reaction and the heat absorbed by the core. Although many hot box cores require postcuring in an oven to complete the cure, warm box cures require no postbake oven curing. Hot Box Binders. Conventional hot box resins are classified simply as furan or phenolic types. The furan types contain furfuryl alcohol, the phenolic types are based on phenol, and the furan-modified phenolic has both. All conventional hot box binders contain urea and formaldehyde. The furan hot box resin has a fast cure compared to that of the phenolic-type system and can therefore be ejected faster from the core box. Furan resin also provides superior shakeout and presents fewer disposal problems because of the lack of phenol. Typical resin content is 1.5 to 2.0%.

A simplified hot box reaction mechanism is: Solid resin Liquid resin þ Catalyst þ Heat ¼ þ water þ heat

Catalyst selection is based on the acid demand value and other chemical properties of the sand. Sand temperature changes of 11  C þ5 units in the acid (20  F) and/or variations of  demand value of the sand require a catalyst adjustment to maintain optimum performance. When a liquid catalyst is used, many operations have winter and summer grades that can be mixed together during seasonal transitions. Both chloride and nitrate catalysts are used. The chloride catalyst is the more reactive. Therefore, the chloride is the winter grade, and the nitrate the summer grade. Hot box resins have a limited shelf life and increase in viscosity with storage. If possible, containers should be stored out of the sun in a cool place and used on a first-in-first-out basis. Hot box catalysts have indefinite storage lives. Pattern temperature should not vary more than 28  C (50  F). Measurements should be made at the highest and lowest points across the pattern. Most production shops run hot box pattern temperatures of 230 to 290  C (450 to 550  F), but the ideal temperature is between 220 and 245  C (425 and 475  F). The most common mistake made with the hot box process is to run too high a pattern temperature, which causes poor core surfaces. This condition results in a friable core finish that is especially detrimental to thin-section cores. The color of the core surface shows how thoroughly the core is cured and is a good curing guideline. The surface should be slightly yellow or very light brown—not dark brown or black. Overall, the phenolic and furan hot box resins are extensively used in the automotive industry for producing intricate cores and molds that require good tensile strengths for low-cost gray iron castings. Warm Box Binders. The warm box resin is a minimum-water (50 mm, or 2 in., in wall

section) require additional fasteners for safe operation. End Plate Wedge and Pin Design. Wedges for holding the end plate secure in the mold are principally of two different types: the tapered wedge and the tapered pin. The maximum allowable stress for tapered wedges and pins is 69 MPa (10 ksi). Wrought steel bar stock, such as cold rolled carbon steel or stainless steel, can be used; cast pins or wedges should not be used. Mold Adapter Tables. A mold adapter table is usually furnished or required with a vertical centrifugal casting machine to facilitate attachment of the molds to the machine. Also furnished or required with the mold adapter table is a mold centering boss or index boss/plug for aligning and centering the mold so that it is concentric with the spinner shaft of the machine. Two methods can be used to secure the mold to the adapter table. One method is to allow for a flanged extension of the bottom mold plate with holes for bolting directly into the table. The other method is to use a flanged extension of the mold bottom plate with dog clamps fastened to the table in the same manner in which a part is held on a mill table or vertical lathe. The vertical centrifugal casting machine is available with a water-cooled bottom mold plate. Water cooling requires the machining of radial grooves or slots in the bottom of the bottom mold plate for water passage. Water cooling is sometimes advantageous for extremely heavy wall castings, which can transfer excessive heat into the bottom of the mold. Pretreatment for Metal Molds. All molds, end plates, and surfaces to which mold wash is to be applied should be treated before use in the following manner:  Preheat mold to approximately 150
C (300
F).  Swab inside surface of mold with concen-

trated aqueous ammonium persulfate.

 Flush mold with water to remove all con-

taminants. The mold can be put into service after this treatment. For new molds, the following treatment is recommended before use:  Preheat mold to 205 to 260
C (400 to 500
F ).  Spray with a coating of mold wash approxi-

mately 0.8 mm (1/32 in.) thick.

 Brush out mold wash and repeat spray pro-

cedure. The mold should then be brushed out and resprayed a third time. The mold is ready for use after this treatment. If the wash still does not adhere, the mold should be cleaned again with ammonium persulfate as described previously. Mold Wash for Permanent Molds. The mold wash, when used on a permanent mold, serves as a refractory insulating material. When applied in sufficient thickness, the mold wash insulates the mold, thus reducing the surface temperature of the mold and increasing its useful life. A wide variety of centrifugal casting mold washes are commercially available, including silica, zirconia, and alumina washes. The centrifugal mold wash must be inert to the molten metal being cast. The insulating characteristic of the mold wash is necessary to retard or slow the initial solidification rate in order to eliminate the formation of cold shuts, laps, droplets, and so on, and to produce a high-quality homogeneous outside surface on the casting. In most cases, a mold wash coating thickness of 0.8 mm (1/32 in.) is desired to obtain satisfactory castings. Spraying equipment is available for applying centrifugal casting mold washes. Water Cooling of Permanent Steel Molds. The temperature of the mold increases with each casting poured. The mold can be operated at temperatures to 370
C (700
F ). However, the usual operating temperatures of the mold should range from 150 to 260
C (300 to 500
F ). To maintain this mold temperature for successive casting production, water cooling of the outside surface of the mold is generally employed. The high velocity of the water impinging on the mold surface will prevent the formation of an insulating steam barrier, which would actually inhibit the extraction of heat from the mold. Quick-opening valves mus be used to prevent warpage of the mold, and the spinning of the mold should be interlocked with the water valves to prevent spraying water on a mold that is not spinning and ultimately warping an expensive mold. In practice, the water cooling is usually activated immediately upon completion of the pouring and allowed to remain on long enough to permit

684 / Centrifugal Casting Total cycle time between pours should be as long as possible to allow the graphite mold to cool to 95 to 150
C (200 to 300
F), thus reducing the time that the graphite will be above its oxidizing temperature of 425 to 480
C (800 to 900
F). Permissible rebores are very influential in determining total mold life. If a casting size is of such tolerance that rebored molds cannot be considered and if there is no larger size of bushing or casting for which the mold can be used, it is obvious that total mold life will be considerably shortened.

Process Details Casting Inside Diameters. When making castings on a vertical centrifugal casting machine, the inside diameter (bore) of the

casting will be tapered in accordance with the following formula: sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi h r21  r22

n ¼ 264

(Eq 1)

where n is the speed of rotation (in revolutions per minute), r1 is the inside radius at the top of the casting (in inches), r2 is the inside radius at the bottom of the casting (in inches), and h is the casting height (in inches). Actually, if the length of the casting does not exceed approximately twice its inside diameter, the amount of taper will be negligible. The optimal speed of rotation results in a centrifugal force of 75 g (75 times the force of gravity) on the inside diameter. It can be seen from Eq 1 that too slow a speed of rotation will result in excessive taper on the inside diameter of the casting.

Casting inside diameter, mm 4000

00 000 2

25

00 300 000 1 1

15

0

80

0

50

0

40

0

30

0 5 0 5 0 20 17 15 12 10

75

50

0 800

0 0) 60 00 (2

00 00 1

16

3000

)

0 ) ) 0 00 ) 20 ) 00 00 50 15 000 1 000 (35 (30 (2 (4 (5

40

25

2000 1500

1000 800 0g

600

30 g 200 g 160 g g 120

500 Mold speed, rpm

400 300

140 g 100 g 75 60 g

45

g

g 250 g 180

15

g

5g

30

g

200 1g

5 (1 00 50 0)

150

100 80 70 60 50

30 0 2 (7 00 (1 50 00 ) 0) 15 (5 0 00 )

the mold to be sprayed with the mold wash at the proper temperature range for the next casting cycle, after extracting the solidified casting. Graphite and Carbon Molds. The choice between using a graphite or a carbon mold depends primarily on the availability to the user. Graphite has a higher rate of heat conductivity than carbon and is therefore sometimes used because of the desired metallurgical properties of the finished casting. Graphite can be easily machined into a variety of intricate mold forms with a very good surface finish. Graphite molds have excellent chill characteristics, with thermal conductivity three times that of iron and a specific heat approximately double that of iron. The chilling ability of a material is roughly equal to the product of its mass multiplied by its specific heat. Graphite is nonreactive with most molten metals. In casting phosphorus bronze, graphite molds are not burned into as iron molds are. Carbon pickup is negligible in casting low-carbon stainless steel because of the quick chilling ability of the graphite, which almost instantaneously solidifies a skin layer of steel and therefore makes carbon pickup impossible. In certain cases where rapid chilling is not desired, an insulating mold wash is used. Graphite molds are extremely resistant to thermal shock, with a thermal conductivity approximately three times that of steel and a low Young’s modulus. The strength of graphite molds increases with temperature. Graphite molds are generally designed with two considerations: minimum wall thickness and weight ratio of mold to casting. A steel sleeve or master die holder is usually employed with a heat shrink-fit over the graphite mold. The steel is preheated to 425
C (800
F) for this shrink-fit; therefore, there is considerable compression on the graphite mold. The graphite mold wall thickness must be sufficient to withstand the stress of the shrink-fit without breaking. For heat shrink-fit applications, the graphite mold wall should not be thinner than 19 mm (3/4 in.). The second factor is the minimum weight ratio of mold to casting to reduce the effects of graphite oxidation and to obtain the proper chilling effect. There are many factors involved in arriving at this ratio, but in general, the ratio should be a minimum of 0.75. The factors that would influence a larger ratio include the heavy load imposed because of rapid succession of casts, greater chill depth desired, availability of graphite mold stock, and general flexibility in machining. The graphite mold can be machined before or after the metal jacket or steel sleeve has been encased around the graphite. The expected life of a graphite mold depends on three major factors: stripping time, total cycle time, and permissible rebores. Stripping time should be as long as possible to permit the casting to shrink away from the mold, particularly if the casting diameter is less than 150 to 200 mm (6 to 8 in.). With this precaution, the casting, with burrs, will not score the graphite mold wall nearly as much as when extracting the casting immediately after solidification when there is only negligible casting shrinkage.

Centrifugal force, multiples of the force of gravity, g Peripheral speed, m/min (ft /min)

40 30

20 100 80

60 50

40

30

20

15

10

8 7

6

5

4

3

2

1.5

1

Casting inside diameter, in.

Fig. 6

Nomograph for determining mold speed based on the inside diameter of the casting and the required centrifugal force. See text for example of use.

Vertical Centrifugal Casting / 685 There are castings for which it is desirable to cast the inside diameter with a predictable taper. Using Eq 1, the exact speed can be calculated to produce an approximate given taper on the inside diameter of the casting. Speed of Rotation. To establish a temperature gradient of the molten metal from the outside diameter toward the center of rotation (that is, directional solidification), it is usually necessary for the mold to be spinning when the metal is poured. In some cases, in order to eliminate defects such as erosion and dirt in sand molds, it is desirable to pour at a slow speed of rotation. However, true centrifugal castings having a wall section of 12.7 mm (½ in.) or less must be poured at spinning speed because the metal in this thin section solidifies quickly. Nomographs are available for determining the proper speed of rotation for centrifugal casting. However, Eq 2 can be used to calculate spinning speed: g ¼ 0:0000142Dn

Crane

Mold lifting tong

Casting

Mold

(Eq 2)

where g is the centrifugal force (in pounds per pound or number of times gravity), D is the inside diameter of the casting (in inches), and n is the speed of rotation (in revolutions per minute). Equation 2 can be easily manipulated to solve for speed. Mold Speed Curves. Mold speeds are determined by the inside diameter of the castings to be made. The mold speed curve shown in Fig. 6 is based on the inside diameter of the casting. The length of the casting is not considered in determining mold speed. For example, the mold speed for producing a casting 100 mm (4 in.) in outside diameter by 75 mm (3 in.) in inside diameter at a centrifugal force of 60 g is calculated as follows. Find the 3 in. diameter at the bottom of the curve. Move vertically from this point until the 3 in. line intersects the diagonal line marked 60 g. From this intersection, move directly to the right-hand edge of the curve; the speed of rotation of the mold in this case should be 1150 rpm. From the standpoint of safety for vertical centrifugal casting, it is highly recommended that the g force acting on the outside diameter of the mold be considered. It is safe practice to limit this force to approximately 200 g on the outside diameter of the mold. Pouring Techniques. For permanent molds, the metal is generally poured approximately 40
C (100
F) higher than the temperature used for the same casting if poured statically in a sand mold. This is because of the more rapid chilling effect of permanent molds. The pouring rates required for successful permanent mold centrifugal casting are quite high compared to those for static casting in sand molds. It is particularly important that the rate of pour at the beginning be very high to prevent cold laps and cold shuts. For most types of centrifugal castings weighing less than 45 kg (100 lb), a pour rate of approximately 9 kg/s (20 lb/s) is recommended. For castings weighing up to

Machine table

Fig. 8

Fig. 7

(a) Typical pouring spout design and (b) pouring spout positioning during casting. ID, inside diameter; OD, outside diameter

450 kg (1000 lb), an initial pour rate of 9 to 23 kg/s (20 to 50 lb/s) is recommended. For castings weighing more than 450 kg (1000 lb), pour rates of 45 to 90 kg/s (100 to 200 lb/s) are recommended. When pouring into a vertically spinning mold, it is important to introduce the molten metal into the mold in such a way as to prevent or minimize turbulence of the molten metal, which can cause splashing, spraying, or droplets and can result in undesirable casting defects. Although many vertical centrifugal castings can be poured directly into the mold from the ladle to produce a quality centrifugal casting, it is more often desirable to use a pouring funnel. With a pouring funnel, the nozzle can

Lifting tong assembly used to extract centrifugal castings from the mold

be lined to the required diameter so that, with a certain riser height of molten metal in the funnel, a controlled pour rate can be obtained for a particular casting weight. In addition, with a pouring funnel, the entry of molten metal into the mold can be made to impinge on the body of the mold with initial metal flow in the direction of mold rotation. This type of pouring will provide superior casting quality by minimizing or eliminating any upsetting turbulence in the flow of molten metal that may cause defects. Figure 7 shows a pouring funnel and funnel position. Extraction of Castings. Commercially available casting pulling tongs (Fig. 8) are recommended for extracting vertical centrifugally cast castings. These pullers engage onto the inside diameter of the casting and are used to lift the casting from the mold. Centrifugal force increases directly as the square of the speed of rotation and directly as the length of the radius from the axis of rotation. Centrifugal force can be tremendous and very destructive. Speeds of rotation should never exceed those required to produce the casting. Molds must be centered on the spinning axis as accurately as possible and must be statically or dynamically balanced, if necessary. The method of attachment of both bottom and top cover plates to the mold is of great importance for withstanding and containing

686 / Centrifugal Casting the force of the molten fluid metal while spinning. All clamping arrangements should be designed to tighten with, or to be unaffected by, centrifugal force. Molds should be firmly clamped to the table, because unbalanced molds may fly off the table during operation. Adequate safety guards with interlocks should be used around all machines to protect workers

from molten metal, which can spray from the mold if too much metal is poured. Heat expansion can occur suddenly and rapidly in the mold body, top and bottom mold plates, and even into the centrifugal casting machine. This sudden expansion can put bending and shearing stresses into fasteners and other retention and clamping devices. A

thorough understanding of the forces involved in centrifugal casting is necessary to ensure the utmost in safety for all concerned. Moisture in a sand mold or moisture in the mold wash can turn into steam when the molten metal contacts it, and the resulting forces would be incalculable.

ASM Handbook, Volume 15: Casting ASM Handbook Committee, p 689-699 DOI: 10.1361/asmhba0005260

Copyright © 2008 ASM International® All rights reserved. www.asminternational.org

Permanent Mold Casting Revised by Don Tyler, General Aluminum Manufacturing Co., and Robert Pischel, FOSECO Metallurgical, Inc.

THE PERMANENT MOLD CASTING PROCESS involves the production of castings by pouring molten metal into permanent metal molds using gravity, low pressure, vacuum, or centrifugal pressure. Simple reusable cores are usually made of metal. More complex cores are made of sand, plaster, ceramics, or even salt. When nonmetal destructible cores are used, the process is called semipermanent mold casting. Basically, the process involves the following steps: 1. A refractory wash or coating is applied to the surfaces of the preheated mold that will be in direct contact with the molten metal alloy. 2. Cores, if applicable, are inserted, and the mold is closed either manually or mechanically. 3. The alloy, heated to above its melting temperature, is introduced into the mold through the gating system. 4. After the casting has solidified, metal cores and loose mold members are withdrawn, the mold is opened, and the casting removed. 5. Steps 2 to 4 are repeated until repair of the refractory coating is required, at which time step 1 is repeated to the extent necessary. 6. The usual foundry practice is followed for trimming gates and risers from the castings. A basic difference between sand casting and permanent mold casting is the metal molds for the latter process. These metal molds are usually made of iron or steel and have a production life of 10,000 to 120,000 or more castings as opposed to sand casting, which requires a new mold for each casting produced. Permanent mold casting is particularly suitable for the high-volume production of castings with fairly uniform wall thickness and limited undercuts or intricate internal coring. The process can also be used to produce complex castings, but production quantities should be high enough to justify the cost of the molds. Compared to sand casting, permanent mold casting permits the production of more uniform castings, with

closer dimensional tolerances, superior surface finish, and improved mechanical properties. In many applications, the higher mechanical properties can justify the cost of a permanent mold when production volume does not. Figure 1 shows castings made by the permanent mold process (Fig. 1a) and the semipermanent mold process (Fig. 1b). Metals that can be cast in permanent molds include the aluminum, magnesium, zinc, and copper alloys and hypereutectic gray iron. Aluminum. The vast majority of permanent mold castings are made in aluminum. Therefore, the major content of this text is devoted to aluminum. High-production permanent mold castings weighing up to 70 kg (150 lb) have been made from aluminum alloys. However, much larger castings can be produced. For example, aluminum alloy engine blocks with a trimmed weight of 354 kg (780 lb) have been produced in a four-section permanent mold having a vertical parting line. Magnesium alloys, despite their comparatively low castability, have been cast in permanent or semipermanent molds to produce relatively large and complex castings. For

example, an 8 kg (17.7 lb) housing for an emergency power unit was poured from alloy AZ91C (UNS M11914) in a semipermanent mold. In another application, 24 kg (53 lb) spoolhead castings 760 mm (30 in.) in diameter were produced from alloy AZ92A (UNS M11920) in a two-segment permanent mold with vertical parting. Copper casting alloys from the silicon bronze, aluminum bronze, and high-copper families are often cast in permanent molds. Alloys with a narrow solidification range are preferred. Conventional materials for these permanent molds are H13 and beryllium-copper, and new materials are being investigated (Ref 1). Permanent mold casting of copper alloys has been more prevalent in Europe and Asia than in North America, but it is gaining wider acceptance. Castings up to 13.6 kg (30 lb) are cast routinely. Larger castings are increasingly difficult and above 22.7 kg (50 lb) are considered impossible. Gray Iron. The production of gray iron castings in permanent molds is seldom practical when the castings weigh more than 13.6 kg (30 lb).

(a)

(b)

Fig. 1

(a) Examples of permanent mold applications: control arm, brake master cylinders and calipers, steering knuckles, valve housing, and engine mounting brackets. (b) Examples of semipermanent mold applications: intake manifolds, thermostat housing, transmission pump housing, engine mounting bracket, and oil pan component. Courtesy of General Aluminum Manufacturing Co.

690 / Permanent Mold and Semipermanent Mold Processes

Gravity Casting Methods Manually operated permanent molds for low production may consist of a simple book-type mold arrangement (Fig. 2a). For castings with high ribs or walls that require mold retraction without rotation, the manually operated device shown in Fig. 2(b) can be used. With either type of device, the mold halves are separated manually after releasing the eccentric mold clamps. Semiautomatic Devices. For high-volume production, manual drives are replaced by two-way air or hydraulic mechanisms. These units can be programmed to open and close in a preset cycle. Therefore, the operation is automatic except for pouring of the metal and removal of the castings. Automatic Tilt Pouring. The mold parting for the static casting devices shown in Fig. 2 is in a vertical plane, which is often the preferred position for mold opening and casting removal. Many castings, however, benefit greatly from tilt pouring, which provides significant advantages over statically poured castings in a vertical plane (Fig. 3). Tilt pouring offers advantages in less pouring turbulence, more consistent filling speeds, less reliance on operator skills, and it facilitates full automation. Also, gating systems can be minimized in many applications; that is pouring gates can also serve as feed risers, and many castings can be poured directly through the riser into the mold cavity without excessive turbulence. An example of a standard commercially available tilt device is shown in Fig. 4. These machines are available in a number of sizes and configurations that are suitable to a large percentage of permanent mold castings. They can accommodate front and back ejection and side, bottom, and top core pull mechanisms. They can be used as individual stand-alone machines as well as in multistation, turntable setups. Figure 5 shows a model of an eight-station automated turntable with A-frame tilt machines. Such a turntable can be outfitted with robotic pouring, core setting, ejection, and casting removal. Steps in the casting process are completed progressively as the casting devices pass through the various work stations. This work cell is suitable for very high production with a minimum of manual labor. Such a setup can easily be run by one operator with assistance for die coating touchup. Identical castings can be run in all stations or a mixture of castings with similar solidification times.

Pouring basin

Web gate

Runner

Vent Core plate

Mold half

Rack Pinion

Sprue

Eccentric mold clamp

Gating system

Core pin (I of 5) Mold cavity Riser

Mold half

(a) Mold half

Gearbox for mold mover

Alignment pin (I of 2)

Pouring basin

Gating system Ejector plate

Eccentric mold clamp

Gearbox

Ejector pin (I of 4) Mold half Base

Core

Alignment key

(b)

Fig. 2

Two types of manually operated permanent mold casting machines. (a) Simple book-type mold for shallowcavity castings. (b) Device with straight line retraction for deep-cavity molds

Molten aluminum Mold cavity Note: Insert may be needed to produce bottom pour ladle effect

Plane parallel to floor

Advantages and Disadvantages There are numerous factors that must be considered when choosing a casting process. Therefore, the decision should be based on thorough engineering analysis and production cost studies for each casting. Due to more rapid heat transfer and solidification times, castings

Fig. 3

Cross section of a mold and pouring basin of a tilt-pouring machine. Pouring basin is filled when mold is horizontal. Mold is tilted to upright position to fill the cavity. Source: Adapted from Ref 2

Permanent Mold Casting / 691

Front ejection Ejection platen

Casting catcher

Fig. 4

Tilt device with casting catcher and front and back ejection mechanisms. Courtesy of CMH Manufacturing Co., Lubbock, Texas

contain entrapped gas such as is possible in high-pressure die castings. Therefore, they are superior to die castings in soundness and pressure tightness and are sometimes chosen for superior quality benefits. However, die castings achieve nearer net shape, require less draft, and can be produced at a lower cost. The lower cost is due to shorter cycle times and reduced degate and finishing costs. Degating is the removal of gates, runners, and risers (feeders). Die casters use trim dies and presses to accomplish this. In permanent mold casting, trim dies and presses are used more frequently for parting line flash removal, but in some cases, where gating contact thickness will allow it, they can also work for degating. The more common methods would be belt sanding, sawing, turning, and milling or a combination of these coupled with robotics for high-production applications. These operations are obviously slower than trim press operations. If part complexity, surface finish, and dimensional accuracy are critical, an investment casting may be more suitable. Extremely complex or irregular shapes may be difficult or even impossible to cast in permanent molds. For very complex shapes with intricate internal coring, sand casting may offer the best solution.

Applications Core set station Eject station

Pouring station

Clockwise table rotation

Fig. 5

Drawing of an eight-station turntable or carousel for automatic permanent mold casting with robotic pouring and ejection. Courtesy of CMH Manufacturing Co., Lubbock, Texa

produced in permanent molds have finer dendrite arm spacing, grain structure, and higher mechanical properties than those cast from a similar alloy in sand. This advantage can translate into thinner walls and significantly lower-weight designs. In addition, cast surfaces

are generally smoother than sand castings, and closer dimensional tolerances can be maintained. When properly produced, permanent mold castings contain less shrinkage and gas porosity than sand castings, and they do not ordinarily

Aluminum permanent mold castings are used widely throughout industry. Oil pans and engine cradles, pistons, cylinder heads, intake manifolds, and other functional parts of internal combustion engines for the automotive, trucking, diesel, and marine industries are typical applications. Other major uses for permanent mold castings include internal and accessory parts for reciprocating and jet-type aviation engines; missiles; forms for concrete; textile machine parts; electric motor housings; portable and hand tool components; support members for outdoor light standards; electric griddles and kitchen pots and pans; and a host of other commercial applications. Since the 1990s, there has been major growth in automotive applications. Prompted by the need to improve gas mileage, aluminum has replaced iron and steel for weight-savings benefits. Uses include cross members, subframes, wheels, and other safety-critical applications such as front and rear steering knuckles and control arms, brake calipers, and master cylinders.

Casting Design Good permanent mold casting designs generally follow the same guidelines as for any casting. The designer needs to have a good knowledge of the characteristics of the many available casting alloys and the various heat treatments and their effect on the production and performance of the casting. There are many sources of information available, including

692 / Permanent Mold and Semipermanent Mold Processes ASM International (Ref 3), the American Foundry Society (Ref 4), and SAE International (Ref 5). For specific materials, the Aluminum Association (Ref 6) and the Copper Development Association (Ref 7) provide information and resources. Among the basic rules for good casting designs are:  Limit drastic changes of thickness to mini-

mize stresses and aid in the casting process.

 Taper sections as liberally as possible    

toward risers to help control directional solidification. Do not interpose thin sections between thick sections and access to risers. Avoid extensive horizontal, flat surfaces. Avoid isolated thick sections that create hot spots that are difficult to feed. Determine casting properties and use commensurate safety factors.

Mold Design A number of factors must be considered when designing and building molds for sound permanent mold castings to be produced efficiently and cost-effectively. These factors are:

Multiple-Cavity Molds. The number of castings per mold is a major consideration in designing the mold; the objective is to have the optimal number of cavities per mold that will yield acceptable castings at the lowest cost. Except for very small and thin castings, the machine cycle time increases as the weight of the metal being cast per mold increases. However, these increases are not directly proportional. A mold with the maximum number of cavities will often produce more castings per unit of time than a mold with a smaller number of cavities that was designed to operate on a shorter cycle. This is because there is a minimum solidification time for every casting, regardless of the number of cavities in the mold. The number of rejects sometimes increases as the number of cavities is increased, but this is usually offset by the greater productivity. Venting. The gap that exists between the mold halves after closing is sometimes large enough to permit air to escape and thus prevent misruns and cold shuts. Frequently, however, vents must be added to allow air to escape as the mold fills. Ejector pins provide an excellent opportunity for venting air. Figure 8 illustrates how ejector pins can be ground with flats for venting purposes, along with two other vent

back (outside) of the mold so that the exterior conforms roughly to the cavity. This permits a more even temperature distribution and heat dissipation. The maintenance of a uniform equilibrium temperature during operation is critical to consistent production of high-integrity castings and is a function of consistent cycle times as well as the contoured mold thickness. Inconsistent or interrupted cycle times result in a loss of progressive temperature gradients and a nondirectional solidification pattern. For castings with heavy sections, the adjacent mold sections are generally heavier. For aluminum castings, a ratio of three or four mold wall thicknesses to one casting wall thickness is often used, but a mold wall this thin cannot always be used in making thin-wall castings without jeopardizing mold stability. Thin-wall castings are sensitive to temperature changes and will misrun readily. Consistent, relatively high mold temperatures are required; this necessitates the use of thicker, more massive molds. Suggested thickness based on cast wall thickness and weight is given for aluminum in Fig. 6. Mold temperature will vary with mold thickness and the weight and section thickness of the casting. Figure 7 gives the relative effect of mold thickness on operating temperature and solidification time.

 Cavity dimensions must compensate for the

shrinkage that occurs as the casting cools.

 Undercuts on the outside of a casting com-

 

 

Mold Wall Thickness. In designing a permanent mold, the part is laid out in the desired orientation, and the mold is designed around it, allowing sufficient space for gating and sealing to prevent metal leakage and for coring and inserts. It is common practice to contour the

Aluminum (water cooled mold for 1.3 mm, or 0.50 in. and over) 0 6 13 4

Weight of part being cast, kg

19

25

100

3

75

2

50

1

25

0 0

0.25

0.50

0.75

1.00

Wall thickness of casting, in.

Fig. 6

500

0

0

Mold wall thickness, mm

Mold wall thickness, in.



Wall thickness of casting, mm

0

0

Entire cavity in one mold sector 45 90

50% of cavity in one mold sector 135 180

100 200 300 Weight of part being cast, LB

400

Suggested mold wall thickness based on casting wall thickness and casting weight. Source: Adapted from Ref 2

13

Mold thickness, mm 25 38 51 64

Mold thickness, mm 13 25 38 51 64

0

0

Mold thickness, mm 13 25 38 51 64 38 mm (1.50 in.) casting

1.9 mm (0.075 in.) casting

0.63 mm (0.025 in.) casting

Solidification time, s



plicate mold design and increase casting cost because additional mold parts or expendable cores are needed. Complicated and undercut internal sections are usually made more easily with destructible cores than with metal cores, although collapsible steel cores or loose metal pieces can sometimes be used instead of expendable cores. Proper selection of mold material and wall thickness is necessary for economy and consistent mold temperature profiles. The gating system carries the metal into the mold cavity at the proper location and fill rate to promote directional solidification. The riser (feeding) system contributes to proper temperature gradients and availability of molten metal to feed the solidifying metal front. Venting must be placed advantageously to allow air and gas to escape ahead of the poured metal. Chills, which are sometimes inserted in the mold to initiate and accelerate solidification in desired locations and directions, must be properly sized and located. Antichills are occasionally needed. Use of both must be judicious.

400 6

300 200

4.5 Cycle time 1.5 min

8

3.5

6

2

10

6

100 10

0 0

Fig. 7

0.5 1.0 1.5 2.0 2.5 Mold thickness, in.

0

0.5 1.0 1.5 2.0 2.5 Mold thickness, in.

0

0.5 1.0 1.5 2.0 2.5 Mold thickness, in.

Relationship of solidification time to mold thickness for various casting section thicknesses for aluminum. Source: Adapted from Ref 2

Permanent Mold Casting / 693

Ejector pin with flat ground to allow venting

Grade 5 bolt

Weld in place

Hex stock

Screen mesh Brass body Slots

Steel body

Sintered vent

Fig. 8

Various vent types used in permanent molds. Courtesy of General Aluminum Manufacturing Co.

Table 1

Recommended permanent mold materials for aluminum Number of pours

Casting size

1,000

10,000

100,000

Small castings 25 mm (1 in.) max dimension

Gray iron(a)

Gray iron with H 14 inserts, H 11, H 13(a)

Medium and large castings

Gray iron(a)

Gray iron(a) H11 die steel SAE 1020 SAE 4140 Gray iron(a) SAE 1020 SAE 4140

Gray iron, H 11, H 13(a)

(a) Meehanite types of gray iron are also suitable. Source: Adapted from Ref 2

Direction of core pull

(a)

Lower cost

Direction of core pull

(b)

Higher cost

Fig. 9

Two types of cavities requiring different coring methods. (a) Corner radii must be sacrificed if metal cores are used. (b) Corner radii can be obtained with sand, plaster, or other expendable core material. Casting cost is higher for option (b) due to the additional operations of core making and removal.

designs that are common for metal molds. Mold coatings with significant texture also provide venting action. However, for chronic areas of misruns, a vented mold works best and prevents the overuse of coatings for this purpose. During operation, coatings lose some of their texture, and thereby, the venting is lost unless the coating is repaired. This practice exposes the mold surface to coating buildup and overspray, which can result in areas of higher insulation, hot spots, and interrupted solidification patterns.

Mold Materials Mold materials are usually chosen on the basis of the number of castings to be produced and the cost of the material and the cavity machining. As shown for aluminum casting (Table 1), gray iron is usually the material of choice for low-production applications. Gray irons also offer more rapid heat-transfer properties than tool steel, which can translate into

shorter cycle times and faster solidification rates and more even die heating than tool steel. Gray iron is also less prone to attack by molten aluminum, can be machined for a fraction of the cost of tool steel, and can be readily cast close to size with the desired contoured wall thickness. However, H13 tool steel is by far the most popular mold material for high-production applications. H13 is the most wearresistant material and has proven to be the most cost-effective material on a per casting basis. It provides the best dimensional accuracy, resistance to thermal stresses, and repairability. By contrast, gray iron is more susceptible to thermal stresses and is very difficult to weld repair. Cores. The cores used in the permanent mold process can be of gray iron, steel, sand, or plaster. Metal cores can be movable or stationary. Stationary cores must be perpendicular to the parting line to permit removal of the casting from the mold, and they must be shaped such that the casting is readily freed. Metal cores not perpendicular to the parting line must be movable so that they can be withdrawn from the casting before it is removed from the mold. When sand or plaster cores are used in a permanent mold, the process is referred to as semipermanent molding. With sand or plaster cores, which are expendable, more complex shapes can be produced than with metal cores. The two cored passageways shown in Fig. 9 illustrate the limitations of cores in permanent mold castings. The cored passageway shown in Fig. 9(a) could be produced with metal cores. Sand or plaster cores would be needed to provide the radius of the passageway shown in Fig. 9(b), which would also be more expensive to produce but would provide the desired design characteristics. Sand cores permit the casting of tortuous passages or of chambers or passages that are larger in section than the opening through the wall of the casting. Because sand cores are expendable, their removal presents no major problem but does require an added operation that impacts production costs. Plaster cores are used in semipermanent molds to provide a surface finish better than that obtainable with sand or coated-metal cores. In rare instances, ceramic cores or cores made from molten salt have been used to provide superior surface smoothness. Plaster and ceramic also have excellent insulating qualities and do not prematurely chill the metal in thin sections. Cored holes in permanent mold castings can usually be held to closer tolerances, on both size and location, than in sand castings. Movable cores and stationary cores can both be machined to close dimensions and can be accurately located. Draft requirements for holes formed by steel cores are 3 to 5 per side. Steel cores require the same coating as that applied to the mold; the dimensional accuracy of cavities made from coated steel cores is affected by the same factors that affect the

694 / Permanent Mold and Semipermanent Mold Processes accuracy of dimensions formed by coated-metal molds. Unless a core is stationary, the clearance that must be provided to permit its withdrawal from the mold permits core movement, which may affect dimensions. In terms of dimensional accuracy, sand cores have approximately the same limitations in permanent molds as in sand molds. Because a permanent mold is more rigid than a sand mold, it provides a more accurate seat for locating a core. Collapsible Cores. It is preferable to avoid designs that require the use of collapsible metal cores (multiple-piece cores) in permanent molds. These cores add to the cost of the castings, and their assembly and removal increase production time. Furthermore, dimensional variations can result from the use of collapsible cores, because they cannot be positioned as securely in the mold as single-piece cores or because of movement of core segments when the casting metal is poured. When castings must be designed to be made with collapsible cores, the designer should allow the loosest tolerances possible. Despite the disadvantages of collapsible cores, they are extensively used in making certain castings. For example, nearly all of the aluminum pistons made for the automobile industry in the United States are cast in permanent molds using fivepiece collapsible metal cores. Casting Ejection. Only the most simple permanent mold castings can be ejected from the mold with no mechanical help. Most castings are ejected by well-distributed ejector pins or are confined in one mold half during opening of the mold and then ejected by the withdrawal of the retaining core or cores. It is important that the casting remain in the correct mold half until ready for ejection. For castings that have similar geometry in both mold halves, there may be a need for a spring-loaded, push-off mechanism in one mold half in order to keep the casting on the ejection side during the opening sequence. Figure 10 illustrates a typical tilt pour, permanent mold rigging system with

Return pin

contoured mold back, mounting spacers, ejection plate and pins, and pouring cup. Gating and Riser System. A well-designed gating and riser system accomplishes the following objectives:  Provides rapid filling to prevent premature

metal stream

 Establishes suitable thermal gradients to



promote directional solidification

 Avoids erosion of the mold  Provides a hot reservoir of metal to feed

solidifying sections



finishing costs



 Maximizes casting yield and minimizes

The design of gating systems still remains as somewhat of an art that depends largely on the experience of the foundry design engineer and a set of internally developed trial-and-error rules and practices. One such commonly used beginning approach is modulus based. The volume-to-surface area ratios (modulus) of various parts of the casting are calculated and then used to establish cavity orientation and to design gate and riser size and placement. This is a very effective method that properly assumes solidification patterns that are dependent on the geometry of the part. Considering this information, the following is a set of “rules of thumb” that can be applied along with local foundry knowledge and practice to design a basic, vertically parted, permanent mold gating system:  Orient the thickest casting sections (high

modulus) toward the top of the mold for easy access to risers (feeders) and to provide hydrostatic pressure to thinner (low-modulus) sections.  Whenever possible, gate (fill the cavity) through side risers, thereby producing the Bumper bolt Ejector plate

Ejector pin

Mount spacer Male half

Leader pin

Female half

Fig. 10



freezing with a minimum of turbulence

 Keeps dross and inclusions out of the casting  Prevents air and gas entrapments in the

Carrier bolt

Return pin



Typical tilt pour permanent mold rigging system. Courtesy of General Aluminum Manufacturing Co.



hottest metal and highest mold temperature at these locations. Minimize the amount of metal flowing over any isolated local areas, which may cause hot spots and erosion of the mold material. Orient the low-modulus sections away from the top and side riser to promote directional solidification from that point toward the side and top, where the molten reservoir is available for feeding. Use surface-area-to-volume (SA/V) ratios when designing risers. Round or spherical risers have the lowest SA/V ratios and therefore remain molten longer for maximum feeding. Riser diameter should be at least 15 to 20% larger than the section being fed. Riser height rarely needs to be greater than 1.5 times the diameter. Side and top riser necks (in-gates) should be as close as possible to the casting cavity and be at least 70% of the section thickness.

These items represent general guidelines to begin the gating/riser design. Much more detailed information can be obtained from research papers and training literature from The American Foundry Society publications and their Cast Metal Institute. As expressed in an earlier section, tilt pouring provides a number of advantages over static methods and is the preferred gravity-pouring method for most successful aluminum foundries. The primary advantage is in the reduction of pouring turbulence, which minimizes potential air and oxide entrapments. Tilt pouring allows the metal to enter the cavity in a tranquil manner and can eliminate the need to slow down the metal flow by the use of down-sprue and choke systems. Filters. To avoid oxide contamination, it may be necessary to incorporate dross traps or filters into the gating system (Fig.11). Perforated steel screens or woven fiberglass have been commonly used for filtration and reduction of metal velocity. However, reticulated ceramic foam filters have proven to be much more effective in not only entrapping potential inclusions but also in reducing metal velocity and pouring turbulence. When properly applied, gating system filters of ceramic foam are the most effective means of controlling pouring turbulence and removing oxides, hard spots, and inclusions. The resulting benefits are improved fluidity, reduced solidification shrinkage, and improved mechanical and fatigue properties. However, it may be difficult to justify the per-piece casting cost for high-volume production. The next-best alternative is in-furnace filtration. Ceramic foam or bonded particle filters have been developed for use in dip wells or launder systems and offer many of the same benefits, except for pouring turbulence control. Design Tools. The advent of computer simulation modeling for the permanent mold process has provided an indispensable tool for the foundry engineer to optimize and prove out his gating and riser designs before the tooling

Riser

Sprue

Permanent Mold Casting / 695

Casting

Dross trap

Fig. 11

Examples of a dross trap (left) and a foam filter (right) designed into the gating system. Source: Adapted from Ref 8

is built. Obviously, this can reduce cost and lead times dramatically, problems are analyzed and resolved more quickly, and process optimization is achieved much sooner than traditional methods of trial and error. There are four basic steps to the modeling process:  Specification of the properties and initial

conditions of all materials used in the process  Development of a geometric model of the part to be cast, along with its rigging system and tooling intended for production  Computer simulation of the process  Evaluation of results and, if necessary, modification of the process to improve some aspect, such as part quality, cycle time, or gating optimization Computer modeling of the permanent mold process has progressed to the point that every aspect of the process can be simulated with very good accuracy. Considering the number and price range of computer modeling systems and services available, simulation work is economically feasible for virtually every permanent mold application. Directional Solidification. Alloys should be cast so that solidification takes place directionally and progressively toward the risers, which are generally to the side or on top of the casting. To achieve this solidification pattern, thinner sections of the casting should be away from the gating system, and heavy sections should be adjacent to it. Directional solidification can also be facilitated by proper attention to mold coating, that is, thinner, less insulating applications away from the risers and thicker, more insulating near and in the riser cavity (see the section “Mold Coatings” in this article). Appropriate use of metal chills, that is, high-thermal-conductive materials such as copper-beryllium, Anviloy, or even aluminum, can be inserted into the mold at strategic locations to reduce or eliminate hot spots that interrupt the solidification pattern. Water or air cascades and circulating water channels can be used in conjunction with metal chills or by themselves

to accelerate local heat transfer to eliminate hot spots and direct the solidification sequence.

Mold Coatings Mold coatings are applied to metal mold and core surfaces to serve as a barrier between the molten metal being cast and the surface of the mold during mold filling and solidification. The primary purposes of mold coatings are:  Protect the base mold metal from the liquid    

metal being cast Control metal flow Control heat transfer Control casting surface finish Influence casting release from the mold

Mold Coating Families. There are two basic families of mold coatings: insulating and lubricating. Combination coatings are sometimes used, but the properties of each are somewhat diminished when mixed. The control of heat transfer through all sections of the mold is the most important characteristic of mold coatings, because it allows for control of both the rate and direction of casting solidification. The relative degree of insulation depends on the physical properties of the raw materials used, the method of application of the coating to the mold, and the coating layer thickness. Mold coatings are composed of three basic components:  Refractory fillers: These are typically com-

posed of refractory powders such as titania, talc, mica, vermiculite, alumina, iron oxide, silica, and graphite.  Binder: The vast majority of mold coatings use sodium silicate with an appropriate SiO2/Na2O ratio. Alternate bonding materials may include potassium silicate or clays.  Carrier: Water of a controlled trace mineral content is the universal carrier. The mixture of fillers and their ratios is the first step to govern the relative degree of heat

transfer during solidification. However, all potential properties of mold coating cannot always be obtained from one coating formulation. Often times, certain desired performance characteristics are mutually exclusive, and it is impossible to formulate one coating to meet all casting demands. Differing casting property demands will often require the use of different coatings. Application. The compactness of the sprayed coating layer, which is governed substantially by the application method, influences the heat-transfer property of the as-applied coating. If the coating layer is not very compact, the contact between the particles is slight, and therefore, the permeability and insulation properties will be greater, but the durability will be reduced. Conversely, if the coating layer is very tight, it will be substantially more durable but less insulating. Most commercial mold coatings are supplied as a heavy suspension or paste, allowing for significant flexibility in customizing the required properties based on the rate of dilution in the foundry. Most silicate-bonded coatings can be stored indefinitely as long as they are covered and not allowed to either dry out or freeze. Exposure to freezing temperatures renders the bonding properties of the silicate useless, and they cannot be reconstituted. Surface Preparation. As in any coating procedure, surface preparation is key. If starting with a mold that has previously been in service, the spent mold coating (including residual silicate) must be removed. Various blasting media are available, including sand, metal shot, glass beads, calcined rice hulls, or nut shells. It is important that the blasting not be overly aggressive so as not to distort mold details or alter dimensions. Dry ice blasting, although suitable for touchup or light removal of damaged coatings while in service, is not aggressive enough for full recoat. The mold surface must be clean and free of grease and oil, including fingerprints. Typically, the mold is preheated to between 205 and 370  C (400 and 700  F). A light mist of water (or mild water/surfactant solution) is sprayed on the mold surface. This slightly oxidizes the mold surface and provides a better anchor for subsequent layers to adhere. Coating preparation for spraying can easily be done in the foundry. It is preferred to use deionized water because trace elements in local water supplies can result in coating inconsistency. Coating manufacturers will have recommended starting points for dilution. These can be varied to suit the specific needs of the foundry or casting. Typically, the more dilute the coating, the more durable it will be, while being slightly less insulating. Conversely, more concentrated coatings will be more insulating but less durable. Once diluted, most mold coatings will require some mechanical suspension through slow, continuous mixing. It is important to avoid any high-shear mixing that may trap air and potentially damage softer refractory particles.

696 / Permanent Mold and Semipermanent Mold Processes When the mold and the coating(s) have been prepared, the application can begin. Spraying is the preferred method of application. Although brushing can result in a thicker coating layer, it does little to actually reduce the rate of heat transfer. Furthermore, due to the physical nature of a brushed layer, it is far more susceptible to spalling, chipping, and flaking, which can actually create casting defects. When the mold has reached the appropriate temperature and preheat burners are removed, an optical pyrometer should be used to determine that the temperature of the mold has peaked and is actually beginning to decline. Primer coatings are often applied as underlayers to improve coating life. Specifically designed primers of high durability are usually recommended, although very dilute main insulating coatings can also be used. Lubricating coating should never be used as primers. Spraying should be done with the spray nozzle as square to the mold surface as possible. The spray application should entail the slow building up of numerous thin coating layers to the desired thickness rather than just a few heavy coats. This will ensure both maximum durability as well as insulation. The objective is to atomize the coating so that it will dry almost instantly on contact with the hot mold to develop the strongest possible bond. Typical coating thickness for maximum insulation in this type of thermal system is between 200 and 300 mm (0.008 and 0.012 in.). Excessive coating thickness will not significantly improve the rate of heat transfer. Excessive coating thickness may potentially be detrimental, producing a layer that will be more susceptible to increased flaking, cracking, or catastrophic failure. In gates, runners, and risers where high insulation is desired, it is recommended that a compositionally higher insulating coating be used in preference to spraying or brushing a heavier thickness of a less insulating coating. Spray pattern and angle of the spray gun/nozzle can also influence the deposition of the coating on the mold surface. Wherever practical, the spray gun should be held perpendicular to the surface being sprayed. Contact of the atomized spray at angles to the mold should be minimized. Optimal spray distance should be 60 to 90 cm (24 to 36 in.). Distances too close to the mold may result in “damp” coating or coating that bounces off the mold surface. Spray distances that are too great may result in a coating spray that dries too much as it travels through the air and is too dry to bond to the mold surface. The way the spray gun is held by the operator and moved across the mold surface also plays a role in the coating deposition pattern. To obtain optimal results, the spray gun should be moved with a constant velocity, that is, with a motion of translation imparted to the gun, wrist, and forearm alike, rather than just a rotating wrist motion around a vertical axis. The drawing on the left in Fig. 12 represents the coating distribution that would result from a perfect sinusoidal rotation of the spray

Rotation of spray gun Horizontal passes

Crossed

80

120 100

160 120

100

100

120

160

120

120

100 80

120

Thickness distribution of coating (average = 100)

Fig. 12

These drawings illustrate the differences in coating distribution across the mold surface as a function of spray gun motion. The drawing on the right illustrates a much more even coat as a result of crossed passes and constant velocity. Source: Adapted from Ref 9

gun by the wrist and shows a heavy buildup of up to 160% of the average deposit, formed near the extremities of the moving spray halo course. Even in the best conditions of the translating movement recommended previously, a similar, although less dramatic, buildup is certain to occur. It is advisable that the operator change direction of the spraying motion intermittently. Also, generously overlapping the passes by 50% or more will considerably improve the evenness at right angles to the travel of the spray gun. The sketch on the right in Fig. 12 shows how crossing the passes can improve coating evenness. This practice should be the rule when possible (on flat surfaces, for example). Coating thickness on hot molds can be measured by commercially available magnetic or ultrasonic measuring devices. Some of these devices may have limited use in high-temperature applications; however, adapting them to permanent mold coating may still be preferable to guessing as to actual coating thickness. When the mold is coated, it can be put into service by reheating it again to the desired operational preheat temperature. If the mold is allowed to cool to room temperature before it is put in service, the subsequent preheat to operating temperature should be done slowly to avoid thermal stress and shock to the coating layer. Coating life may vary considerably, not only from foundry to foundry but from casting to casting. Mold complexity and geometry, pouring temperature, cycle time, pouring rate, alloy, and nature of the coating play a part in the potential life of a coating. Catastrophic coating failure or wear are not the only limiting factors of coating life. Even if a coating remains intact on the mold, over time the coating layer will begin to compress due to thermal cycling and pressure. This results in the loss of internal coating porosity and begins to adversely affect the rate of heat transfer. Coatings that have lost a significant degree of their insulating capability over time while remaining intact on the mold are considered “dead.” In such a case,

the mold must be taken out of service, completely cleaned, and recoated. In other cases, only a section of the mold may experience excessive wear. In those cases, in-process touchup may be possible. There are two critical differences in touchup versus initial mold coating. First, the operating temperature of the mold is often greater for touchup than for initial coating, so as not to adversely affect production. This requires a more dilute version of the initial coating, since the atomized coating will now have to travel through a hotter environment before depositing on the mold surface. If the coating is too concentrated, it may completely dry before contact with the mold, leaving a powdery coat that will not adhere to the mold. Secondly, using a mask over the surrounding area to be touched up is often essential. This will ensure that overspray does not overbuild in areas that do not require the additional coating. In addition, other areas of the mold may be coated with alternate coatings for additional property requirements. Masking will ensure that those areas will not be compromised. Surface Quality. Coatings also improve the surface quality of the casting. However, there is often a compromise between surface quality (smoothness) and the ability to move metal over long, flat, thin sections, especially in aluminum and magnesium castings. For any skinforming alloy, such as most aluminum and magnesium alloys, a slightly rough coating surface texture promotes metal flow by constantly rupturing the initial metal skin during flow. Very smooth coatings often promote miss-runs in long, flat, thin castings where the solidifying metal skin can grow uninterrupted. Release. Lubricating coatings can be used over a primer or over an insulating coating to promote the release of the casting from the mold, especially in flat, deep-draw sections or areas with minimal draft. Insulating coatings should never be placed on top of lubricating coatings. By their very nature, the durability of lubricating coatings is less than for the vast majority of insulating coatings, so touchup on

Permanent Mold Casting / 697 a mold in service is quite common and necessary. In extreme cases, touchup between every casting may be required. Coating application is not necessarily a black art but rather a comprehensive set of applied physical parameters that can and should be closely controlled and monitored for best-practice reproducible results. Mold coating, however, still remains an issue of compromise, with casting and mold requirements that often compete against each other. It is always best to prioritize casting requirements, fulfilling each as possible in order, understanding that not all issues can be successfully satisfied with one coating. One measure of a good permanent mold foundry is its ability to control this element of production. It can spell success or failure just as much as other factors.

Mold Life Mold life can vary from as few as 100 pours to more than 250,000 pours, depending on the variables discussed later in this section. With proper care and maintenance, a mold for an aluminum piston made in H13 tool steel, for example, can be expected to produce 250,000 castings before requiring repair. After the production of 250,000 more castings, the repaired mold will require a major overhaul. With repeated repairs and overhauls, the mold can produce as many as 3.5 million castings before being discarded. However, a piston mold, with its relatively simple design, will have a much longer life than a mold that requires elaborate internal coring and external inserts, such as a cylinder-head mold. Mold life is likely to be longer in the casting of magnesium alloys than in the casting of aluminum alloys of similar size and shape; this is because molten magnesium does not attack ferrous metal molds. However, the difference in mold life for magnesium alloys depends to a great extent on the effectiveness of the mold coating used. In the casting of gray iron, mold life is expected to be short compared to the casting of similar shapes from aluminum alloys. Major variables that affect the life of permanent molds are:  Pouring temperature: The hotter the casting

metal is poured, the hotter the mold is operated, which leads to rapid weakening of the mold metal.  Weight of casting: Mold life decreases as casting weight increases.  Casting shape: Mold walls are required to dissipate more heat from castings having thick sections than from those having thin sections. When there is a significant variation in the section thickness of a casting, a temperature differential is set up among different portions of the mold. As the temperature differential increases, mold life decreases.  Cooling methods: Water cooling is more effective than air cooling, but it substantially decreases mold life due to thermal stresses.

 Heating cycles: Generally, a continuous run,





  







in which the mold is maintained at a uniform temperature, provides maximum mold life. Repeated heating and cooling over a wide temperature range will shorten mold life. Preheating the mold: This is done to operating temperature with a gas flame or electric heaters, and it greatly increases mold life. Thermal shock from starting a mold from well below operating temperature is one of the principal causes of mold failure. Mold coating: This protects the mold from erosion and soldering by preventing the metal from contacting or overheating mold surfaces, thus increasing mold life. Mold materials: See Table 1. Storage: Improper storage can lead to excessive rusting and pitting of mold surfaces, which will reduce mold life. Cleaning: The common practices for cleaning molds are abrasive blasting, dipping in caustic solution, and wire brushing. Dipping in caustic can be hazardous to the operator. Wire brushing and abrasive blasting can cause excessive mold wear if not carefully controlled. Glass beads are the safest abrasive blast material; their use minimizes dimensional changes due to erosion from the abrasive blast. The use of dry ice pellets has come into prominence as a blasting medium and is by far the least abrasive material. However, it may be periodically necessary to use a more abrasive medium on certain coatings with the best adherence properties. Gating: A poor gating system can greatly reduce mold life by causing excessive turbulence, erosion, and excessive localized hot spots at the gate areas. Method of mold operation: Although the same materials are used to make molds and cores for both automatically operated equipment and hand-operated equipment, the life of the tool materials on hand-operated equipment is shorter because of the abuse the tooling must withstand. Tools for automatic equipment may last up to twice as long as for hand-operated equipment. End use of casting: If the structural function of a casting is more important than its appearance and dimensional accuracy, a mold can be used longer before being discarded.

Influence of Mold Design. In addition to the aforementioned factors, mold design has a marked effect on mold life. Variation in mold wall thickness causes excessive stress to develop during heating and cooling, which in turn causes premature mold failure from cracking. Abrupt changes in thickness without generous fillets also cause premature mold failure. Small fillets and radii lead to reduced mold life because checking and cracking, as well as ultimate failure, often start at these points. Usually, less draft is required on external mold surfaces than on internal mold surfaces, because of the shrinkage in the casting. A 5

draft is desirable, but 2 on external and 3 on internal mold surfaces can be used. Lower draft angles, however, decrease the number of castings that can be made between mold repairs. Projections in the mold cavities contribute greatly to reduced mold life. These projections become extremely hot, which increases the possibility of extrusion, deformation, and mutilation when the casting is removed. Whenever possible, mold life can be extended by using inserts to replace worn or broken projections.

Mold Temperature If the mold temperature is too high, excess flash develops, castings are too weak to be extracted undamaged, and mechanical properties and casting finish are impaired. When mold temperature is too low, cold shuts and misruns are likely to occur, and feeding is inhibited, which generally results in shrinkage, hot tears, and sticking of the casting to molds and cores. The variables that determine mold temperature include:  Pouring temperature: The higher the pour  

  

ing temperature, the higher the temperature of the mold. Cycle frequency: The faster the operating cycle, the hotter the mold. Casting weight: Mold temperature increases as the weight of molten metal increases. Casting shape: Isolated heavy sections, cored pockets, and sharp corners not only increase overall mold temperature but also set up undesirable thermal gradients. Casting wall thickness: Mold temperature increases as the wall thickness of the casting increases. Mold wall thickness: Mold temperature decreases as the thickness of the mold wall increases. Thickness of mold coating: Mold temperature decreases as the thickness of the mold coating increases.

After the processing procedure has been established for a given casting operation, mold coating, cycle frequency, chills, and antichills have significant effects on mold temperature. Mold coating is difficult to maintain at an optimal thickness, primarily because the coating wears during each casting cycle and because it is difficult to measure coating thickness during production. The most widely used method for controlling coating thickness is periodic inspection of the castings. Improper coating thickness is reflected by objectionable surface finish and loss of dimensional accuracy. Preheating of Molds. In many casting operations, molds are preheated to their approximate operating temperature before the operation begins. This practice minimizes the number of unacceptable castings produced during establishment of the operating temperature.

698 / Permanent Mold and Semipermanent Mold Processes Molds can be preheated by exposure to direct flame, although this method can be detrimental to the molds because of the severity and nonuniformity of heat distribution. Customized heaters are often built for molds. Preheating of the mold in an oven or by infrared heating are the best methods because the thermal gradients are of smaller magnitude. Unfortunately, this may be impractical for larger molds. Final mold operating temperatures are achieved after the first few production cycles.

Control of Mold Temperature Optimal mold temperature is the temperature that will produce a sound casting in the shortest time. For an established process cycle, temperature control is largely achieved through the use of auxiliary cooling or heating, control of coating thickness, and cycle-time consistency. Auxiliary cooling is often achieved by forcing air or water through passages in mold sections adjacent to the heavy sections of the casting. Circulating water is by far the more effective method of auxiliary cooling, particularly when coupled with chill inserts, as illustrated in Fig. 13. Scale buildup in cooling water passages must be minimized in order to achieve optimal results. When cooling water is maintained free of solid particulate, scaling is minimized and heat transfer can be managed accurately and effectively to facilitate consistent, predictable, and precise solidification patterns and shorten cycle times for greater productivity. The problem of scale formation from particulate-contaminated water can be solved by the use of recirculating systems containing demineralized, purified water. Water treatment systems such as reverse osmosis and deionization are readily available and simple to maintain for high-production operations. These systems can be coupled with particle counters and automatic controls that can eliminate scaling problems entirely. Water flow is regulated manually to each mold section with the aid of a flow meter. Shutoff valves are used to stop the water flow when the casting process is interrupted, thereby minimizing the loss of mold operating temperature. Strategically placed Rib (I of 4)

Casting

Iron mold Copper antichill

thermocouples or timers can be used to signal on and off water cooling. In addition to the control of water flow, the temperature of the inlet water affects the performance of the mold cooling system and must be controlled to within a suitable maximum range. If water or another liquid coolant is used, it must never be allowed to contact the metal being poured, or a steam explosion will result. The intensity of a steam explosion increases as metal temperature increases. In addition, water will react chemically with molten magnesium. A mold coating of controlled thickness can equalize solidification rates between thin and heavy sections. Chills and antichills can be used to adjust solidification rates further, so that freezing proceeds rapidly from thin to intermediate sections and then into heavy sections and finally into the feeding system. Chills are used to accelerate solidification in a segment of a mold. This can be done by directing cooling air jets against a chill inserted in the mold (Fig. 13) or, more simply, by using a metal insert without auxiliary cooling. Chilling can also be achieved by removing some or all of the mold coating in a specific area to increase thermal conductivity. Chills can be used to increase production rate, to improve metal soundness, and to increase mechanical properties. Antichills serve to slow the cooling in a specific area. Heat loss in a segment of a permanent mold can be reduced by directing an external heating device, such as a gas burner, against an antichill inserted in the mold (Fig. 13). This is an effective but cumbersome method that adds cost and can significantly reduce mold life due to thermal stresses. A simpler, less costly method to slow cooling in a local area would be to cover the exterior mold surface with insulating materials and/or apply high-insulating mold coatings to the cavity surface.

Dimensional Accuracy The dimensional accuracy of permanent mold castings is affected by short-term and long-term variables. Short-term variables are those that prevail regardless of the length of the run:  Cycle-to-cycle variation in mold closure or 

A

Gas burner

A

Fig. 13

Copper chill Section A-A

  

in the position of other moving elements of the mold Variations in mold closure caused by foreign material on mold faces or by distortion of the mold elements Variations in thickness of the mold coating Variations in temperature distribution in the mold Variations in casting removal temperature

Air jet

Air- or water-cooled chills and flame-heated antichills can be used to equalize cooling rates in casting sections of varying thickness.

Long-term variables that occur over the life of the mold are caused by:  Gradual and progressive mold distortion

resulting from stress relief, growth, and creep

 Progressive wear of mold surfaces primarily

due to cleaning Given proper attention to the aforementioned factors, recommended permanent mold dimensional standards are illustrated in Fig. 14. Chart E1 represents basic tolerances for features contained on one side of the mold parting line that reflect variation caused by expansion and contraction of the mold and the metal during solidification. For dimensions crossing the parting line or involving cores or surfaces formed by moving mold members, charts E2 and E3 represent the additional tolerances that apply for such features. These features exhibit additional variation due to the hydrostatic pressure of the liquid metal. The amount of additional variation across the parting line is therefore related to the projected area of the casting at the parting line. Dimensional variations can be minimized to within those illustrated in Fig. 14 by keeping heating and cooling rates constant, by operating on a fixed cycle, and by maintaining clean parting faces. It is particularly important to select mold cleaning procedures that remove a minimum of mold material. Mold Design. Both the mold thickness and the design of the supporting ribs affect the degree of mold warpage at operating temperatures. Supporting ribs on the back of a thin mold will warp the mold face into a concave form. This mold design error can alter casting dimensions across the parting line by as much as 1.6 mm (0.06 in.). Adequate mold lockup will contribute to the control of otherwise severe warpage problems. Mold erosion resulting from metal impingement and cavitation due to improper gating design contributes to rapid weakening of the mold metal and to heat checking. These mold design errors contribute to rapid mold deterioration and resulting dimensional variation during a long run. Mechanical abrasion due to insufficient draft or to improperly designed ejection systems also contributes to the rapid variation of casting dimensions. Sliding mold segments require clearance of up to 0.38 mm (0.015 in.) to function under varying mold temperatures. This clearance and other mechanical problems associated with sliding mold segments contribute to variations in casting dimensions. Sand cores further aggravate the problem. Mold Operation. Metal buildup from flash can prevent the mold halves from coming together and can cause wide variations in dimensions across the parting line, even in a short run. Mold coatings on the cavity face are normally applied in thicknesses from 0.076 to 0.15 mm (0.003 to 0.006 in.). Poor mold maintenance can allow these coatings to build to more than 1.5 mm (0.060 in.) thick, causing extreme variation in casting dimensions. Inadequate lubrication of sliding mold segments and ejector mechanisms will contribute to improper mold lockup and consequent variation in casting dimensions. Variation in the casting cycle

Permanent Mold Casting / 699 Standards for aluminum permanent mold castings E1 E1

E2

E2

P/L

Slide core

P/L

E2 E2

E3

E3

E1

(E1) Tolerance

(E2) Additional tolerance

Fully machined mold cavites

Cast to size molds

Permanent mold

mm

in.

Basic tolerance up to 25 mm (1 in.)

±0.38

±0.015

Additional tolerance for each additional 25 mm (1 in.)

±0.050

Fig. 14

mm ±0.8

in. ±0.030

±0.002 ± 0.080 ±0.003

Permanent mold

Projected area 2

cm Up to 65 65_320 320_650 650_1600 1600_3200

(E3) Additional tolerance

2

Permanent mold to metal core

Projected area 2

cm

in. Up to 10 10_50 50_100

mm ±0.25 ±0.38

in. ±0.010 ±0.015

±0.50

±0.020

100_250 250_500

±0.6

±0.025

Up to 65 65_320 320_650 650_1600

±0.8

±0.030

1600_6500

2

in.

mm

Up to 10 10_50 50_100 100_250

±0.25 ±0.38 ±0.38

in. ±0.010

±0.6

±0.022

250_1000

±0.8

±0.032

±0.015 ±0.015

Recommended dimensional standards for permanent mold castings. Source: Ref 10

and in metal temperature will contribute to dimensional variations. Wear Rates. The dimensions of many mold and core components change at a relatively uniform rate; therefore, it is possible to estimate when rework or replacement will be required. To maintain castings within tolerances, it is sometimes necessary to select mold component materials on the basis of their wear resistance; for example, H13 tool steel has better wear resistance than mild steel or cast iron.

Surface Finish The surface finish on permanent mold castings depends mainly on:  Surface of the mold cavities: The surface fin-

ish of the casting will be no better than that of the mold cavity. Heat checks and other imperfections will be reproduced on the casting surface.  Mold coating: Excessively thick coatings, uneven coatings, or flaked coatings will degrade the casting finish.  Mold design: Enough draft must be provided to prevent galling or cracking of casting surfaces. The location of the parting line can also affect the surface finish of the casting.

 Gating design and size: These factors have

Portions of this article are adapted from “Permanent Mold Casting,” revised by Charles E. West and Thomas E. Grubach, in Casting, Volume 15, ASM Handbook, ASM International, 1988, p 275–285.

3. ASM International, www.asminternational. org 4. American Foundry Society, www.afsinc. org 5. SAE International, www.sae.org 6. Aluminum Association, www.aluminum. org 7. Copper Development Association, www. copper.org 8. R. Fuoco and E. Correˆa, “The Effect of Gating System Design on the Quality of Aluminum Gravity Castings,” IPT-Technological Research Institute, Sao Paulo, Brazil 9. F. Chiesa, Controlling Permanent Mold Coating Application Parameters, Mod. Cast., Oct 1996 10. Standards for Aluminum Permanent Mold Casting, 14th ed., Aluminum Association, March 2000

REFERENCES

SELECTED REFERENCE

a marked effect on casting finish because of the influence on the rate and smoothness of molten metal flow.  Venting: The removal of air trapped in mold cavities is important to ensure smooth and complete filling.  Casting design: Surface finish is adversely affected by severe changes of section, complexity, requirements for change in direction of metal flow, and large, flat areas.

ACKNOWLEDGMENT

1. M. Sadayappan, J. Thomson, F. Fasoyinu, and M. Sahoo, Establishing Process, Design Parameters for Permanent Mold Cast LeadFree Copper Alloys, Mod. Cast., Feb 2002 2. Aluminum Casting Technology, 2nd ed., AFS Inc., 1993

 J.G. Kaufman and E.L. Rooy, Aluminum

Alloy Casting: Properties, Processes, and Application, ASM International and AFS, 2004

ASM Handbook, Volume 15: Casting ASM Handbook Committee, p 700-708 DOI: 10.1361/asmhba0005261

Copyright © 2008 ASM International® All rights reserved. www.asminternational.org

Low-Pressure Metal Casting Gregory G. Woycik, CMI Equipment and Engineering, Inc. Gordon Peters, Intermet-PCPC

LOW-PRESSURE METAL CASTING has been used since the 1950s to produce highvolume, high-integrity castings in alloys ranging from aluminum to zinc. In recent history, aluminum has become the most widely used alloy in this process (Ref 1), due in large part to the increased popularity of aluminum among automotive designers who desire lightweight alternatives for today’s vehicles. Low-pressure casting, as it is most widely practiced, is a variation of permanent mold casting (Ref 2) (although it is used to a limited degree to fill sand or plaster molds, too); thus, it is frequently called low-pressure permanent mold. It may also be referred to as low-pressure die casting, perhaps deriving from the European practice of referring to all metal mold processes as die casting (low-pressure die casting, gravity die casting, high-pressure die casting, etc.). Low-pressure casting is a process where molten metal is introduced to the mold by the application of pressure to a hermetically-sealed metal bath, forcing the molten metal up through a narrow-diameter fill (stalk) tube from a furnace usually residing below the casting machine (although there is a version using electromagnetic forces to lift metal into the mold, and then the furnace may be an open hearth located beside the casting machine). The process can be considered for low to high volumes of castings from 5 to 100 kg (11 to 220 lb) and usually incorporates the use of iron or steel permanent molds. Recent developments in sand molding technology have made precision sand molds a viable choice for high-volume lowpressure casting as well. A wide range of casting core options, such as expendable sand and shell cores and mechanical single- or multipiece permanent cores, are successfully used in the low-pressure process. The large majority of casting produced in low pressure are aluminum; however iron-, magnesium-, and copper-base alloys are successfully cast with this process also. The mechanical and fatigue properties of low-pressure cast components are typically 5 to 6% greater than those of gravity cast components of a similar alloy. Controlled application of the pressure yields a smooth, nonturbulent cavity fill from the bottom to the top,

which is highly desirable when casting light alloys such as aluminum and magnesium. The single-point entry of metal to the casting and minimal or nonexistent risers minimizes subsequent processing to remove these items, reducing overall manufacturing costs. For these reasons, the vast majority of light-alloy road wheels (Fig. 1a, b) for the automotive sector are produced using low pressure. The process is also well adapted for the production of high-integrity

Fig. 1

As-cast aluminum wheels low-pressure casting

produced

by

components such as cylinder blocks, cylinder heads (Fig. 2), pistons, automotive suspension components, brackets, and housings.

Conventional Low-Pressure Casting The low-pressure process incorporates the use of a hydraulically operated machine to vertically open and close a casting mold that is situated over a ready supply of molten metal. A basic schematic of a low-pressure machine is given in Fig. 3. The machine consists of an upper (moving) and lower (stationary) platen, to which the upper (cope) and lower (drag) mold halves are attached, respectively. The moving platen is raised and lowered hydraulically to allow for casting and extraction of a component. Standard mold design allows the casting to remain in the upper mold as it is opened and subsequently ejected from the mold for manual or automatic removal. The stationary platen is typically positioned directly over a sealed, airtight furnace with a supply of

Fig. 2

Low-pressure cast motorcycle cylinder head. Courtesy of Progress Casting, an ATEK company, Minneapolis, MN

Low-Pressure Metal Casting / 701

Fig. 3

Schematic of low-pressure casting machine with an electric resistance crucible furnace

Fig. 4

Low-pressure wheel casting cell with two casting machines and robotic casting extraction/quench. Courtesy of CMI Equipment and Engineering, Inc., AuGres, MI

molten metal. Figure 4 shows a typical low-pressure wheel casting cell comprised of two low-pressure machines and automated casting extraction robot. A fill/stalk tube is immersed in the molten metal through the top of the sealed furnace and brought in contact with the lower mold by either lowering the machine platen to the furnace or by raising the furnace to meet the platen. In former times, the fill/stalk tubes were typically constructed of steel or cast iron to which a protective refractory coating was applied to prevent deterioration in the molten metal. High-strength, nonwetting refractory materials are now the norm in most lowpressure applications, and, although initially more expensive, they typically last many weeks instead of days and are thus very cost-effective. An intermediate tube is sometimes used between the furnace fill tube and the lower mold when there may be interference of the machine platen and the top of the furnace. When an intermediate tube is used, it should be kept as short as possible, and it is imperative to incorporate some external heating to avoid metal freezing in the tube. Fiber gaskets are normally used between any junctions in the stalk tube, intermediate tube, and lower mold to maintain an airtight, leakproof seal. The fill tube extends below the metal surface between 305 and 760 mm (12 and 30 in.), depending on the style, size, and depth of the casting furnace. This gives the low-pressure process a distinct advantage in that the surface of the molten bath remains undisturbed during the casting process, thus avoiding the entrainment of surface oxides into the melt, as occurs when ladles or dippers are used to deliver metal to a casting mold. The entrainment of oxides in the melt has been routinely shown to reduce the mechanical properties of aluminum castings. For a complete discussion of this topic, see Ref 3 as well as the articles “Casting Practice: Guidelines for Effective Production of Reliable Castings” and “Filling and Feeding System Concepts” in this Volume. To provide optimum performance and consistency to the process, almost all low-pressure machines are controlled by a programmable logistic controller (PLC). The PLC will control all moving functions of the casting machine, pressurization of the casting furnace, mold cooling/heating circuits (oil, water, or air), automatic casting extraction, and manipulation of casting furnace for refilling. The PLC can also track important historical data of the casting cycles, such as mold temperature, cycle time, metal temperature, pressure profiles, and cooling cycles, which may be used to gage productivity or troubleshoot casting problems. Most low-pressure machines include automatic casting extraction due to the size and number of castings produced by the process. A casting extractor can be a simple shuttle pan onto which the casting is ejected or a robotic arm that extracts the casting and takes it to a subsequent operation.

702 / Permanent Mold and Semipermanent Mold Processes A machine operator is required to begin the casting cycle, monitor the machine performance, visually inspect castings, and make process adjustments as necessary. The operator will also inspect the mold between cycles to verify that it is ready for the next casting cycle. The sequence of operation begins with a properly coated and preheated mold in the “open” position. The lower platen is situated above the casting furnace, with the fill tube sealed in position and extending down into a full bath of molten metal. When the cycle is initiated, the machine moves the upper mold down to seal the mold envelope with the lower half. Guide pins in the mold are typically used to accurately position the two mold halves as they are brought together. Vertical position sensors confirm that the mold is indeed closed, at which point pressure in the furnace begins to increase along a preprogrammed pressure or filling profile. The filling profile is established on a partby-part basis to provide a smooth, constant metal filling front as the metal travels up the fill tube and through the mold. Complex molds may require several different “ramps” in the profile to account for changes in the crosssectional area of the mold cavity as the metal fills. Innovations in predictive pressure control can now provide very accurate and repeatable control of a filling profile, which is paramount to casting quality. The pressure is programmed to reach a maximum at the point where the mold is completely full. The maximum pressure is dictated by the casting requirements and is limited to the capability of the furnace seals but seldom exceeds 150 kPa (1.5 bar) in most applications. The pressure is held during casting solidification to keep the mold full and to provide feed metal to the in-gate or sprue. Solidification under pressure also maintains intimate contact of the casting to the mold wall, promoting heat transfer and rapid solidification. Rapid solidification results in a finer microstructure, which increases the mechanical properties of the casting. Hydrostatic pressure through the molten metal also aids in minimizing the occurrence of visible gas porosity in the casting. This effect is only realized until the sprue freezes and pressure is no longer transferred through the molten metal. As solidification of the casting progresses to the sprue, the pressure in the furnace is released, allowing the metal in the fill tube to drop back into the furnace. It is important that the casting does not solidify below the sprue choke point and become “die locked” in the mold. Figure 5 shows the sprue area of a lowpressure mold and the choke point in the sprue bushing. This condition often makes the casting impossible to remove from the mold and causes significant interruption to the casting process. Often, the tooling must be removed from the machine to extract the casting by drilling out the frozen sprue area. Some processes only reduce the pressure enough to allow the liquid metal front to drop

away from the solidifying sprue enough to avoid die lock but keep the majority of the fill tube full. This prevents potential oxide buildup on the internal surface of the fill tube as the metal drops back into the furnace, which could be carried into subsequent castings. Many lowpressure molds incorporate the use of a disposable filter screen that is placed just above the sprue before each shot to prevent large oxide skins from being carried into the casting during the fill. Typical low-pressure aluminum permanent molds incorporate the use of a replaceable sprue bushing that is made from an insulating ceramic material, which aids in retaining heat in the critical in-gate of the casting. If a casting does freeze up in the sprue bushing, it can be knocked out and replaced, usually without removing the tooling from the machine, thus reducing any production downtime for this problem. After a predetermined time to allow the casting to cool sufficiently, the machine will open the mold. By design, the casting remains in the upper mold half and travels up with the mold. The casting may be allowed to cool further or may be immediately ejected from the mold by actuating the ejection cylinder, which pushes ejector pins through the mold, thus pushing the casting out to a waiting operator or automatic unloading device. The casting is removed from the area to a cooling line or quench bath. The operator checks for any flash that remains in or around the mold and removes it, after which the cycle is ready to begin again. Total cycle time for the low-pressure process is dependent on the number and size of castings being made as well as the mold cooling methods employed but typically ranges from 2 to 7 min for aluminum castings. Low-Pressure Furnace. In all cases, the molten metal supply to the low-pressure casting process is from below the machine/mold. This requires a furnace of sufficient size to produce many castings, yet small enough to fit under the casting machine. Aluminum furnace capacities can range from 450 to 1500 kg (990 to 3300 lb), depending on the style of furnace and size of casting machine. Furnaces can be gas fired or electric resistance crucible style, where the crucibles are mainly silicon carbide. In magnesium low-pressure casting, steel crucibles are commonly used. Another common low-pressure furnace is a refractory-lined furnace heated by electric radiant (glow bar) or electric resistance (immersion) heaters. Immersion-style heaters have almost twice the average efficiency of radiant glow bar heaters and cost less to operate over time. Crucible-style furnaces will typically support a single fill tube, while refractory-lined furnaces can accommodate 1 to 5 fill tubes. Figure 6 shows a refractory-lined immersion-style low-pressure casting furnace. Regardless of furnace style, each must be refilled periodically as metal is used up. A critical feature in maintaining a successful

Fig. 5

Sprue and sprue bushing area of a low-pressure casting mold. Castings must not solidify below the choke point, to avoid a die-locked condition.

Fig.

6

Refractory-lined immersion-heated pressure furnace. Courtesy of Equipment and Engineering, Inc., AuGres, MI

lowCMI

production cycle is the ability to refill the casting furnace with no or minimal interruption to the casting cycle. Refractory-lined furnaces are equipped with a fill door through which freshly treated and degassed molten metal can be added. A refractory-lined filling trough is typically used to aid in filling the furnace through the fill door (Fig. 7). The refill occurs after the pressure has been released during a casting cycle, while the casting is cooling in the mold. It must be completed before another pressure cycle may begin, or the furnace will not build pressure to make another casting. Crucible furnaces may be filled in a similar fashion, or, in some cases, the entire furnace is swapped out for another with a full supply of fresh metal. As the metal in the casting furnace is depleted, both the volume of space that must be pressurized and the height that the metal must travel to reach the mold increase. These two factors must be considered to assure that the casting mold is filled consistently every time. Through PLC programming, a compensation factor can be added to the filling profile to assure that

Low-Pressure Metal Casting / 703

Fig. 7

Fig. 8

Fill trough for replenishing molten metal in a low-pressure furnace. Courtesy of CMI Equipment and Engineering, Inc., Augres, MI

Low-pressure wheel mold with four moving side cores to create the external geometry of a wheel

the metal front reaches the mold at the same time. This compensation factor can be determined in various ways and can be based on calculated metal volume decrease per shot, weight of remaining metal in the furnace, or metal level in the furnace, among others. Direct feedback from a load cell or metal-level sensor can provide the best information on which to determine a compensation factor but is also the most costly. Pressurization of the casting furnace is typically achieved by using dry compressed air. The air must be substantially free of oil and moisture, because these elements will lead to

contamination of the metal bath through excessive oxidation and hydrogen pickup over time as the moisture is vaporized above the molten bath. Dry air is acceptable for most low-pressure casting applications; however, for some high-integrity components, an inert gas such as nitrogen or argon is substituted for dry air, either intermittently or completely, to avoid the problems described previously. Low-Pressure Tooling. The molds used in low-pressure casting can range from hard sand molds to lost foam flasks but are most typically constructed of tool steel or cast iron. In applications where only forced-air cooling is used, iron molds can be used very effectively and at a lower cost than steel. To take full advantage of the enhanced mechanical properties capable in the low-pressure process, water-cooled steel molds are used. Strategically placed internal water cooling channels can be added to steel molds to promote rapid directional solidification of the casting from the extremities back to the sprue. This technique results in a finer casting microstructure and increased mechanical properties as well as reduced castingto-casting cycle time. Low-pressure casting machines can be fitted with computercontrolled water (and air) cooling solenoids that allow customization of casting solidification to eliminate shrinkage defects. To achieve the maximum quality of cast components in low pressure, care must be exercised when designing low-pressure tooling. Since each casting is typically filled through a single entry point, and external risers are not normally used, the in- gate or sprue of the casting must be positioned so that good directional solidification back to the sprue is achieved. This may be enhanced through the use of water or air cooling, as described previously. In the absence of good directional solidification, castings may yield significant unwanted shrinkage

porosity. Successfully designed low-pressure castings have some of the best casting-toshot-weight yields of any process, due to the lack of risers and the single sprue feeding. Casting yields over 90% are typically achieved in aluminum castings. It is important to note that because the casting fills from the bottom up, the metal front must force the air out of the cavity through properly located vents. If good venting practices are not followed, any air trapped in the mold can accumulate in localized areas, resulting in unfilled features of the casting. Standard practice incorporates the use of vented ejector pins, because they are typically located in the upper mold and on the uppermost surfaces of the casting, where air is most likely to accumulate. Ejector pin vents are typically self-cleaning of any flash; however, if excessive soldering buildup occurs, the venting capability may be diminished. In situations where ejector pins are not feasible or where casting geometry may cause air to be trapped, several styles of in-mold vents can be used to exhaust the cavity. It is important to mask in-mold vents when applying mold coatings so that they are not coated over, rendering them useless. Vents that become plugged during casting must be cleaned or replaced to maintain their effectiveness. Proper vent cleaning/replacement should be exercised during every mold maintenance interval. As is the case for gravity permanent molds, low-pressure permanent molds for aluminum castings made from iron or steel must be coated with a refractory coating to prevent soldering of the aluminum to the iron-base mold. There are many commercially available permanent mold coatings with custom-tailored thermal properties ranging from highly insulating to chill coats. The proper application of mold coating is critical to the longevity and performance of the coating, and strict adherence to manufacturer’s recommended application techniques must be exercised. Cores. The low-pressure process can readily accommodate the use of almost any core technology. Hydraulically operated mechanical cores are easily integrated into the PLC control logic of the machine and automatically controlled. These cores can be used to create internal passages in castings or, as in the case of wheels, used to create external geometry of the casting. Figure 8 shows a typical low-pressure wheel mold installed in a low-pressure machine. Four side cores make up the external geometry of a wheel. The low-pressure process can also accommodate semipermanent molds, where cores can be expendable sand, shell, or plaster. The incoming metal velocity in low pressure is slow and does not typically displace or erode these types of cores. Automatic core- setting equipment can be easily integrated into the casting process. Air-set precision sand has been widely used to make cores; however, it is possible to create entire casting molds as well. With some

704 / Permanent Mold and Semipermanent Mold Processes adaptations to the low-pressure machine, these types of molds may be positioned over a lowpressure furnace and fill tube and filled in much the same manner as iron or steel molds. With this approach, the sand casting will benefit from smooth, controllable bottom-up filling, the trademark of the low-pressure casting process.

Counterpressure Casting The counterpressure casting process is an evolution/extension of the basic low-pressure die casting process. The equipment used in the counterpressure casting process is similar to a low-pressure die casting machine and is essentially a vertical press with a crucible furnace mounted under the press. A unique feature of the PCPC (Intermet Corp.) process is that the

horizontally parted die is surrounded by an airtight mold chamber. The basic pressure counterpressure casting (PCPC) machine is shown in Fig. 9. The PCPC process is capable of producing high-quality, cost- effective aluminum castings if proper attention is paid to critical aspects of the process. Aluminum castings that are to be used in safety-critical automotive applications or military applications are candidates for the PCPC process (Ref 4-7). Counterpressure casting has a long-established history of process development and practical implementation. Due to the high integrity of the castings this process can produce, counterpressure castings have replaced aluminum forgings and squeeze castings, with significant savings to the customer. Ductile iron parts have been replaced at a 40 to 50% weight savings, a good value to the customer. History of Counterpressure Casting. Counterpressure casting evolved out of research conducted at the Bulgarian Academy of Sciences in the 1950s. In 1961, Angel Tontcheve Balevsky and Ivan Dimov Nikolov were awarded a patent for a method for casting under pressure. This patent described the use of pressure acting inside a mold cavity and was termed counterpressure. The process was then named counterpressure casting. In 1964, the first vertical semiproduction counterpressure casting machine was constructed for casting aluminum alloys. To date (2008), counterpressure casting machines are producing millions of castings per year. Automotive safety-critical castings, such as wheels, front steering knuckles, control arms, and rear knuckles, are being produced. Power train components such as engine blocks and cylinder heads have also been produced. Military applications such as tank track wheels have also had counterpressure castings replace more expensive aluminum forgings. Pressure Differential (DP) and Mold Filling. In standard low-pressure casting, the height of the metal above the metal bath is calculated by: H ¼ p=g

Fig. 9

Basic pressure Components

counterpressure

Machine

where H is the height of the metal column above the furnace metal level, g is the specific mass of the alloy melt, and p is the pressure applied above the alloy melt. Since there are two pressures acting in counterpressure casting (Fig. 10), the height of the alloy melt is given by: H ¼ ðp2  p1 Þ=g

Fig. 10

Counterpressure casting method

where H is the height of the metal column above the furnace metal level, g is the specific mass of the alloy melt, p1 is the pressure applied to the cavity in the mold chamber, and p2 is the pressure applied above the alloy melt in the furnace. Since DP is defined at p2  p1, the formula can be simplified to resemble the calculation for low pressure by substituting DP for p2  p1.

Thus, the formula for calculating the height of the metal in the fill tubes over the metal bath is: H ¼
p=g

For example, if the density of the aluminum alloy is 2550 kg/m3 and the DP is 300 mbar, then: H ¼ 0:30bar=2550 kg=m3 H ¼ 30 kPa=2550 kg=m3 H ¼ 30; 000 Pa=2550 kg=m3

Since 1 kg = 9.81 N, then: 30; 000 Pa ¼ 3058:1 kg=m2 H ¼ 3058:1 kg=m2 =2550 kg=m3 H ¼ 1:20 m

By knowing the diameter of the fill tube and thus its area, this can be compared to the area of the mold in the x- y plane at any z height. The correct fill profile for the part being cast can be established as a result of using the conservation of mass flow principle. Counterpressure Casting Cycle Steps. The following describes the basic counterpressure casting cycle steps (Fig. 11): 1. The die closes to initiate the casting cycle. The mold chamber is set up to seal at the same time as the die halves. 2. The furnace is pressurized at the same rate as the mold chamber. 3. Once the furnace pressure and the air inside the mold chamber reach the pressure set point, the pressure holds for an equalization period. This period is approximately 5 s. After the equalization is over, the pressure in the furnace increases. The pressure in the chamber remains constant as it is bled off when the metal rises up the riser tubes and compresses the air displaced from the tubes and die cavities. This pressure difference is called the delta pressure. During this phase, the delta pressure is slowly increasing to permit the die to be filled. The delta pressure curve should resemble the fill profile of a typical low-pressure fill profile. The metalfilling characteristics are controlled by the delta pressure profile programmed into the casting machine. 4. Once the die is filled, the delta pressure is held constant for a period of time to stabilize. This is the delta P-maintain time programmed into the casting machine (usually 5 s). After the delta P-maintain time expires, the pressure in the chamber is released, and the delta pressure increases by the value of the system pressure setting. 5. The pressure is maintained on the furnace (feeding time) until the sprue has solidified. Once the sprue has solidified, the furnace pressure is released, and the casting cools until it has hardened enough to withstand the forces applied to it during demolding

Low-Pressure Metal Casting / 705

Fig. 11

Counterpressure casting process steps

and ejection from the upper die half. The die then opens, and the casting is ejected during the mold-open phase of the casting cycle. The casting can now be processed through the remaining manufacturing process, such as degating, deflashing, inspection, heat treatment, and packaging. Advantages of Counterpressure Casting. The advantages of the PCPC process are as follows:  High yield rate: This translates to lower

operating costs, reduced melt loss, and less material handling. Yields are similar to low pressure or vacuum riserless/pressure riserless casting and are typically approximately 85 to 90%. On larger castings, yields above 90% are possible.  Good fill control: Fill is controlled by programmed pressure curves, and precise fill dynamics can be achieved.  Good mechanical properties: Properties are typically better than sand and gravity

permanent mold and low pressure, but PCP is capable of matching the properties of the more expensive squeeze casting or semisolid/thixocasting processes. The net result of these advantages translates into very high- performance castings at relatively low costs. The good mechanical properties of PCPC castings are attributed to:  Filling characteristics: Since the casting

process is conducted with a working pressure greater than atmospheric pressure, the metal is less prone to turbulence during the filling process. This reduces the potential for oxide films to be generated that will reduce the mechanical properties of the casting.  Feeding characteristics: This includes the increased feeding pressure that the crucible furnace can generate and the ability to create a near-impulse increase in net feeding pressure when the pressure in the mold chamber

is released. By controlling how much pressure is applied to the casting during solidification, areas prone to solidification shrinkage can have improved internal integrity. This results in less microporosity from solidification shrinkage, and thus, mechanical properties are improved.  Mold/casting heat-transfer efficiency: Due to the increased pressure imparted on the casting from the furnace, the casting is kept in contact with the mold surface with greater force, and the heat from the casting can be extracted from the casting with greater efficiency. This results in smaller secondary dendritic arm spacing and thus better mechanical properties.  Melt cleanliness: Since the crucible furnaces are completely exchanged once the metal is consumed during the casting process, each new crucible of metal has the opportunity to be fluxed, alloyed, degassed, and thoroughly cleaned. This is a big advantage in metal quality compared to other low-pressure processes, because the casting machine

706 / Permanent Mold and Semipermanent Mold Processes furnaces are typically continually topped off, and the whole furnace is not cleaned unless the machine is taken out of production. Effective degassing of a fixed furnace in continuous operation is essentially impossible, a because gas is introduced each time it is topped off. This increased control on metal preparation contributes to improved mechanical properties.  Low levels of gas porosity: Since the solidification of the casting is conducted at pressures greater than atmospheric as a result of the counterpressure applied to the mold chamber, any dissolved hydrogen will stay in solution longer and will be less likely to evolve out and create microporosity in the casting. This results in less gas microporosity and better mechanical properties. Special considerations for PCP operations include preventative maintenance, die access, and die coating considerations. Preventative Maintenance. Since the control of the filling profile depends on two air systems working in harmony, any problems with sealing in either of the systems can result in uncontrolled mold filing, casting problems, machine damage, and potential operator injury. There are several rubber seals on counterpressure casting machines that must be replaced on a periodic basis. This must be done and the schedule rigidly adhered to. If this is not done, the casting machine operating equipment efficiency will suffer. This results in slightly increased repair and maintenance costs when compared to low- pressure casting. Die Access. Since there is a steel chamber surrounding the die, access to the mold is restricted. The mold cannot be seen during the mold-closed phases of the casting cycle. In order to detect if there is a die spill (metal that flows between the parting line of the die halves and onto the casting machine base), the operator must rely on a spill detection wire to signal the machine to depressurize when molten aluminum hits it. If this wire is set up incorrectly or not installed, the machine can completely fill up the mold chamber with molten aluminum, which will spray out of the mold chamber once the rubber seal is burned up. The upper die access is limited to the die cavity faces. Access to the outside of the upper die is only possible if the chamber is removed. This can result in longer die change times and make die repairs on the machine more laborintensive. The mold chamber also significantly reduces the effect of radiation heat transfer and convective heat transfer during the casting cycle. Thus, the PCPC casting process must rely on effective water or air cooling circuits in the die. Die Coating Considerations. Like gravity permanent mold and low-pressure permanent mold casting, the counterpressure casting process relies on die coatings. These are applied to the cavity surfaces to promote complete filling of the die cavity, to control solidification,

Fig. 12

Example of vacuum riserless/pressure riserless cast automotive chassis and suspension components

and to provide lubrication for demolding the casting from the die cavity. Due to the increased metal pressure exerted on the casting during the solidification phase of the casting cycle, metal will infiltrate into the pores of the die coating and increase abrasion of the coating. This demands that excellent die coating procedures be in place in a counterpressure casting operation. If the die coating practices are neglected, casting performance will suffer due to premature mold coating wear.

Vacuum Riserless/Pressure Riserless Casting The vacuum riserless/pressure riserless casting (VRC/PRC) process for casting aluminum is a modified permanent mold casting method that incorporates the use of vacuum and pressure to bottom fill multiple mold cavities from a hermetically sealed molten-metal source. The use of vacuum in the production of permanent mold castings dates back to World War II for the production of aircraft cylinder heads and was further developed by Alcoa throughout the 1940s to the 1960s with the addition of pressurized filling, expendable cores, and watercooled molds (Ref 8). In the late 1990s, the process was reengineered by CMI Equipment and Engineering for use as a high- volume casting process for automotive components, with manufacturing costs ranging between standard permanent mold and squeeze casting costs.

The VRC/PRC castings are readily heat treatable, and mechanical properties approach those of squeeze castings. Products such as safety-critical automotive subframes, steering knuckles, control arms, and brackets are produced with the VRC/PRC process (Fig. 12) A schematic of a typical VRC/PRC casting machine is shown in Fig. 13. The equipment consists of a hydraulically operated machine that houses the mold package in order to open and close the mold as well as eject the castings. Figure 14 shows a typical production machine. The machine is situated over a molten-metal casting furnace that can be an electrically heated silicon carbide crucible furnace or a refractory-lined electric immersion-heated furnace. Immersion-heated furnaces (Fig. 6) provide significant efficiency advantages over the crucible-style furnaces and have become the furnace of choice in new installations. A key feature of immersion furnaces is that the lower platen/mold combination can act as the furnace lid, providing a hermetically sealed chamber to which pressure may be applied to fill the mold cavities. Multiple fill tubes are connected to the lower mold or molds, through which metal is introduced. By design, the entire surface of the molten-metal bath is accessible, providing significant flexibility in positioning molds and fill tubes to maximize productivity. A “family” of individual molds, each with a single fill tube, can rest on a base plate and be filled simultaneously. In the same fashion, a large singlemold cavity incorporating multiple fill tubes

Low-Pressure Metal Casting / 707

Fig. 13

Schematic of vacuum riserless/pressure riserless casting machine

Fig. 14

Vacuum riserless/pressure riserless production casting machine. Courtesy of CMI Equipment and Engineering, Inc., AuGres, MI

may be used. If expendable cores are used, they are placed in the mold either manually or by an automated mechanism. When a casting cycle is initiated, the machine closes the mold, providing a critical airtight seal around the parting line of the mold. A controlled vacuum is applied through engineered passages (vacuum track) and vents in the upper mold half to evacuate air and other contaminating atmosphere from the cavity. As a result of the vacuum, molten metal begins to rise into the cavities. The amount of vacuum is dependent on the number and size of castings being filled but is limited to only what is necessary to fill the cavities. Pressure is then applied to the molten-metal bath to maintain the metal

height in the fill tube and cavity, providing feed metal until solidification proceeds to the fill tube. Both vacuum and pressure are PLC controlled and follow a preprogrammed profile to achieve shot-to-shot consistency. Another significant feature of the process is that the fill tubes protrude only a few inches below the metal surface. Close coupling of the fill tubes to the molten metal allows lower metal temperatures, minimizes oxide formation, and provides maximum feeding of the casting. As a result of the minimum fill tube depth, a unique method of replenishing the casting furnace is employed so that a nearly constant metal level in the furnace is maintained. Fresh metal is transferred to a holding furnace via a filtered launder system from a reverberatory melting furnace, where it is thoroughly degassed before being dosed to the casting furnace after every shot. The solidification of the castings in the VRC/ PRC process must proceed from the extremities back to the feed metal in the fill tubes to provide sound, shrink-free castings. To facilitate rapid directional solidification, strategically engineered and sequenced water cooling is used. Up to 150 electrically operated and PLC-programmable water solenoid valves are incorporated on the largest machines, providing optimum thermal control of the casting process. In addition to minimizing cycle time, the rapid solidification results in a finer microstructure, which improves the mechanical properties of the castings. Automotive subframes manufactured using the VRC/PRC process, A356-T6, exhibit tensile strength of 310 MPa (45 ksi), yield strength of 225 MPa (33 ksi), and elongation of 13%. The extensive thermal control of the mold also provides repeatable process temperatures throughout the casting cycle, minimizing process variation during production. After the castings have solidified, the machine opens the top mold. By design, the castings remain in the upper mold and are subsequently ejected onto an automated extraction

mechanism. The castings are removed to a quench station, and the machine begins another cycle. One operator is necessary to prep the mold between casting cycles by checking for and removing any flash or debris remaining from the previous shot, visually inspect castings, and touch up any areas where mold coating may be compromised. The VRC/PRC process is a hybrid of the lowpressure process and as such has similar casting yields of 85 to 95%. Production rates for this process are dependent on casting size, complexity, and weight but typically range between 100,000 and 600,000 castings per year. Casting weights produced by VRC/PRC typically range between 3 and 25 kg (7 and 55 lb); however, these should not be considered as the limits of the process. High-volume-production VRC/ PRC machines can be built to accommodate molds as large as 132 by 168 cm (52 by 66 in.). Smaller machines will accommodate 107 by 130 cm (42 by 51 in.) molds (Ref 9). In a typical installation, the casting furnace resides in a shallow pit below the VRC/PRC machine. The entire casting machine can be raised off the furnace and shuttled forward to gain access to the metal bath and mechanical components of the furnace for cleaning and service. The casting furnace is typically cleaned at least once per shift to avoid dross buildup at the air-metal interface and near the filling port. The VRC/PRC machines are equipped with a tiltout upper platen capable of carrying the weight of the entire mold. This feature aids in loading molds into the machine by allowing overhead crane access to the platen. The tooling can be secured to the upper platen and subsequently tilted up to the locked position. The upper platen is then lowered to secure the bottom (drag) mold to the fixed platen. Supporting mold straps that hold the mold together are removed, and the machine is ready to start production. Due to the large number of water cooling lines employed, a quick-disconnect manifold is standard for the upper and lower molds to speed the process of mold changes. This also helps avoid improperly connected water lines, which can cause significant problems when relying on proper directional solidification to produce high-integrity castings.

ACKNOWLEDGMENTS Gordon Peters thanks Thomas Prucha of AFS and George Roussev of CPC for their assistance in supplying supporting material on counter pressure casting. REFERENCES 1. J.L. Jorstad and W.M. Rasmussen, Aluminum Casting Technology, 2nd ed., The American Foundry Society, 1993, p 181 2. T.E. Prucha, Metal Mold Processes — Gravity and Low Pressure Technology, International Conference on Structural Aluminum

708 / Permanent Mold and Semipermanent Mold Processes Castings, Proceedings (Orlando, FL), The American Foundry Society, Nov 2003 3. J. Campbell, Castings, 2nd ed., Elsevier Butterworth-Heinemann, 2003 4. T. Prucha, Recent Advancements for the design and Manufacturing of Reliable Structural Aluminum Castings, AFS International Conference on Structural Castings, (Orlando, FL), 2003 5. Y. Arsov, E. Momchilov, K. Daskalov, G. Bachvarov, The Theoretical and

Technological Fundamentals of Gas CounterPressure Casting. Prof. Marin Drinov Academic Publishing House, Sofia, Bulgaria, 2007 6. Wurker, Lars, Aluminum Chassis Parts Produced by Counter Pressure Casting, Casting Plant and technology international, Vol. 1, 2006, p 38–49 7. G. Ruff, T. Prucha, J. Barry, D. Patterson, Pressure Counter Pressure Casting (PCPC) for Automotive Aluminium Structural Components, SAE 2001-01-0411

8. T.F. Arnold, “Vacuum Permanent Mold Casting,” Aluminum Company of America, Permanent Mold Castings Division 9. VRC/PRC Offers Squeeze Casting Properties at Permanent Mold Costs (Case History: Vacuum Riserless Casting and Pressure Riserless Casting), Mod. Cast., Vol 94 (No. 6), June 1, 2004, p 48(1)

ASM Handbook, Volume 15: Casting ASM Handbook Committee, p 709-711 DOI: 10.1361/asmhba0005265

Copyright © 2008 ASM International® All rights reserved. www.asminternational.org

Low-Pressure Countergravity Casting Paul Mikkola and Gary Scholl, Metal Casting Technology, Inc.

THE METHOD of countergravity mold filling has been practiced for many years. The technique is the result of applying a differential pressure between the molten metal and the mold to be filled, much like a soda straw is used to drink a liquid. During the last 30 years, the technology has matured, with many variations currently in practice. The differential pressure countergravity processes are used to produce aluminum and nonferrous alloys, low-alloy steels, stainless steel, high-alloy steels, and superalloys in weights ranging from a few grams to approximately 20 kg (44 lb). Production volumes range from 50 castings per year to 180,000 per day. The process is widely used for investment castings, precision sand molds made of resin or coldbox bonded sands, and semipermanent molds. The markets commonly served by these processes include automotive, hand tools, sports, medical, military, and aerospace. The general principle of the process involves placing a gas-permeable mold in an isolating chamber such that the sprue (inlet runner) extends downward from the bottom of the chamber. For investment casting, the molds are a common tree design with numerous part cavities attached around a central runner. The process starts with the mold and melt surface at the same pressure (typically 101 kPa, or 1 atm). The extended sprue is then positioned below the melt surface, and a differential pressure is created between the mold cavity in the chamber and the top of the melt. Usually, this is accomplished by applying a controlled vacuum to the mold chamber, but it can also be accomplished by applying a controlled pressure to the melt surface. This differential pressure causes metal to flow up the sprue and into the mold cavities. After the mold is completely filled, the liquid metal is held in position with this differential pressure while solidification takes place. For investment casting, typically only the castings and their in-gates are allowed to solidify before the differential pressure is removed, causing the molten metal in the central runner to be returned to the furnace for use in the next casting cycle. The process offers many advantages. Because metal is taken from below the center of the melt surface, slag that collects on the

melt surface (or migrates to the furnace lining) is not washed into the castings. The metal filling the mold is exposed to less atmosphere, and fill rates can be closely controlled. This enables routine filling of thin sections at lower metal temperatures because the liquid front encounters less resistance. There are also fewer opportunities for reoxidation and splashing, which cause entrainment of inclusions in the casting. Less superheat is required to attain fillout since the metal does not need the intermediate step of transferring into pouring ladles. This saves energy, eliminates ladle costs, and improves production throughput. The ability to cast at lower metal temperatures generally results in smaller grain sizes. By returning liquid metal from the central sprue, both energy and metal yield are improved over gravity pouring. These processes normally provide 70 to 90% metal yields, with only a small in-gate that needs to be removed. The process is accomplished in a chamber that is independent

Fig. 1

of the mold geometry, and mold filling is accomplished by computer-controlled differential pressure. Therefore, the casting process is easier to automate, which reduces labor and process variation. In the investment casting process, the single largest advantage is in sprue and mold loading. Because the central runner is removed before it solidifies, parts can be spaced more closely on a sprue, with no need to allow space for cutoff saws. This advantage can enable 10 to 15% more castings per sprue, which improves mold costs and throughput. Several production implementations that use differential pressure countergravity mold filling methods are discussed.

Mold Filling Countergravity Low-Pressure Air (CLA) Process. The CLA process (Fig. 1) has been used to produce investment castings in a wide

Schematic diagram of the countergravity low-pressure air process steps. (a) Preheated investment shell mold is placed into a casting chamber. (b) Chamber with mold is positioned and lowered into a molten-metal source. Vacuum is applied to the chamber, and mold filling is complete. (c) After casting and in-gates have solidified, vacuum is released, and all of the central sprue flows back into the melt.

710 / Permanent Mold and Semipermanent Mold Processes

Fig. 2

Schematic showing the steps of the countergravity low-pressure vacuum process. (a) Metal is melted in a vacuum chamber that is then flooded with argon. (b) A preheated mold is introduced into a separate upper chamber that is evacuated and then flooded with an equal pressure of argon before opening the separation valve. (c) The melt is raised, causing the fill pipe to be submerged in the metal, and then a controlled vacuum is applied to the upper chamber, causing filling of the mold cavities. (d) After solidification of the in-gates, the mold chamber is again flooded with argon, causing liquid metal to be returned to the furnace. The melt is lowered, and the separation valve is closed.

Fig. 3

Schematic showing, steps of the countergravity low-pressure inert atmosphere process. (a) Metal is melted in a vacuum or inert atmosphere, the chamber is filled with argon at +102 kPa (+1 atmt), and a hot mold is introduced into an independent mold chamber. (b) The mold chamber is then positioned above the melting furnace, and the fill pipe is lowered into the melt chamber but above the bath. Vacuum is applied to the mold chamber, purging the mold cavity with argon. (c) The fill pipe is then lowered into the molten bath, and additional vacuum is applied to the mold chamber, causing liquid metal to flow into the mold. (d) After a dwell time to allow the gates to solidify, the vacuum is released, and the remaining molten metal is returned to the furnace crucible.

variety of air melt alloys for over 30 years. These parts are supplied to a wide variety of industries. Applications for the automotive industry include steering system components, transmissions parts, diesel precombustion chambers, fuel injection components, rocker arms, and others. Parts are supplied to the aircraft and aerospace industries, such as temperature probes, fuel pump impellers, missile wings, brake parts, pump housings, and structural parts. Other applications include golf club heads, machine components, wood router tool

bits, tin snip blades, small wrenches, lock components, gun parts, valves and fittings, and power tools. Countergravity Low-Pressure Vacuum (CLV) Process. The CLV process (Fig. 2) is similar to the CLA process except that the mold and melt are contained in chambers that support both hard vacuum and argon atmospheres, to protect reactive alloys. The two chambers are joined by a slide valve that enables the downward-extending fill tube to contact the melt. For highly reactive alloys such as MAR-M-509

and Rene´ 125, which are noted for hafnium and zirconium inclusions, the CLV process consistently delivers parts with low levels of oxide inclusions. Castings produced by this process are used in gas turbine engines. Thin-wall components (0.5 mm, or 0.02 in.) previously made as weldments can be produced, thus enabling design freedom for shaped castings that maximize heat transfer, ease assembly, and reduce thermal fatigue. These design improvements often enable higher operating temperatures. Important cost and quality improvements have been achieved in cast airfoils, turbine seals, conduits, fuel swirls, and clamps. Countergravity Low-Pressure Inert Atmosphere (CLI) Process. The CLI process (Fig. 3) combines attributes of the CLA and CLV processes. Here, the melt and mold chambers are independent, enabling integration of the CLI melt chamber with a conventional casting environment. This allows reactive alloys to be cast using essentially the same casting machine that is used to produce high-volume steel castings. This process is used for high-volume automotive turbocharger wheels and vanes as well as gas turbine engine components for temperature probes and struts. Countergravity Pressure Vacuum (CPV) Process. The CPV process (Fig. 4) is similar to other countergravity reactive alloy processes except that the mold and melt are contained in chambers that remain under hard vacuum. Argon atmosphere is applied locally to the melt box, providing differential pressure for casting. The hard vacuum surrounding the melt box and mold chamber reduces the tendency for any oxygen in the outside room atmosphere to come into contact with the molten metal. The lack of any gas atmosphere in the mold during casting enhances fillout and eliminates the variable of mold permeability on casting fill. For the highest reactive alloys, the CPV process has demonstrated the lowest levels of oxide inclusions. Castings produced by this process are used in gas turbine engines. Wall thicknesses down to 0.5 mm (0.020 in.) have been successfully cast at metal temperatures only 28  C (50  F) over the solidus temperature of the alloy, thus providing a very fine, uniform grain structure. The higher detail, finer grain size, and reduced oxide inclusion count help to enable product designs that will operate at higher temperatures, improving efficiency of the gas turbine engine. Important cost and quality improvements have been achieved in cast airfoils, turbine seals, afterburner seals, and combustor liners. Supported Shell (Mold) Technique. The supported shell technique (Fig. 5) is an enhancement of the previously mentioned processes and enables use of thinner investment casting shell molds and sand molds. The products produced are the same as indicated previously. Loose Sand Vacuum (LSVAC) Process. The LSVAC process (Fig. 6) is used to produce thin-wall automotive components. Exhaust

Low-Pressure Countergravity Casting / 711

Fig. 5

Fig. 4

Schematic showing the steps of the countergravity pressure vacuum process. (a) Metal is melted in a vacuum chamber, and a hot mold is introduced into an isolated casting chamber. The mold fill pipe protrudes from the bottom of the support plate. The mold chamber is then sealed to the support plate, and both mold and casting chambers are evacuated. (b) An interlock between the melt and the casting chambers is opened, the melt is moved into the casting chamber, and the fill pipe is submerged in the melt. (c) Argon pressure is applied to the casting chamber, causing metal to fill the mold. The mold chamber may or may not be pressurized to achieve the desired differential pressure.

Schematic showing supported shell techniques. (a) The mold is placed in an empty chamber so that the fill pipe protrudes from the bottom. (b) Sand or other particulate media completely fills the void between the shell mold and the chamber wall. (c) A porous vacuum head/separator is placed atop the chamber, and casting operations proceed as described for the countergravity low-pressure air process in Fig. 1.

Fig. 7

Fig. 6

Schematic showing the loose sand vacuum process. (a) Several molds, produced using any of the bonded sand technologies, are placed atop a sheet of aluminum foil and protrude slightly below an open bottom tube, with the mold fill runners facing downward. (b) The region between the casting tube, the molds, and the foil is then filled with a particulate support medium, and a porous vacuum head/separator is placed at the top of the tube. (c) Vacuum is applied to the chamber through the vacuum head, and the assembly is moved to the melt furnace. The bottom of the mold assembly is submerged into the melt, thus causing metal to flow into the individual molds. (d) After the metal has solidified, the mold tube assembly is moved to the shakeout area, and vacuum is discontinued.

manifolds are produced in stainless steel. The process has been demonstrated to produce hollow camshafts in chill iron and nodes for space frame construction in low-alloy steels. Countergravity Centrifugal Casting (C3) Process. The C3 process (Fig. 7) makes use of the supported shell technique to hold the mold

in the casting chamber. Centrifugal force not only retains liquid metal in the part cavities, but after the central runner has been removed, it creates a strong, radial pressure gradient that enhances feeding and floats lighter oxides and gases to the vacant center. Unlike castings that are connected to a central runner and solidify

Schematic showing the steps of the countergravity centrifugal casting process. (a) A mold is placed in a chamber using the supported shell technique. (b) The mold is filled using the countergravity low-pressure air or countergravity lowpressure inert atmosphere casting technique. (c) The mold chamber begins to spin, and atmospheric pressure returns to the chamber, allowing liquid metal in the central sprue to return to the furnace but retaining liquid metal in the parts and in-gates by centrifugal force. Spinning continues until the parts and in-gates have completely solidified.

at different metalostatic pressures based on their height on the runner, each part cast using the C3 process solidifies under exactly the same pressure conditions. This promotes an exceptionally high level of uniformity when compared to parts produced with other casting processes. Many parts originally cast using the CLA and CLI supported shell processes are now cast using the C3 process.

ASM Handbook, Volume 15: Casting ASM Handbook Committee, p 715-718 DOI: 10.1361/asmhba0005266

Copyright © 2008 ASM International® All rights reserved. www.asminternational.org

High-Pressure Die Casting William A. Butler, General Motors (Retired), Consultant for NADCA

HIGH-PRESSURE DIE CASTING is characterized by the use of a source of hydraulic energy that imparts high velocity to molten metal to provide rapid filling of a metal die. The die absorbs the stresses of injection, dissipates the heat contained in the metal, and facilitates the removal of the shaped part in preparation for the next cycle. The hydraulic energy is provided by a system that permits control of actuator position, velocity, and acceleration to optimize flow and force on the metal as it fills the cavity and solidifies.

Die Casting Alloys and Processes The variety in die casting systems results from trade-offs in metal fluid flow, elimination of gas from the cavity, reactivity between the molten metal and the hydraulic system, and heat loss during injection. The process varieties have many features in common with regard to die mechanical design, thermal control, and actuation. Four principal alloy families are commonly die cast: aluminum-, zinc-, magnesium-, and copper-base alloys. Lead, tin, and, to a lesser extent, ferrous alloys can also be die cast. Parts produced by the die casting process are shown in Fig. 1. The primary categories of the conventional die casting process are the hot chamber process and the cold chamber process. These main process variations have been enhanced in recent years to produce high-integrity die castings, which provide mechanical properties suitable for use in structural and safety applications previously not considered for conventional high-pressure die castings. Hot Chamber. The hot chamber process is the original process invented by H.H. Doehler. It continues to be used for lower-melting materials (zinc, lead, tin, and magnesium alloys). Hot chamber die casting places the hydraulic actuator in intimate contact with the molten metal. The process minimizes exposure of the molten alloy to turbulence, oxidizing air, and heat loss during the transfer of the hydraulic energy. However, the prolonged intimate contact between the molten metal and the system components presents severe materials problems

in the production process, and the process is limited in the size of the part that can be produced. Details are presented in the article “Hot Chamber Die Casting” in this Volume. The cold chamber process solves the materials problem of higher-temperature operation by separating the molten metal reservoir from the actuator for most of the process cycle. Cold chamber die casting requires independent metering of the metal and immediate injection into the die, exposing the hydraulic actuator to the molten metal for only a few seconds. This minimal exposure allows the casting of higher-melting-temperature alloys such as aluminum, copper, and even some ferrous alloys. The cold chamber process is used to make parts of all sizes, from small brackets and connectors to automotive housings and engine blocks. A further discussion is found in the article “Cold Chamber Die Casting” in this Volume. High-integrity die casting processes have been developed in recent years that expand the

Fig. 1

potential markets for high-pressure die castings. These processes enhance various aspects of the conventional die casting process to provide castings with very low levels of porosity that can be heat treated and/or welded to meet stringent product requirements. Examples of highintegrity die castings are shown in Fig. 2. There are three different types of high-integrity die casting processes: squeeze casting, semisolid casting, and high-vacuum die casting. Squeeze casting is a term commonly used to refer to a die casting process in which liquid alloy is cast without turbulence and gas entrapment and subsequently held at high pressure throughout the freezing cycle to yield high-quality heat treatable components. Direct squeeze casting originally was developed as a liquid forging process in which liquid metal was poured into the lower half of a vertically oriented die set and subsequently closed-die forged. This is really not a die casting process. There are also several indirect squeeze casting approaches involving

Typical high-pressure die castings. Courtesy of NADCA (Ref 1)

716 / High-Pressure Die Casting to a high-pressure casting process using extremely low vacuums, in the range of 10 kPa (100 mbar) or less, in order to assure heat treatable castings. High-vacuum die casting relies on the vacuum system to effectively eliminate cavity gases from the shot sleeve and die, thereby eliminating the possibility of gas entrapment. In most other respects, the process is much like conventional die casting and uses thin, trimmable gates coupled with high metal-injection velocities to fill thin-wall sections. High-vacuum die casting generally produces thin-wall parts in which ribs are often used to provide strength. A complete discussion of vacuum die casting is included in the article “Vacuum High-Pressure Die Casting” in this Volume.

Advantages of High-Pressure Die Casting

Fig. 2

Typical semisolid high-integrity die casting products. Courtesy of NADCA (Ref 1)

injection of metal into a cavity via massive gates, which allow adequate feeding of solidification shrinkage while the casting freezes. In commercial practice today (2008), there are vertical or horizontal injection systems with the parting line of the die oriented horizontally or vertically, respectively. Squeeze castings can be made from the full range of heat treatable (and non-heat-treatable) aluminum alloys used in the permanent mold processes. Only the hot-short 2xx, 5xx, and 7xx aluminum alloys, sometimes used in sand and plaster mold casting, are not suitable for squeeze casting. Hot shortness is a tendency of an alloy to crack during or immediately after solidification (Ref 2). Squeeze castings are not limited to the higher-silicon (higher-fluidity) alloys needed for die casting. These processes are discussed in more detail in the article “Squeeze Casting” in this Volume. Semisolid metal (SSM) casting differs from squeeze casting in that it uses a novel semisolid material. In all other respects, however, the basic rules remain. The same low turbulence during filling followed by application of high pressure are critical elements for good process control. The principal difference is that the higher-viscosity semisolid metal allows higher metal velocities to be implemented before the onset of turbulence, and, of course, the metal is already half-solidified at the time of casting. These attributes allow SSM casting to achieve remarkable thin-walled components and high production rates. The feed material for SSM casting shown in Fig. 3 is aluminum A357. Feed material is typically 50 to 60% solid, with a globular

microstructure comprised of solid spheroids suspended in the melted portion to produce the consistency shown. There are three types of SSM processes: thixocasting, rheocasting, and thixomolding. Feedstock for SSM was originally prepared in specially cast billets produced on continuous casting systems, equipped with electromagnetic stirrers. The billets were cut to length and then heated to the proper temperature by induction heating before being placed into the casting machine. Recently, several systems have been developed to prepare on-demand feedstock for SSM casting. These systems offer the potential of reducing the need for special billets and billet heating, thereby simplifying the process and reducing cost. A complete discussion of SSM casting methods is provided elsewhere in this Volume. Vacuum die casting uses a vacuum pump to evacuate the air and gasses from a die casting die cavity and metal delivery system before and/or during the injection of metal. The term vacuum die casting can mean any vacuum in the die cavity below atmospheric pressure. Many die casting companies use vacuum to aid their die casting process. Some manufacturers use vacuum to not only evacuate the air and gasses from the die cavity but also to suck liquid metal into the shot sleeve prior to injection into the cavity. Vacuum is used not only in conventional high-pressure die casting but also to some degree in the high-integrity processes of squeeze casting and SSM casting. There is another term, high vacuum die casting, which describes a process used to produce high-integrity castings, that generally refers

The high-pressure die casting production rate is much higher than for gravity or low-pressure casting. The ability to produce castings with close dimensional tolerances greatly reduces machining operations. Die castings have good surface finish, which is a prime requirement for plating. Die castings can be produced with much thinner wall thickness, reducing overall casting weight. In the die casting operation, more complex parts can be produced, thereby reducing the number of components in an assembly. Inserts can also be cast in place during high-pressure die casting.

Fig. 3

Semisolid metal (SSM) feedstock material. This SSM is A357.0 (UNS A33570), an Al-Si-Mg alloy. Courtesy of NADCA (Ref 1)

High-Pressure Die Casting / 717

Disadvantages of High-Pressure Die Casting The high tooling costs of high-pressure die casting make short production runs generally uneconomical. There are a limited number of alloys suitable for die casting, and internal porosity restricts the heat treatment or welding of the finished castings, except for those produced by the high-integrity processes described previously. Iron or steel alloys are normally not die castable. There are restrictions in die casting on the casting size and casting wall thickness, which eliminate the possibility of die casting some parts. Die casting machine and machine maintenance costs are higher than for other casting processes due to their complexity and high pressure requirements.

Product Design for the Process Product design and die design are intimately related. The high-speed nature of the die casting process allows the filling of thin-wall complex shapes at high rates (of the order of 100 parts per hour per cavity). This capability places additional demands on the casting designer, because traditional feeding of solidification shrinkage is almost impossible. The inability to feed in the traditional sense demands that machining stock be kept to a minimum; high-integrity surfaces should be preserved. A factor in cost is the parting-line topology. The parting line is the line on the casting generated by the separation between one die member and another. The simplest and lowest-cost die has a parting line in one plane. Casting design should be adjusted if possible to provide flat parting lines. Draft is required on the die casting walls perpendicular to the parting line or in the direction of die motion. An important characteristic of good design is uniform wall thickness, which is necessary for obtaining equal solidification times throughout the casting. Die castings have wall thicknesses of approximately 0.64 to 3.81 mm (0.025 to 0.150 in.), depending on casting shape and size. Bosses, ribs, and filleted corners always cause local increases in section size. In particular, bosses that must be machined require consideration of the entire product-manufacturing cycle. The machinist will find it easier to drill into a solid boss; cored bosses may require floating drill heads in order to align the drill with the cast tapered hole that preserves the high-integrity skin of the casting. Cores and slides provide side motions for undercuts. A core body is generally round and buried within the cover or ejector die. A slide body has a rectangular or trapezoidal shape and crosses the parting line of the die. As with the cover and ejector dies, the impression steel is often separate from the holder steel. Cores and slides are actuated by various methods, including hydraulic cylinders, rack and pinion, and angle

pins. Innovative die design permits radial die motion at a price of die expense. There are die casting processes that use complex-shaped disposable cores similar to those in gravity casting processes. Cores and slides provide the casting designer with tremendous flexibility at the expense of an increase in die complexity. A standard set of cores—fixed core pins for small holes that are screwed in, or bolted-in inserts— can be used to reduce die construction cost and to permit rapid replacement. In certain cases, a reentrant shape needs to be cast into the part where there is no space for core/slide mechanisms. In such a case, the die designer can use a loose piece. A loose piece is placed in the die before each shot is made. It is then ejected from the die with the casting and separated manually or by a fixture. Although it provides design flexibility, the load/unload sequence required for loose pieces slows the process, thus increasing cost. Similarly, the die casting process can allow the part designer great flexibility in local material properties by the use of cast-in inserts of other materials, such as steel, iron, brass, and ceramics. The bond between insert and casting is physical, not chemical, in nature. Therefore, the insert should be clean and preheated. The insert should be designed to prevent pullout or rotation under working loads; knurling, grooves, hexagons, or flats are commonly used for this purpose. Proper support of hollow inserts will prevent crushing of the insert under the high metal-injection pressure. The wall thickness of the casting surrounding an insert should be no less than 2.0 mm (0.080 in.) to prevent cracking by shrinkage, hot tearing, and excessive residual stresses. Trimming. The die-cast part is ejected from the die with a variety of appendages (gates, overflows, vents, flash, and robot grasping lugs) that must then be removed. This secondary process is called trimming. Although trimming can be done manually, the high production rates characteristic of die casting demand automation. Trim presses are used to remove the excess material. Castings are often trimmed immediately after the casting process, because their higher temperature reduces the strength of the metal. It is common for the die casting machine and trimming operation to be located in a cell arrangement, with a robot providing the material handling. Trimming conditions directly influence the design of the part and the die casting process, especially the gating and parting-line definition. Trimming is facilitated by flat parting lines. The relatively rough edge that results from trimming may be acceptable and is often left as is. In some cases, this rough edge is not acceptable and must be removed by machining or grinding. The direction of flash must be such that the edge is machinable. Dimensional variation is determined by die design, the accuracy of die construction, and process variation. The most accurate dies are those machined using computer numerical control methods. Close control of alloy composition,

temperature, casting time, and injection pressure will lead to more consistent casting dimensions. The minimum variation in dimensions is found for those features contained entirely within one die half. Therefore, machining locators should ideally be placed in the same die half. Tolerances are a function of casting size and projected area. Features across parting lines have added variation because of the accuracy of repeated die closing. Die temperature, machine hydraulic pressures, and die cleanliness are the principal factors to be controlled. Finally, further dimensional variation occurs if the feature is in a moving die member such as a slide or core. Detailed product specifications for both conventional and high-integrity die castings are available from the North American Die Casting Association (NADCA) (Ref 1). In summary, a cost-effective die casting design demands proper attention to the dimensional variation of the process. Inattention to dimensional factors will lead to an inability to provide consistent products within economic process conditions. The product designer and the die caster must therefore initiate a dialog early in the product cycle.

Metal Injection The distinguishing characteristic of the die casting process is the use of high-velocity injection. The short fill time (of the order of milliseconds) allows the liquid metal to move a great distance despite a high rate of heat loss. Proper process performance depends on the delivery of molten metal with high quality, as defined by temperature, composition, and cleanliness (gas content and suspended solids). The molten alloy is prepared from either primary ingot or secondary alloys. A melting furnace is used to provide the proper temperature and to allow time for chemistry adjustment and degassing. The alloy is often filtered during transfer to a holding furnace at the casting machine. Most die casting machines provide the ability to control the piston acceleration in a linear fashion. Parabolic velocity curves are also available on some controls. This phase of injection can be accomplished in several steps. The third phase of injection is activated as the cavity is close to being filled. This intensification phase draws on an accumulator of highpressure hydraulic fluid or multiplies pressure using conventional piston intensifiers. This increases the pressure on the metal to force the rapidly freezing alloy into incipient shrinkage cavities. The gate is the controlling entry point into the casting. The gate serves a fluid flow need, but it must later be removed from the casting by trimming. Therefore, the gate cross section should be the smallest in the gating system. The cross section is determined by the desired fill time and flow rate that the casting machine can provide.

718 / High-Pressure Die Casting

7000 Machine power 6000 Theoretical Fill rate

Metal pressure, psi

5000

Machine

__________

Hyd. cyl .size

__________

Hyd. press. used

__________

Dry shot speed

__________

Plunger tipsize

__________

4000 Die line(Gate area)

Operating Window 3000

Gate velocity limits 2000 Operating Point

1000

0 0 200 300 1400 2100

400 2800

500 3500

600 4200

700 4900

800 5600

900 6300

1000 7000

Q , in.3/s

Fig. 4

PQ2 diagram generated by process control software. The volumetric flow rate, Q, is represented with a nominal 0 to 1000 scale to clearly show its nonlinearity. The 0 to 7000 scale models a specific process. Courtesy of NADCA (Ref 1)

The selection of a fill time for the casting is based on experience and experimentation. The limited selection of plunger diameters for a given machine restricts the design. The cold chamber process links the volume of metal to the plunger diameter by filling the shot sleeve approximately one-half to two-thirds full. Flow Modeling. It has been recognized by the die casting industry that the ability of the casting machine to provide this metal volume flow, while keeping the dies closed during injection, must be considered. The tool that has been developed for this purpose is called the PQ2 diagram (Fig. 4). It can be shown that the pressure, P, on the metal and hydraulic system is proportional to the square of the injection velocity and therefore the volume flow rate, Q. The line with the negative slope is the machine characteristic line. The charac-

teristic line moves to the left and right with changes in hydraulic pressure, shot valve throttling, and plunger diameter. The line that starts at the origin of the graph is a measured relationship of pressure to flow rate for the particular casting and gate being cast. The effect of adjusting the gate area is that the angle of the line changes. In order to have a successful operation, the operating point should be inside the operating window. The flow characteristics are nonlinear, as illustrated by the Q axis, 0 to 1000 scale, on the PQ2 diagram (Fig. 4). The 0 to 7000 scale below this nominal scale is developed for the specific process being evaluated. Optimization of these various parameters for the casting and machine provides the process engineer with a powerful tool for process definition and debugging. A number of micro-

computer programs are available to base gate design on this hydraulic approach. The overflow is the final component in the fluid flow system. Although they add to the weight of remelt, overflows do serve a variety of purposes. They can act as a reservoir for metal to be removed from the cavity, and they can provide an off-casting location for ejection pins, robot holds, or instrumentation points. REFERENCES 1. North American Die Casting Association (NADCA), 241 Holbrook Dr., Wheeling, IL 60090, (847) 279–0001, www.diecasting.org 2. J.G. Kaufman and E.L. Rooy, Aluminum Alloy Castings, ASM International, 2004, p 295

ASM Handbook, Volume 15: Casting ASM Handbook Committee, p 719-723 DOI: 10.1361/asmhba0005267

Copyright © 2008 ASM International® All rights reserved. www.asminternational.org

Hot Chamber Die Casting Frank E. Goodwin, International Lead Zinc Research Organization

DIE CASTING ALLOYS that do not attack die casting machine components used for injection or delivery of metal to the die during long periods of immersion can be hot chamber die cast. These casting metals are processed at relatively low temperatures and include zinc alloys with casting temperatures between 400 and 450  C (750 and 840  F). Also, lead and tin alloys, together with other systems having casting temperatures below 250  C (480  F), are routinely hot chamber die cast. The injection end of a hot chamber die casting machine is shown in Fig. 1.

Melting Process The holding furnace, which can also serve as a melting furnace if a central melting facility is not used, has traditionally been gas fired and constructed of cast iron. In recent years, great improvements in energy efficiency have been

realized by constructing these pots from ceramic material and using either gas or electric immersion heaters to supply heat to the bath. If a central melting facility is used, liquid casting alloy is delivered either by ladle or by using a launder system. If melting is carried out directly in the pot, it is recommended that ingots of no greater than 6.8 to 9.0 kg (15 to 20 lb) be used. Each is gradually lowered into the bath using a clockwork type of mechanism. The casting alloy is in the form of a long, thin rectangular ingot with a hole through which the hook of the lowering mechanism is inserted. This design of ingot is termed a margash shape (Fig. 2). Gradual lowering of the ingot into the bath rather than sudden additions of whole ingots greatly improves temperature uniformity of the bath and reduces dross and oxides. For zinc, the pot should be equipped with controls so that the temperature of the molten bath can be maintained within 6  C (10  F). The furnace capacity required depends on the size of the casting machine and the production rate.

Shot cylinder

Accumulator

Overflow

Pump

Overflow gate Die cavity

Control valve Ingate nozzle Gooseneck

Gooseneck plunger

Generally, a holding furnace should be able to hold at least four times the amount of metal required for 1 h of operation. If the furnace is also used for melting, its capacity is typically 450 to 9000 kg (1000 to 20,000 lb), although immersion tube furnaces can range up to 18 metric tons (40,000 lb). Total melting furnace capacity is usually five to seven times the amount of metal required per hour.

Injection Components In the molten metal pot, a gooseneck plunger and cylinder arrangement is located that allows introduction of casting metal to the cylinder when the plunger is retracted. These machine components are normally made from tool steel or stainless steel. Also, the sealing rings in the plunger are made of a similar material. This plunger-and-cylinder arrangement injects metal into the gooseneck, so called because of the shape of this channel, which is typically made of cast iron. The gooseneck connects the melting pot to the die casting machine. The gooseneck is shaped to allow for smooth flow of the rapidly injected casting metal and is also maintained at a temperature that avoids solidification of the casting metal in the gooseneck channel. The end of the gooseneck that delivers metal to the die casting machine has a semispherical end that allows it to be fit into the back of the cover or stationary half of the die casting die in a conventional machine. This spherical joint is tightly clamped to prevent leakage of casting alloy.

Hydraulic fluid reservoir Sprue pin

Runner

Die casting machine frame

Fig. 1

Holding furnace

Basic components of a typical hot chamber die casting system

Fig. 2

Two margash-form zinc alloy ingots for feeding die casting alloy to the holding furnace. Note the hole that accepts a hook that slowly lowers the metal into the furnace. Source: Courtesy of Allied Metal Company, Chicago, IL

720 / High-Pressure Die Casting The injection cycle for a conventional hot chamber die casting machine is shown in Fig. 3. To begin the cycle, the die is closed, and the gooseneck is filled with liquid casting alloy. The plunger then pushes the liquid alloy through the gooseneck and nozzle into the die cavity. The alloy is held under pressure until it solidifies. After a predetermined time, the die opens, and the cores, if any, retract. The casting stays in the ejector half of the die. The plunger returns, pulling metal back through the nozzle and gooseneck. The ejector pins push the casting out of the die. As the plunger uncovers the filling hole, molten metal flows through the inlet to refill the gooseneck. Casting Machines. In recent years, so-called four-slide machines have become increasingly popular. These feature a metal platen on which four quadrants of a die casting die are mounted to open and close (Fig. 4). Typical die dimensions are 160 by 160 mm (6 by 6 in.), with casting weights of 245 to 650 g (8.7 to 23 oz). Cycle times can reach 450 shots per hour. In a four-slide machine, the gooseneck is fitted to the back of the platen rather than any portion of the die. The portions of the die close together

Materials of construction for the gooseneck nozzle seat are chosen to resist high wear and have been made from nitrided alloy steel, hot worked tool steels such as H13, high-speed tool steel, and stainless steel. The gooseneck, plunger, and cylinder are always designed to allow easy replacement. Pressure. Injection energy is produced by either hydraulic or pneumatic means. Injection pressure used for zinc die casting alloys usually ranges from 10.3 to 20.6 MPa (1500 to 3000 psi). The lower pressures are used for simpler castings, while higher pressure is for more complex ones. The best practice is to use the lowest pressure that will produce acceptable castings; however, a minimum pressure of 10.3 MPa (1500 psi) is usually required for obtaining an acceptable combination of soundness, surface finish, and mechanical properties. The injection characteristics of the machine can be changed by increasing or decreasing the size of the cylinder or the plunger that fits into it, although some adjustment of plunger stroke length can be made using a machine control system; the amount of metal required for different dies is best adjusted to the die by using a proper-sized cylinder and plunger combination. Movable die half

Fixed die half Nozzle

Ejector pins

Gooseneck

v, F

Plunger Cavity Pot

to form a leakproof cavity into which the casting metal is injected.

Distinctions Between Hot and Cold Chamber Processes Beyond the injection section, a conventional hot chamber die casting machine closely resembles a cold chamber die casting machine. A major difference is the cycle time at which hot chamber die casting machines can typically run. This is due to two reasons. The use of the hot chamber injection system allows for much higher cycling of the submerged plunger in the cylinder, because metal is readily available to fill the injection cylinder after each cycle. In a cold chamber machine, metal must be either manually or automatically ladled into the cold chamber cylinder, which will always require more time. Second, casting metals processed in hot chamber machines have lower melting points than those processed in cold chamber machines; therefore, less heat extraction is needed to solidify the metal before ejection of the part and completion of the casting cycle. The design of the runner and gate channels in the die that transmit liquid casting alloy to the casting cavity also differs in some ways from cold chamber dies. As in the cold chamber process, hot chamber die casting dies can be constructed in singlecavity, multiple-cavity combination, or unit dies. However, dies that are designed to produce zinc alloy castings can seldom be used to produce castings of aluminum alloys or other metals that are cast at higher temperatures, because zinc alloys can be cast at thinner sections, smaller radii, and closer tolerances than aluminum,

Chamber Cross head Slide (a)

V

Die block

(b)

V

Core

Fig. 4

(c)

Fig. 3

Operating sequence for the hot chamber die casting process. (a) Die is closed, and hot chamber (i.e., gooseneck) is filled with molten metal. (b) Plunger pushes molten metal through gooseneck and nozzle and into the die cavity. Metal is held under pressure until it solidifies. (c) Plunger returns, pulling molten metal back through nozzle and gooseneck. Die opens and cores, if any, retracts. Casting stays in ejector die until ejector pins push casting out of ejector die. As plunger uncovers filling hole, molten metal flows through inlet to refill gooseneck.

Four-slide mechanism for production of complex shapes in a single casting. The multi-slide tool is made up of the die block, slide, crosshead and cover plate. Each die block has either a cavity and/or cores on its face, which together form the complete cavity and runner profile into which the molten metal is injected. These die blocks are mounted onto sliders, which fit precisely into a crosshead, ensuring repeatable opening and closing operations. A cover plate, bolted onto the top of the tool, holds all the components together.

Hot Chamber Die Casting / 721 magnesium, or copper alloys. However, a die design for casting the higher-melting-point alloys can be used for casting zinc alloys. The precision capabilities of zinc casting alloys vary with the casting size, and they are described in the most recent edition of NADCA Product Specification Standards for Die Castings (Ref 1). For zinc die castings, both a normal and precision tolerance level are given. These allow many components in zinc to be cast to near-net shape or even net shape condition, avoiding the need for further operations, such as machining, that would add cost to the part. On the opposite side of the gooseneck nozzle bushing, the orifice that allows metal to enter the die cavity faces the beginning of the feeding system in the moving or ejector side of the die, termed the sprue. The purpose of the sprue is to allow for uniform introduction of the injected liquid metal to the runner system that feeds the die cavities. The volume of this sprue is much smaller than the volume of the biscuit typically used in cold chamber die casting and can be a much smaller proportion of the untrimmed casting weight in comparison with a part made in a cold chamber die. The design of the gate and runner systems used in hot chamber die casting is otherwise similar to those in cold chamber die casting.

Gate and Runner Design An example of a typical runner, gate, and overflow configuration for a faucet fixture casting is shown in Fig. 5. This casting is both highly decorative and functional; the use of overflows allows hot metal to fill areas critical to appearance. Casting metal enters the sprue, where it is distributed to one or more runner systems that feed casting cavities. The main runner for a casting can be divided into branch runners, but the sum of the cross sections of all branch numbers cannot exceed the cross section

of the main runner. Generally, slightly tapered runner systems are used to accelerate the flow and prevent air entrainment. Each runner ends with a narrow gate in which the casting alloy is further accelerated as it fills the die cavity. General rules to follow in selecting a feed system configuration are:  Minimize runner length  Provide uniform flow across the die cavity,

such that air is forced out of the cavity ahead of the molten zinc into overflows and/or vents Feed from the side with the most critical surface requirements Flow across the short dimension of the cavity Avoid flow across large openings by center gating (Fig. 6) or provide runners across large openings Select runner and gate configurations that control flow direction through the cavity so that the metal tends to follow desired paths

   

The dark sections shown in Fig. 5 and 6 are trimmed off, leaving the final unshaded casting. The lines drawn on the casting are segment lines, which are a useful way to visualize the intended metal flow paths in the casting cavity.

Temperature Control The temperature of the die casting dies for zinc die casting is relatively low, ranging from 160 to 245  C (325 to 475  F); therefore, hot worked tool steels are not generally required for dies. However, for extremely long runs and high dimensional accuracy, hot worked tool steels such as H13 will provide optimal die life. The die hardness is also less critical for alloys of higher casting temperature. Steels prehardened by the manufacturer to any maximum hardness content with reasonable machinability can be used. The typical hardness is 29 to 34 HRC (Rockwell C hardness) or 280 to 320 HB Overflows

Tangent runner 1 Uniform runner 1 Second cavity Uniform Fan gate runner 3 (main runner)

2

1

9

10

3

(Brinell hardness). For the slides and cores, hardenable stainless steels such as 440B can be used in many cases. If significant wear occurs, such as in slides, hot worked tool steels such as H13 can be used. Because these tool steels respond to nitriding, they can be selectively hardened, and therefore, a component made from one of them can be treated for different properties in different areas. For example, the main portion of cores and slides can be nitrided for wear resistance, and the end can be masked to resist nitriding for better resistance to heat checking and spalling. Lubricating the slides and cores with molybdenum disulfide or colloidal graphite helps to ensure smooth action and minimize wear. With hot chamber die casting of zinc alloys, the surface life of the die casting die is normally much longer than that for casting aluminum, magnesium, or copper alloys. It is not unusual for die to last for one million shots, and the very small castings made on four-slide machines can reach one billion shots before extensive maintenance is required. Maximum die life depends largely on having well-designed dies, trained operators, and a rigidly enforced program of machine and die maintenance. Lower die temperatures, near 160  C (325  F), are normally used for thicker-section die castings, whereas the higher end of the range, 245  C (475  F), is used for thin-section castings. When hardware finish is required, temperatures at the high end of the range are generally required, regardless of the casting thickness. Some casting shapes require localized heating or cooling of the die above or below the established temperature. Metal overflows are often used to heat die areas surrounding the parameters of castings that have thin sections far from the main runner. This method of local heating helps to fill thin sections and to improve casting finish. Conversely, water channels are frequently placed behind the runner area immediately adjacent to the sprue to provide localized cooling and to prevent soldering of the molten metal to the die. Control of die temperature is one of the most important parameters affecting production rates and casting quality. A number of die design programs are available to allow the designer

4

5 6

Sprue base

Turning vane Uniform runner 2

7 8

Typical cross section through runner and ingate

Fig. 5

Tangent runner 2

Sketch of general runner, gate, cavity, and overflow configuration of a faucet fixture casting. The numbered sections indicate the flow pattern in the finished part. The dark runner area and the overflow areas are trimmed from the finished part.

Fig. 6

Center-gated casting. An alternative method of gating a large casting with a large center opening. Dark areas will be trimmed from finished part. Lines indicate flow pattern from center to perimeter.

722 / High-Pressure Die Casting to determine proper placement of cooling lines in the die and the coolant capacity required. Generally, dies are designed with excess cooling capacity, and then a temperature control system is used to modulate the flow of coolant in the die. Heating and cooling of the die can be accomplished using oil as a heat-transfer medium. Water cooling can also be used, but in this case, cartridge heaters or other means of preheating the die prior to beginning the casting campaign are required. Many die casters control die temperature by observing the casting being produced and then manually adjusting valves to regulate the flow of die cooling water through passages in the die until a satisfactory casting is produced. If the cooling passages are properly located, if the machine is operating consistently in a very repetitive manner, if the machine operator is skilled and is giving sufficient attention to the die temperature, and if the part does not require close temperature control, this method is satisfactory. However, as casting conditions become more critical, as they are with thin-wall die castings, it is virtually impossible to control the flow of cooling water manually so as to obtain satisfactory castings consistently. Automatic temperature control should be employed to continuously monitor die temperature and adjust the coolant flow rates to maintain the optimal die temperature for critical applications. Studies on some large castings indicate that scrap is reduced between 15 and 40% when temperature is automatically controlled (Ref 2–4). The reasons for the scrap reduction are:  No coolant flow occurs during start-up until

the die is up to temperature, reducing startup time and amount of scrap produced during start-up.  Die temperature variations are reduced during steady running.  Whenever an interruption occurs in the production process, the coolant is automatically shut off, reducing the number of scrap parts that must be made to bring the die back up to operating temperature.

Alternatively, a hot oil medium can be used instead of die coolant, which provides accurate control of die temperature. In either case, a basic die temperature controller of the type diagrammed in Fig. 7 is suggested. The basic components are a thermocouple, located approximately 6.4 mm (0.25 in.) below the die cavity surface, a temperature controller that determines whether the temperature as measured at the thermocouple location is above or below a setpoint temperature, a solenoid valve to regulate coolant flow, and a manual valve to control the flow rate. If the manual valve is not used, the temperature will tend to swing through large extremes, and a generally unsatisfactory performance will result. For complex castings, or where more control is desired, a multichannel temperature controller is desired. This monitors temperatures from thermocouples located at selected locations in the die and controls the flow of coolant to each selected part of the die. An example is the location of thermocouples on an automobile headlight bezel (halfsection view) (Fig. 8).

Ejection and Postprocessing Robotics. Operations such as die lubrication, casting removal, and automation of the die casting process using robotics are common to hot chamber and cold chamber machines. Trimming of the die casting after it is ejected from the machine is automatically performed for many zinc die castings because of their small size. This requires that the gate thickness be carefully controlled. Tumble degating of many zinc die castings is possible. The conduit connector (Fig. 9) is an example of a high-volume part designed for tumble degating. However, for larger castings, a trim press may be required. In finished trimming, the part must almost always be quenched in a water-soluble coolant or lubricant similar to the lubricants used for machining zinc castings. Failure to include soluble oil usually causes the trim die plates to solder, with consequent tearing of the trimmed parts.

Quenching several hundred pounds of shots per hour in a small tank requires either a heat exchanger for the quenching fluid or connection to a larger remote reservoir so that heat can be dissipated naturally. Trim presses that remove the sprue and runner system, together with overflows and flash, typically deliver these trimmed portions of the shot to a tank that allows for recycling of this scrap after cleaning. In-die degating can also be designed, allowing for simplification of the postprocessing area. Recycling. Clean scrap from trimming operations, together with scrap casting, can be recharged into the die casting circuit in unlimited portions, although the use of 50% maximum of scrap per charge is recommended. There are safeguards that should be employed to ensure that remelt does not disrupt the desired alloy chemistry. The ability to remelt zinc process scrap many times without losing properties is of significant advantage to the zinc die caster; however, caution should be taken to keep this material clean and free of unwanted substances. It should be stored separately from other metals if it is accumulated in batches. If conveyed back to a central melter, conveyers and the furnaces should be covered when maintenance work is being done overhead. Floors and tables should also be kept clean. If there is any doubt as to the purity of the scrap, it should not be used in any proportion until it has been analyzed. Scrap returned for recycling must be free of oil and moisture. A safety hazard is created when oil and moisture are present on the metal being charged into the furnace. Some zinc scrap may be electroplated. This material should be remelted separately and added back in small quantities or, better still, sold outright. The electrodeposited metals will separate and float to the top of the bath, where they can be skimmed off. Agitation will increase copper, nickel, and chromium levels and therefore should be avoided. Die castings that have been joined using lead- or tin-base solders should not be added to the scrap feed under any circumstances. The very low limits

A Thermocouple B

Die cavity

C

Controller

Fig. 9

Water main Solenoid valve

To drain

Fig. 7

Manually operated valve

Components of a commercial single-channel die temperature-control system

Fig. 8

Location of thermocouples in a die for a headlight bezel. A, B, and C indicate approximate positions of thermocouples to be mounted 1 6.4 mm (4 in.) below the die cavity surface.

Design of a casting for automatic degating. A two-cavity die is used to make electrical conduit connector castings. Use of thin gates and placement of the mass of the casting at the end of a runner allow the use of tumbling to automatically degate this high-volume part. Zinc casting metal enters through a central sprue and is divided to flow through the two runners in opposite directions. One of the castings is attached, while the other has been removed to show the location of the gate. Source: Courtesy of Bridgeport Fittings, Bridgeport, CT

Hot Chamber Die Casting / 723 of impurities in casting composition specifications make only very small quantities of contamination by lead, tin, or cadmium capable of bringing a large holding pot or melting furnace out of specification. This includes castings that may have pressed-in bronze bearings that, in many cases, contain lead or tin—introducing such items into the scrap feed should be avoided at all cost. Even when zinc alloys are highly contaminated, a zinc refinery can refine die-cast scrap, allowing for complete recycling. Worldwide, 400,000 metric tons of zinc alloy die castings are recycled, mainly from shredded automobile scrap (Ref 5). Fluxing of zinc alloys during melting and holding is generally not required and can change the percentage of alloy elements. However, 1.4 to 2.3 kg (3 to 5 lb) of a chloride or fluoride flux

should be added for each 450 kg (1000 lb) of metal when the melting stock partially or totally comprises trimmings, gates, and rejected castings. A few pounds of flux per ton of alloy will reduce the magnesium content, and greater flux additions can make the magnesium disappear completely. Therefore, it is necessary to control temperatures continuously, to flux properly, and to check the analysis, particularly for aluminum and magnesium contents. Adjustments should be made as required.

REFERENCES 1. NADCA Product Specification Standards for Die Castings, 6th ed., North American Die Casting Association, 2006

2. B.K. Dent and R. Fifer, Production Operation with the ILZRO-Battelle Multichannel Temperature Controller, Paper 5572, Transactions of the Seventh SDCE International Die Casting Congress, The Society of Die Casting Engineers, 1972 3. R. Reddi, “Temperature Control of Die Casting Dies,” Paper G-T75-125, Presented at the Eighth SDCE International Die Casting Congress, March 17–29, 1975 (Detroit, MI), The Society of Die Casting Engineers 4. Zinc, (No. 1), Zinc Institute, Inc., 1974, p 1 5. “Zinc Recycling—The General Picture,” International Zinc Association—Europe, 2005

ASM Handbook, Volume 15: Casting ASM Handbook Committee, p 724-726 DOI: 10.1361/asmhba0005268

Copyright © 2008 ASM International® All rights reserved. www.asminternational.org

Cold Chamber Die Casting Robert J. McInerney II, BuhlerPrince

THE COLD CHAMBER die casting process is used with higher-melting-point alloys, such as aluminum and magnesium. Since the cold chamber is located outside of the furnace, as compared to hot chamber, it requires a means of moving the molten metal from the holding furnace to the cold chamber. The cold chamber is attached between the die casting machine front platen and the die. The transport of the molten metal is typically done with a ladle mechanism, either manually or automatically, when casting aluminum alloys. Casting cycle times can range from 10 s for a small machine up to 2 min for a large machine.

Machine Components The cold chamber high-pressure die casting machine performs several functions during the

Fig. 1

die casting process and is essentially two machines that work together. The two machines are called the clamp end and the shot end. The machine functions are controlled, directly or indirectly, and machine variables are monitored by the processing engineer. Figure 1 shows the main components of a horizontal cold chamber die casting machine. Items to the left of the front plate are considered part of the clamp end, and items to the right of the front plate are considered part of the shot end. The machine is powered by a high-performance hydraulic system, consisting of pumps, cylinders, valves, and oil/nitrogen accumulators. Oil/nitrogen accumulators are used because they store a large amount of potential energy. This energy can be released easily and accurately in terms of time and pressure by the control valves in the system. Clamp End. The functions of the clamp end of the machine include opening and closing the die,

Typical parts of a horizontal cold chamber die casting machine

developing clamping force, and providing power for sliding cores and for ejecting the casting from the die cavity. The machine must open and close the die on a straight line to keep the die parting line parallel. In most cases, the faces of the die are flat, and their plane determines the parting line. If this is not done properly, there will be incomplete closure and interference between the die and the casting. This interference can cause the casting to be bent and the tool to wear prematurely. The design of the machine is robust to provide the needed rigidity. To open and close the die straight:  The machine must be level.  The plate support rails and plate supports must

be in good condition for operation and leveling.

 The tie bars and locking mechanism must be

parallel with the base.

Cold Chamber Die Casting / 725 The shot end includes the cold chamber, plunger rod, and plunger tip (Fig. 2). The primary mechanical requirement of the shot end is to provide good alignment so that proper clearance is maintained between the plunger tip and cold chamber. Misalignment will result in a drag, which will affect plunger velocity. The cold chamber container is also called the shot sleeve. To the right of the plunger rod are the hydraulic system piston (shot cylinder), shot accumulator and intensifier accumulator, and control valves that provide velocity to the plunger. This velocity is changed at predetermined positions, providing the output force (pressure). The velocity changes are linear or parabolic. This output force allows the plunger to move forward and is increased after the cavity is full for the intensification phase.

Process Parameters

Fig. 2

Typical parts of the shot end of a horizontal die casting machine

Fig. 3

Shot trace from machine control panel. Process variables monitored are plunger position and velocity and cylinder rod and shot cylinder head pressure.

 The front plate must be perpendicular to the       

base. The four plates must not be bent. There must be caved-in areas where the die makes contact with the plate. The tee slots that hold the die to the plate must be in good condition (no cracks). The tie bars and load indicators must be in good working condition. The linkage and bearing housing can have only a minimal amount of wear. The traveling plate support must carry the traveling plate (no weight can be on the tie bars). Die carriers must be used when needed.

The machine must also provide the correct amount of clamping force to seal the die parting line, operate core slides and locks, and hold the die closed against the force of the molten metal. The required force to hold the die closed is determined by the projected area of the casting and the amount of pressure on the metal. Mechanical locks are often needed. More details on die components and fluid flow dynamics are found in the article “Die Casting Tooling” in this Volume. Shot End. The purpose of the shot end is to inject the molten metal (the shot) into the die. This injection must be done at a controlled velocity and pressure.

The shot profile, as seen on a machine control screen for a typical aluminum die casting process, is given in Fig. 3. The velocities in this process are called slow shot and fast shot. The slow shot is typically split into two different phases: a pour hole close speed and a speed used to control the wave of molten metal in the cold chamber as the cold chamber is filling. The shot must start forward with as smooth a motion as possible as soon as the ladle mechanism finishes the pour, to prevent metal from splashing out of the shot sleeve (cold chamber). The size of the die cavity, including gating and vents to be filled, determines the size of the shot. This pour hole close velocity (0.25 m/s, or 10 in./s) indicates the plunger forward motion until the pour hole is closed. The second portion of the slow shot phase is the critical slow shot. The plunger diameter and the percentage that the cold chamber is filled determine this critical slow shot velocity (0.4 m/s, or 15 in./s). Some die casting machines are capable of providing an engineered linear or parabolic acceleration rate to the critical slow shot velocity, reducing the possibility of air entrapment even more. Fast Shot. The final velocity phase provided by the casting machine is the fast shot velocity (2.8 m/s, or 110 in./s). This fast velocity is determined by the required time to fill the cavity and the plunger velocity needed to push the metal through the gate for atomized flow. Figure 3 shows a complete shot profile, including the shot velocity and the related shot cylinder piston and rod pressures. From the plunger distance traveled and its velocity, most of the shot cycle is consumed by the slow shot portion (2 s), and the fast shot takes approximately 0.1 s. The die must be filled quickly to minimize heat loss in the melt. The intensification phase lasts a few milliseconds. Intensification Phase. The final phase of the cold chamber casting shot process is the intensification phase. The aluminum alloy will

726 / High-Pressure Die Casting shrink approximately 7% as it solidifies, and as the casting shrinks, there will be tears, cracks, and voids in the casting. To help overcome this shrinkage, which would result in a porous, lowquality casting, there must be a means of pushing metal through the gate to fill the shrinkage. The intensification system on the casting machine is designed to increase the output force of the shot cylinder, which produces higher flow pressure to force metal through the gate and narrowing flow paths. This will help reduce or eliminate the shrink porosity.

Component Size. Sizing of the cold chamber and plunger tip is highly dependent on the casting that is being cast. For small castings, the die can be designed with multiple cavities to match the machine capabilities and to increase productivity. The process engineer will use the casting size, wall thickness, and final function to determine the volume of melt required and the corresponding correct cold chamber size. Knowing these variables will also allow the engineer to calculate the correct shot velocities and intensification phase pressure.

The size of the cold chamber/plunger tip diameter may change from 50 mm (2 in.) for a small casting, such as a power saw component, to 180 mm (7 in.) for a large die that produces automobile parts. The die casting process and tooling are also discussed in the articles “High-Pressure Die Casting” and “Die Casting Tooling” in this Volume.

ASM Handbook, Volume 15: Casting ASM Handbook Committee, p 727-731 DOI: 10.1361/asmhba0005269

Copyright © 2008 ASM International® All rights reserved. www.asminternational.org

Squeeze Casting* Rathrindra DasGupta, National Science Foundation

Squeeze Casting Squeeze casting (Ref 1–4) is considered a high-integrity process and has emerged as an alternative to conventional casting techniques (gravity permanent mold, low-pressure permanent mold, sand casting) and forging. Squeeze casting builds on conventional (high-pressure, high-velocity) die-casting practices and, in recent years, has been widely used to manufacture automotive components (essentially aluminum) requiring high impact strength, high fatigue strength, and pressure tightness or wear resistance. Two types of squeeze-casting machines are commercially available: vertical squeeze cast (VSC) and horizontal-vertical squeeze cast (HVSC). Figure 1 shows the schematic of a typical HVSC machine that is more widely used. Major features of the HVSC machine include:  Horizontal die clamping  A vertical, high-pressure delivery system  A tilt-docking injection unit in which the

 Smaller dendrite arm spacing and a fibrous

 Use of localized squeeze pins to eliminate

eutectic silicon morphology resulting from the rapid solidification rate  Heat treatable. Unlike conventional (highpressure, high-velocity) die castings, squeeze castings can be subjected to solution treatment with minimal or no blistering.  Used for sections >2.5 mm (0.1 in.)

shrinkage porosity in areas otherwise difficult to feed Avoidance of varying wall thicknesses (thick-thin-thick) to aid in metal flow Casting layout in the die (Fig. 3). It is evident from the comparison that a more optimized flow pattern is achieved with the vertical gating and not the horizontal gating (vortex formation). Avoidance of sharp corners (to achieve nonturbulent flow) and undercuts Use of heat treated, premium-grade H13 steel or equivalent for the cavity/retainer Use of larger gates and runners (in comparison to conventional die casting) to assure planar cavity fill and to aid in directional solidification Flat area for the gate, since the gate must be sawed off and not trimmed, as in conventional die casting

Casting and Tooling Design for Squeeze Casting The casting/tooling design process (Ref 1, 2) for squeeze casting (HVSC) requires the following considerations:  Location of cooling channels or high-ther-

mal-conductivity-material inserts (Fig. 2) in the tool to extract heat from the tool quickly and to achieve directional solidification

sequence involves die closure, transfer of molten metal into a vertical chamber (shot sleeve), tilting of the shot unit to injection position, docking of the shot sleeve, and injection of molten metal into the die cavity

 

  



Cover half

Eject half Moving platen

Cavity Stationary platen

Molten metal

In the squeeze-casting process, degassed and filtered molten metal is ladled into a vertical shot sleeve and slowly forced into a preheated and lubricated die cavity at a speed slower than that used for conventional die casting. Solidification of metal then occurs under a continuous application of pressure. The typical advantages of squeeze casting include:

Shot cylinder

 Minimal air entrapment in the matrix result-

ing from reduced turbulence during the flow of metal into the die cavity  Reduction or absence of shrink porosity due to high cavity pressure (70 to 100 MPa, or 10 to 15 ksi) and rapid solidification rate (from the contact of the molten metal with the lubricated die surface)

Fig. 1

Schematic of a horizontal-vertical squeeze-cast machine

*Reprinted with permission from the North American Die Casting Association

728 / High-Pressure Die Casting

Factors Affecting Squeeze-Cast Products (Ref 1, 2) As in any casting process, monitoring and control of select process parameters influencing the integrity of squeeze castings is also necessary to produce defect-free (absence of porosity and/or inclusions) components. The following sections outline a few of the key process characteristics.

Cube insert

Metal Temperature and Metal Cleanliness  Metal temperature control (to less than +/ 5 After filling completes, heavy section is very hot.

Fig. 2

High-thermal-conductivity inserts

 C, or +/ 3  F) helps minimize hydrogen absorption by the melt (thus retarding the occurrence of gas porosity in castings), sludge formation, and/or oxidation of the molten metal (thereby hindering the occurrence and entrapment of hard spots in castings).  The cleanliness of the metal is typically achieved via continuous degassing (lance or porous plugs) of molten metal in the furnace and the use of ceramic filters. The overall cleanliness of the molten metal can be easily determined using either the reduced pressure test or the Prefil-Footprinter (ABB Group) (see the article “Measurement of Metal Quality” in this Volume).

Casting (Cavity) Pressure  Pressure levels of 70 to 100 MPa (10 to 15

ksi) are generally used to help reduce shrink porosity and to promote rapid solidification of the molten metal in the die. The magnitude of cavity pressure is dictated by the part geometry, and any increase in cavity pressure beyond the optimal pressure required for a given component does not necessarily contribute to significant improvements in mechanical properties. Oxides stay on the flat surface

Less metal pressure drop

Type of Lubricant  Care must be taken in selecting the type of

lubricant in squeeze-casting operations, since the reaction between the molten metal and the lubricant may lead to undesirable gases (e.g., hydrogen) being generated. Entrapment of these gases in castings can contribute to the formation of blisters during heat treatment.

Gate Position, Area, and Velocity  Gates are normally >3 mm (0.1 in.) thick to

Fig. 3

Casting layout in the die. Vertical gating (right) provides a more optimized flow of metal.

avoid premature freezing during component solidification. Gate placement is at the thickest part of the component to promote directional solidification.

Squeeze Casting / 729  Gate velocity is under 500 mm/s (20 in./s).

The slow gate velocity helps reduce turbulence and entrapment of gas in the metal.

Presolidified Layers (Cold Flakes) from the Shot Sleeve  Cold flakes form on the inner wall of the

shot sleeve when molten metal, in contact with the inner walls of the shot sleeve, solidifies prior to injection of the metal into the die cavity. Entrapment of cold flakes in castings contributes to a nonknitting condition and thus can affect product performance, including premature failure of the casting. Maintaining the appropriate sleeve temperature is therefore critical in preventing the occurrence of cold flakes.

melt cleanliness, and shot velocity to help reduce gas entrapment.

Squeeze-Cast Applications Squeeze casting has been successfully used in manufacturing a variety of products using traditional nonferrous casting alloys (particularly aluminum) (Ref 1–4). Examples of squeeze-cast applications include: Axle Carrier. Cast from a modified 383 aluminum alloy with a T6 temper and shown in Fig. 4, the part requires high strength (yield strength >280 MPa, or 40 ksi; tensile strength >350 MPa, or 50 ksi) and high stiffness.

Axle Cover. This part requires pressure tightness. It is shown in Fig. 5 cast from a modified 383 aluminum alloy with F temper. Front Steering Knuckle. The mechanical property requirements for these knuckles are: yield strength, >207 MPa (30 ksi); tensile strength, >276 MPa (40 ksi); and elongation, >6%. An example cast in A356.2-T6 aluminum alloy is shown in Fig. 6. Suspension Links. The mechanical property requirements for the links are: yield strength, >207 MPa (30 ksi); tensile strength, >276 MPa (40 ksi); and elongation, >7%. Various

Heat Treatment  Too high a solution-treatment tempera-

 

 

ture may cause incipient melting of the casting; on the other hand, too low a temperature may cause inadequate dissolution of Mg2Si and other phases. Too long a time at temperature may coarsen the eutectic silicon particles, thereby contributing to reduced ductility. Care must be taken in selecting the appropriate quench medium (air, water, or polymer) as well as the quench medium temperature. Although the use of hot water as a quenchant remains popular, forced-air cooling or quenching in a polymer bath following solution treatment are techniques often employed to minimize distortion of very large and complex components. The quench delay is typically less than 30 s. Too high an aging temperature or too long a time at temperature must be avoided to prevent overaging (softening) of castings.

Fig. 4

Axle carrier cast from modified 383-T6 aluminum alloy

Fig. 5

Axle cover cast from modified 383-F aluminum alloy

Porosity  A critical feature for the proper use of

squeeze casting is the application of process simulation and modeling prior to casting development trials to predict areas that may be difficult to feed, that is, determining areas prone to shrink porosity. Optimization of cooling line configurations and cooling sequences is one way to reduce or eliminate porosity in these locations. The other option involves the use of separate squeeze pins (localized squeeze), as described earlier.  Gas entrapment in squeeze castings can lead to the occurrence of blisters during heat treatment. Thus, attention must be paid to proper vent design/location, use of appropriate lube-and-dilution ratio,

730 / High-Pressure Die Casting examples cast in A356.2-T6 aluminum alloy are shown in Fig. 7.

Commonly Used Alloys, Properties, and Microstructures (Ref 1–5) Table 1 compares the tensile, hardness, and impact properties of select squeeze-cast aluminum alloys with those obtained from conventional casting processes (e.g., gravity permanent mold, or GPM, and conventional die casting). Key observations from Table 1 are: 

Fig. 6

For a given heat treatment process and aluminum alloy (A3256.2, 357, or 206), squeeze casting exhibits higher ductility than GPM; however, both processes reveal comparable strengths. The higher ductility of squeeze-cast components is attributed essentially to the lack/absence of porosity in the matrix. In addition, squeeze castings yield a more refined microstructure (more rounded eutectic silicon particles) than GPM. Typical microstructures of squeeze-cast A356.2-T6 and 390-T6 alloys are shown in Fig. 8 and 9, respectively.  Reduced porosity and refinement of microstructure are factors also responsible for squeeze-cast modified 383-F (F = as-cast) exhibiting higher tensile properties than conventional die-cast (high-pressure die cast, or HPDC) 383-F aluminum alloy.  In comparing the impact properties, it is evident that the 390-T6 aluminum alloy exhibits the lowest impact strength. The lack of ductility in this alloy (275 (>40)

2–4 (1.5–3.0) 7–10 (5.2–7.4) >700 (>102) >325 (>47)

38–42 (22–24) 11.5–13 (6.4–7.2) ... >125 (>18) >50 (>7.3) 7.0–7.1 5.2–5.4 (3.2–3.4) 0.45 (108  10 3) 80,000 (34) 50 (20)

35–38 (20–22) 11–13 (6.1–7.2) >200 (>29) ... ... 7.0–7.1 5.2–5.4 (3.2–3.4) 0.45 (108  10 3) 80,000 (34) 50 (20)

Tensile properties

Fatigue properties Rotating bending limit, unnotched, MPa (ksi)

 

Endurance ratio Fatigue strength reduction factor







Physical properties Thermal conductivity, 20–100  C (68–212  F), W/m. C (Btu/ft h.  F) Thermal expansion coefficient, 20–400  C (68–750  F), mm/m.  C (in./in. 1% creep strain limit, 10,000 at 350  C (660  F), MPa (ksi) at 400  C (750  F), MPa (ksi) at 500  C (930  F), MPa (ksi) Density, g/cm3 Ultrasonic velocity, km/s (miles/s) Specific heat capacity, 100  C (212  F), J/g.K (Btu/lb  F) Enthalpy, 26.6–200  C (79.9–392  F), J/kg (Btu/lb) Specific electrical resistance, 20  C (68  F), mO.cm (mO.in.)









F)




Table 2

Fig. 8

Stress-strain curves of pearlitic and ferritic compacted graphite iron castings. Straight line indicates modulus of elasticity 144 GPa (20.9  106 psi). Source: Ref 14

Properties of compacted graphite irons in accordance with ASTM A 842 Minimum tensile strength

Fig. 9

Correlation among hardness, tensile strength, and elongation in compacted graphite irons. Source: Ref 17

Minimum yield strength

Grade(a)

MPa

ksi

MPa

ksi

Elongation in 50 mm (2 in.), %

Hardness(b), HB

BID(b), mm

250(c) 300 350 400 450(d)

250 300 350 400 450

36.3 43.5 50.8 58.0 65.3

175 210 245 280 315

25.4 30.5 35.5 40.6 45.7

3.0 1.5 1.0 1.0 1.0

179 max 143–207 163–229 197–255 207–269

4.50 min 5.0–4.2 4.7–4.0 4.3–3.8 4.2–3.7

(a) Grades are specified according to the minimum tensile strength requirement given in MPa. Values in ksi are not part (b) Hardness range is nonmandatory information. Brinell impression diameter (BID) is the diameter (in mm) of the impression of a 10 mm diam ball at a load of 3000 kgf. (c) The 250 grade is a ferritic grade. Heat treatment to attain required mechanical properties and microstructure shall be the option of the manufacturer. (d) The 450 grade is a pearlitic grade usually produced without heat treatment with addition of certain alloys to promote pearlite as a major part of the matrix.

Table 3 Mechanical and physical properties of ISO 16112-grade compacted (vermicular) graphite iron Grade

GJV-300 GJV-350 GJV-400 GJV-4500 GJV-500




Tensile strength, MPa

Yield strength, MPa

Elongation, %

Modulus of elasticity, GPa

Thermal conductivity, W/m K

300–375 350–425 400–475 450–525 500–575

210–260 245–295 280–330 315–365 350–400

2.0–5.0 1.5–4.0 1.0–3.5 1.0–2.5 0.5–2.0

130–145 135–150 140–150 145–155 145–160

47 43 39 38 36




For all grades: Poisson’s ratio = 0.26, CTE = 11  10 6/K, specific heat capacity = 0.475 J/g K (100 C)

are worth noting. In general, CG iron has lower hardness than an FG iron of equivalent strength because of the higher amount of ferrite in the structure. For the same elongation, CG iron has considerably less yield strength than SG iron. Effect of Composition. The tensile properties of CG irons are much less sensitive to variations in CE than those of FG irons. Even at CE near the eutectic value of 4.3, both pearlitic and ferritic CG irons have higher strengths than does low-CE, high-duty, unalloyed FG cast iron (Fig. 11). An increase in the silicon content up to 2.6% benefits strength and hardness in both the as-cast and annealed conditions (Fig. 12). This is true even though the matrix becomes more ferritic, because silicon strengthens the ferrite. The same is true in the case of as-cast elongation, because there is a decrease in the amount of pearlite with an increase in silicon, while

876 / Cast Irons

Fig. 10

Tensile strength versus hardness for spheroidal graphite (SG), compacted graphite (CG), and flake graphite (FG) irons. Source: Ref 18

Fig. 11

Effect of carbon equivalent on the tensile strength of spheroidal, compacted, and flake graphite irons cast in 30 mm (1.2 in.) diameter bars. Source: Ref 3

elongation in the annealed condition decreases (Ref 18). Although increasing the phosphorus content slightly improves strength, a maximum of 0.04% P is desirable to avoid lower ductility and impact strength. The pearlite/ferrite ratio, and thus the strength and hardness of CG irons, can be increased by the use of a number of alloying elements, such as copper, nickel, molybdenum, tin, manganese, arsenic, vanadium, and aluminum (Ref 7, 8, 10, Ref 19–21). After annealing to a fully ferritic structure, it is possible to increase the yield point of CG iron by 24% when using 1.5% Ni (Table 4). This is because of the strengthening of the solid solution by nickel (Ref 4). However, additions of copper, nickel, and molybdenum may increase nodularity (Ref 16). Figure 13 shows some correlations of microstructure and properties for iron-carbonaluminum CG irons produced by the in-mold process. Effect of Structure. One of the most important variables influencing the tensile properties of CG irons is nodularity. As nodularity increases, higher strength and elongation are to be expected, as shown in Table 5 and Fig. 14 and 15, although nodularity must be maintained at levels under 20% for the iron to qualify as CG iron. However, spheroidal graphite contents of up to 30% and even more must be expected in thin sections of castings with considerable variation in wall thickness. As previously discussed, the pearlite/ferrite ratio can be increased by using alloying elements. Another way of increasing or decreasing this ratio is by using heat treatment.

Table 4 Effect of heat treatment and alloying with nickel on the tensile properties of compacted graphite iron measured on a 25 mm (1 in.) section size Tensile strength

Yield strength

Heat treatment

Iron matrix(a)

MPa

ksi

MPa

ksi

Elongation, %

Hardness, HB

Nickel, %

As-cast Annealed(b) Normalized(c) As-cast Annealed(b) Normalized(c)

60% F 100% F 90% P ... 100%F 90% P

325 294 423 427 333 503

47.1 42.6 61.3 61.9 48.3 73

263 231 307 328 287 375

38.1 33.5 44.5 47.6 41.6 54.4

2.8 5.5 2.5 2.5 6.0 2.0

153 121 207 196 137 235

0 0 0 1.53 1.53 1.53

(a) F, ferrite; P, pearlite. (b) Annealed, 2 h at 900  C (1650  F), cooled in furnace to 690  C (1275  F), held 12 h, cooled in air. (c) Normalized, 2 h at 900  C (1650  F), cooled in air

Fig. 12

Effect of silicon on mechanical properties of compacted graphite irons produced by the in-mold process. Carbon equivalents range from 4.33 to 4.45. Source: Ref 18

Compacted Graphite Iron Castings / 877

Fig. 13

Effect of aluminum on the structure and mechanical properties of compacted graphite irons produced by the in-mold process. Source: Ref 10

Table 5

Properties of compacted graphite iron as a function of nodularity Tensile strength

Nodularity, %

10–20 20–30 40–50

MPa

ksi

Elongation, %

320–380 380–450 450–500

46–55 55–65 65–73

2–5 2–6 3–6

Thermal conductivity, W/(m

50–52 48–50 38–42


K)

Shrinkage, %

1.8–2.2 2.0–2.6 3.2–4.6

Source: Ref 22

increased nodularity and higher chilling must be considered. This is particularly true for overtreated irons. While it is possible to eliminate the carbides that result from chilling by heat treatment, it is impossible to change the graphite shape, which remains spheroidal with the associated consequences. Other factors influencing the cooling of castings, such as shakeout temperature, can also influence properties. Thin-wall castings (2.5 to 4 mm, or 0.1 to 0.16 in.) are of particular interest to designers of engines in the transportation industry. This tendency of increased nodularity in thin sections can work to the advantage of the design. In engine block design, for instance, the thick section may contain 5 to 15% nodularity, while the thin sections (1.85). Gray iron is considerably less notch sensitive, with a notch factor of less than 1.5 (Ref 16). Statistical analysis of a number of experimental data allowed the calculation of a relationship between fatigue strength and tensile strength measured in MPa of CG irons (Ref 16).

Another study (Ref 27) shows similar relative results with higher stress amplitudes (Fig. 23b). A 36% decrease of the alternating bending fatigue strength has been observed on unnotched bars with casting skin compared to machined bars. This compares with a 50% reduction for FG iron of comparable strength and a 32% reduction for ferritic SG iron.

Cast irons are generally good in fresh water and atmospheric conditions. Seawater presents problems due to galvanic action between carbon and the matrix. No grade or type of cast iron will resist all corrosive environments. Many types of cast irons are used in acidic environments. Unalloyed cast irons often behave well in alkalis. Caution should be used when comparing corrosion rates for uniform corrosion, because short-term tests may not reflect long-term performance. Also, localized corrosion, such as graphitic corrosion in which the carbon is selectively leached, or stress-corrosion cracking may be more degrading forms of corrosion. Detailed information on the corrosion resistance of cast irons is available in Ref 28.

Applications The applications of CG irons stem from their intermediate position between FG and SG irons. Compared to FG irons, CG irons have certain advantages:  Higher tensile strength at the same carbon

equivalent, which reduces the need for expensive alloying elements such as nickel, chromium, copper, and molybdenum  Higher ratio of tensile strength to hardness  Much higher ductility and toughness, which result in a higher safety margin against fracture

880 / Cast Irons Modern car and truck engines require that manifolds work at temperatures of 500  C (930  F). At this temperature, FG iron manifolds are prone to cracking, while SG iron manifolds tend to warp. Compacted graphite iron manifolds warp and oxidize less and thus have a longer life. ACKNOWLEDGMENTS This article is adapted from “Metallurgy and Properties of Compacted Graphite Iron,” ASM Specialty Handbook: Cast Irons, which in turn was based on:  D.M. Stefanescu, Compacted Graphite Iron,

Properties and Selection: Irons, Steels, and High-Performance Alloys, Vol 1, ASM Handbook, ASM International, 1990, p 56–70  D.M. Stefanescu, R. Hummer, and E. Nechtelberger, Compacted Graphite Irons, Casting, Vol 15, ASM Handbook (formerly Metals Handbook 9th ed.), ASM International, 1988, p 667–677

REFERENCES

Fig. 21

Effect of pearlite content on the Charpy V-notch impact strength for as-cast compacted graphite (CG) irons compared to spheroidal graphite (SG) iron at room temperature (21  C, or 70  F). Source: Ref 18

 Lower

oxidation and growth at high temperatures  Less section sensitivity for heavy sections Compared to SG irons, the advantages of CG irons are:     

Lower coefficient of thermal expansion Higher thermal conductivity Better resistance to thermal shock Higher damping capacity Better castability, leading to higher casting yield and the capability for pouring more intricate castings  Improved machinability Compacted graphite iron can be substituted for FG iron in all cases in which the strength of FG iron has become insufficient but in which a change to SG iron is undesirable because of the less favorable casting properties of the latter. Examples include bed plates for large diesel engines, crankcases, gearbox housings, turbocharger housings, connecting forks, bearing brackets, pulleys for truck servodrives, sprocket wheels, and eccentric gears.

Because the thermal conductivity of CG iron is higher than that of SG iron, CG iron is preferred for castings operating at elevated temperature and/or under thermal fatigue conditions. Applications include ingot molds, crankcases, cylinder heads, exhaust manifolds, and brake disks. The largest industrial application by weight of CG iron produced is for ingot molds weighing up to 54 metric tonnes (60 tons). According to a number of reports summarized in Ref 29, the life of ingot molds made of CG iron is 20 to 70% longer than the life of those made of FG iron. In the case of cylinder heads, it was possible to increase engine output by 50% by changing from alloyed FG iron to ferritic CG iron (Ref 1). The specified minimum values for cylinder heads are 300 MPa (43 ksi) tensile strength, 240 MPa (35 ksi) yield strength, and 2% elongation. New emission standards and the need for higher-efficiency diesel engines have increased the requirements of cylinder liners to 340 to 400 MPa (49 to 58 ksi) with a modulus of elasticity of 130 to 150 GPa (19 to 22  106 psi) in order to withstand the peak pressures.

1. R.D. Schelleng, Cast Iron (with Vermicular Graphite), U.S. Patent 3,421,886, May 1965 2. E.R. Evans, J.V. Dawson, and M.J. Lalich, Compacted Graphite Cast Irons and Their Production by a Single Alloy Addition, Trans. AFS, Vol 84, 1976, p 215 3. G.F. Sergeant and E.R. Evans, The Production and Properties of Compacted Graphite Irons, Br. Foundryman, Vol 71, 1978, p 115 4. K.P. Cooper and C.R. Loper, Jr., A Critical Evaluation of the Production of Compacted Graphite Cast Iron, Trans. AFS, Vol 86, 1978, p 267 5. H.H. Cornell and C.R. Loper, Jr., Variables Involved in the Production of Compacted Graphite Cast Iron Using Rare EarthContaining Alloys, Trans. AFS, Vol 93, 1985, p 435 6. G.F. Ruff and T.C. Vert, Investigation of Compacted Graphite Iron Using a High Sulfur Gray Iron Base, Trans. AFS, Vol 87, 1979, p 459 7. J. Fowler, D.M. Stefanescu, and T. Prucha, Production of Ferritic and Pearlitic Grades of Compacted Graphite Cast Iron by the In-Mold Process, Trans. AFS, Vol 92, 1984, p 361 8. K.R. Ziegler and J.F. Wallace, The Effect of Matrix Structure and Alloying on the Properties of Compacted Graphite Iron, Trans. AFS, Vol 92, 1984, p 735 9. E. Nechtelberger, R. Hummer, and W. Thury, Aluminum- Alloyed Cast Iron with Vermicular Graphite, GiessereiPrax., Vol 24, 1970, p 387

Compacted Graphite Iron Castings / 881

Fig. 22

Dynamic tear energy versus temperature from 100 to 100  C ( 150 to 212  F). (a) compacted graphite (CG) iron. (b) spheroidal graphite (SG) iron. Source: Ref 18

Table 7 Impact toughness of a cerium-treated compacted graphite (CG) cast iron and two magnesium-titanium-treated CG cast irons Structural condition and graphite type

Iron

Cerium-treated

>95 % ferrite (as-cast); 95% CG, 5% SG

Magnesium (0.018%) and 100% ferrite (annealed); CG titanium (0.089%) treated Magnesium (0.017%) and Ferritic (as-cast); CG titanium (0.062%) treated

Test temperature 

C



F

Impact bend toughness(a) J

Notched-bar(b) impact toughness




ft lbf




J

ft lbf

20 20 40 20

68 4 40 68

32.1 23.7 26.5 19.5 26.7 19.7 13.5–19 10–14

6.5 4.6 5.0 5.4

4.8 3.4 3.7 4.0

20

68

6.8–10.2 5–7.5

3.4

2.5

(a) Unnotched 10 by 10 mm (0.4 by 0.4 in) (Charpy) testpiece. (b) V-notched 10 by 10 mm (0.4 by 0.4 in.) (Charpy) testpiece. Source: Ref 16

Table 8 Fatigue strengths and endurance ratios for five compacted graphite (CG) irons from rotating-bending tests Tensile strength Matrix structure

As-cast As-cast As-cast As-cast As-cast

ferrite (>95% ferrite) ferrite pearlite pearlite pearlite (70% pearlite)

SG, spheroidal graphite. Source: Ref 16

Fatigue strength

Graphite type

MPa

ksi

MPa

ksi

Fatigue endurance ratio

Hardness, HB

95% CG CG CG CG + SG CG

336 388 414 473 386

48.7 56.3 60 68.6 56.0

211 178 185 208 186

30.6 25.8 26.8 30.2 27

0.63 0.46 0.45 0.44 0.48

150 184 205 217 ...

10. F. Martinez and D.M. Stefanescu, Properties of Compacted/Vermicular Graphite Cast Irons in the Fe-C-Al System Produced by Ladle and In-Mold Treatment, Trans. AFS, Vol 91, 1983, p 593 11. R. Hummer, unpublished research at the Austrian Foundry Research Institute, A.-Nr. 31.253, 1987 12. G. Nandori and J. Dul, Untersuchungen Uber den Abklingeffekt bei Gusseisen mit Kugelgraphit Durch Messung der La¨ngen—und Temperatura¨nderung wahrend der Erstarrung, Giesserei-Prax., No. 18, 1978, p 284 13. D.M. Stefanescu, I. Dinescu, S. Craciun, and M. Popescu, “Production of Vermicular Graphite Cast Irons by Operative Control and Correction of Graphite Shape,” Paper 37, presented at the 46th International Foundry Congress (Madrid, Spain), 1979 14. E. Nechtelberger, H. Puhr, J.B. von Nesselrode, and A. Nakayasu, “Cast Iron with Vermicular/Compacted Graphite—State of the Art Development, Production, Properties, Applications,” paper presented at the 49th International Foundry Congress (Chicago, IL), April 1982 15. D.M. Stefanescu, F. Martinez, and I.G. Chen, Solidification Behavior of Hypoeutectic and Eutectic Compacted Graphite Cast Irons—Chilling Tendency and Eutectic Cells, Trans. AFS, Vol 91, 1983, p 205 16. E. Nechtelberger, The Properties of Cast Iron up to 500  C, Technicopy Ltd., 1980 17. C.F. Walton and T.J. Opar, Ed., Iron Castings Handbook, Iron Casting Society Inc., 1981, p 381–397 18. K.P. Cooper and C.R. Loper, Jr., Trans. AFS, Vol 86, 1978, p 241 19. C.R. Loper, M.J. Lalich, H.K. Park, and A.M. Gyarmaty, “Microstructure-Mechanical Property Relationship in Compacted/ Vermicular Graphite Cast Iron,” Paper 35, presented at the 46th International Foundry Congress, (Madrid, Spain), 1979 20. N.N. Aleksandrov, B.S. Milman, L.V. Ilicheva, M.G. Osada, and V.V. Andreev, Production and Properties of High-Duty Iron with Compacted Graphite, Russ. Cast. Prod., Aug 1976, p 319 21. R.B. Gundlach, Trans. AFS, Vol 86, 1978, p 551 22. Spravotchnik po Tchugunomu Ljitiu [Cast Iron Handbook], 3rd ed., li Mashinostrojenie, 1978 23. S. Dawson, “Compacted Graphite Iron: Mechanical and Physical Properties for Engine Design,” Materials in Powertrain VDI, Dresden, Germany, Oct 28–29, 1999 24. K.H. Riemer, Giesserei, Vol 63 (No. 10), 1976, p 285 25. K.B. Palmer, BCIRA J., Report 1213, Jan 1976, p 31 26. A.F. Heiber, Trans. AFS, Vol 87, 1979, p 569

882 / Cast Irons

Fig. 24

Correlation between ultrasonic velocity and nodularity. SG, spheroidal graphite, CG, compacted graphite, FG, flake graphite. Source: Ref 7

27. M. Shidida, Y. Kanayama, and H. Nakayama, Strength and Crack Growth Behaviors of Compacted Graphite Vermicular Cast Iron in Rotating Bending, 29th Japan Congress on Materials Research, 1986, p 23–28 28. T.C. Spence, Corrosion of Cast Irons, Corrosion: Materials, Vol 13B, ASM Handbook, ASM International, 2005, p 43–50 29. D.M. Stefanescu and C.R. Loper, Jr., Giesserei-Prax., No. 5, 1981, p 74

Fig. 23

Fatigue (S-N) curves for ferritic, pearlitic, and higher-nodularity compacted graphite (CG) (Table 7) from rotating-bending tests. Arrows indicate specimens did not fail. SG, spheroidal graphite. (a) Source: Ref 25. (b) Source: 27

Compacted Graphite Iron Castings / 883

Fig. 25

Tensile strength related to (a) ultrasonic velocity and (b) resonant frequency for cast irons of various graphitic structures. SG, spheroidal graphite. Source: Ref 3

ASM Handbook, Volume 15: Casting ASM Handbook Committee, p 884-895 DOI: 10.1361/asmhba0005326

Copyright © 2008 ASM International® All rights reserved. www.asminternational.org

Malleable Iron Castings Revised by C.J. van Ettinger, Gieterij Doesburg

MALLEABLE IRON is a cast ferrous metal that is initially produced as white cast iron and is then heat treated to convert the carbon-containing phase from iron carbide to a nodular form of graphite called temper carbon. The history of malleable iron goes back to Europe. The French scientist and metallurgist R.A.F. de Re´aumur published the first technical discussions of malleable iron castings (720 to 1722). In America, the inventor Seth Boyden introduced malleable iron (Ref 1). Boyden knew of the whiteheart malleable iron of Europe and foresaw a potential market for such a material; he started a series of experiments to find out how to produce malleable iron. Because of the chemical composition of the pig iron used, Boyden found the fractures of his specimens “black and grey” (blackheart) because of the presence of free carbon after the heat treatment process. In 1831, he was granted the first American patent for malleable iron; subsequently, the production of castings began in his foundry in Newark, New Jersey. There are two types of ferritic malleable iron: blackheart and whiteheart. Only the blackheart type is produced in the United States. This material has a matrix of ferrite with interspersed nodules of temper carbon. Cupola malleable iron is a blackheart malleable iron that is produced by cupola melting and is used for pipe fittings and similar thin- section castings. Because of its low strength and ductility, cupola malleable iron is usually not specified for structural applications. Pearlitic malleable iron is designed to have combined carbon in the matrix, resulting in higher strength and hardness than ferritic malleable iron. Martensitic malleable iron is produced by quenching and tempering pearlitic malleable iron. Malleable iron, like ductile iron, possesses considerable ductility and toughness because of its combination of nodular graphite and low-carbon metallic matrix. Because of the way in which graphite is formed in malleable iron, however, the nodules are not truly spherical, as they are in ductile iron, but are irregularly shaped aggregates. Malleable iron and ductile iron are used for some of the same applications in which ductility and toughness are important. In many cases, the choice between malleable and ductile iron is based on economy or availability rather than on properties. In certain applications, however,

malleable iron properties have a distinct advantage. It is preferred for thin-section castings; for parts that are to be pierced, coined, or cold formed; for parts requiring maximum machinability; for parts that must retain good impact resistance at low temperatures; and for some parts requiring wear resistance (martensitic malleable iron only). Ductile iron has a clear advantage where low solidification shrinkage is needed to avoid hot tears or where the section is too thick to permit solidification as white iron. (Solidification as white iron throughout a section is essential to the production of malleable iron.) Malleable iron castings are produced in section thicknesses ranging from approximately 1.5 to 100 mm (0.06 to 4 in.) and in weights from less than 0.03 to 180 kg (1 oz to 400 lb) or more. Composition. The chemical composition of malleable iron generally conforms to the ranges given in Table 1. Alloying elements such as copper (max 1.0%), nickel (max 0.8%), and molybdenum (max 0.5%) are also sometimes present. When the alloy elements are not used, then all other elements besides carbon, silicon, manganese, and sulfur should be called trace elements. Trace elements such as chromium, tin, titanium, vanadium, aluminum, copper, and molybdenum should be kept as low as possible and as consistent as possible, because they will retard or promote the graphitization process.

Melting Practices Melting can be accomplished by batch cold melting or by duplexing. Cold melting is done in coreless or channel-type induction furnaces, electric arc furnaces, or cupola furnaces. In duplexing, the iron is melted in a cupola or electric arc furnace, and the molten metal is transferred to a coreless or channel-type induction furnace for holding and pouring. Charge materials (foundry returns, steel scrap, ferroalloys, and, except in cupola melting, carbon) are carefully selected, and the melting operation is well controlled to produce metal having the desired composition and properties. Minor corrections in composition and pouring temperature

are made in the second stage of duplex melting, but most of the process control is done in the primary melting furnace. Molds are produced in green sand, silicate CO2-bonded sand, or resin-bonded sand (shell molds) on equipment ranging from highly mechanized or automated machines to that required for floor or hand molding methods, depending on the size and number of castings to be produced. Cores are generally produced in cold-box-type bonded sand on highly automated core-shooters. Cores are usually coated with a core-wash to protect the core sand against penetration. In general, the technology of molding and pouring malleable iron is similar to that used to produce gray iron. Heat treating is done in high-production controlled-atmosphere continuous furnaces or batch-type furnaces, again depending on production requirements. After it solidifies and cools, the metal is in a white iron state, and gates, sprues, and feeders can be easily removed from the castings by impact. This operation, called spruing, is generally performed manually with a hammer because the diversity of castings produced in the foundry makes the mechanization or automation of spruing very difficult. After spruing and shotblasting, the castings proceed to heat treatment, while gates and risers are returned to the melting department for reprocessing. Malleable iron castings are produced from the white iron by a two-stage annealing process. First- and second-stage annealing processes are described in detail in the section “Control of Annealing” in this article. After heat treatment, ferritic or pearlitic malleable castings are cleaned by shotblasting, gates are removed by shearing or grinding, and, where necessary, the castings are coined

Table 1 Chemical composition of malleable iron Element

Carbon Silicon Manganese Sulfur Phosphorus

Composition, %

2.16–2.90 0.90–1.90 0.15–1.25 0.02–0.20 0.01–0.05

Malleable Iron Castings / 885 or punched. Close dimensional tolerances can be maintained in ferritic malleable iron and in the lower-hardness types of pearlitic malleable iron, both of which can be easily straightened in dies. The harder pearlitic malleable irons are more difficult to press because of higher yield strength and a greater tendency toward springback after die pressing. However, even the highest-strength pearlitic malleable can be straightened to achieve good dimensional tolerances. Control of Melting. Metallurgical control of the melting operation is designed to ensure that the molten iron will have a certain composition and will:
Solidify white in the castings to be produced
Anneal on an established time-temperature

cycle set to minimum values in the interest of economy
Produce the desired graphite distribution (nodule count) upon annealing Changes in melting practice or composition that would satisfy the first requirement listed previously are generally opposed to satisfaction of the second and third, while attempts to improve annealability beyond a certain point may result in an unacceptable tendency for the as-cast iron to be mottled instead of white. The common elements in malleable iron are generally controlled within approximately  þ0.05 to  þ0.1%, using emission spectrometers. Thermal analysis is used for controlling the white solidification of the iron. A limiting minimum carbon content is required in the interest of mechanical quality and annealability, because decreasing carbon content reduces the fluidity of the molten iron, increases shrinkage during solidification, and reduces annealability. A limiting maximum carbon content is imposed by the requirement that the casting be white as-cast. The range in silicon content is limited to ensure proper annealing during a short-cycle high-production annealing process and to avoid the formation of primary graphite during solidification. Manganese and sulfur contents are balanced to ensure that all sulfur is combined with manganese and that only a safe, minimum quantity of excess manganese is present in the iron. An excess of either sulfur or manganese will retard annealing in the second stage and therefore increase annealing costs. The chromium content is kept low because of the carbide-stabilizing effect of this element and because it retards both the first-stage and secondstage annealing reactions. A mixture of gray iron and white iron in variable proportions that produces a mottled (speckled) appearance is particularly degrading to the mechanical properties of the annealed casting, whether ferritic or pearlitic malleable iron. Primary control of mottle is achieved by maintaining a balance of carbon and silicon contents. Because economy and castability are enhanced when the carbon and silicon contents of the base iron are in the higher portions of their respective ranges, some malleable iron

foundries produce iron with carbon and silicon contents at levels that may produce mottle and then add a balanced, mild carbide stabilizer to prevent mottle during casting. Bismuth and boron in balanced amounts accomplish this control. A typical addition is 0.01% Bi (as metal) and 0.001% B (as ferroboron). Bismuth retards graphitization during solidification; small amounts of boron have little effect on graphitizing tendency during solidification but accelerate carbide decomposition during annealing. The balanced addition of bismuth and boron permits the production of heavier sections for a given base iron or the utilization of a higher-carbon higher-silicon base iron for a given section thickness. Tellurium can be added in amounts of 0.0005 to 0.001% to suppress mottle. Tellurium is a much stronger carbide stabilizer than bismuth during solidification but also strongly retards annealing if the residual exceeds 0.003%. Less than 0.003% residual tellurium has little effect on annealing but has a significant influence on mottle control. Tellurium is more effective if added together with copper or bismuth. Because tellurium is toxic, it is no longer used in the foundries. Residual boron should not exceed 0.0035% in order to avoid nodule alignment and carbide formation. Also, the addition of 0.005% Al to the pouring ladle significantly improves annealability without promoting mottle.

the metallic matrix as lamellar pearlite, tempered martensite, or spheroidite. Molten iron produced under properly controlled melting conditions solidifies with all carbon in the combined form, producing the white iron structure fundamental to the manufacture of either ferritic or pearlitic malleable iron (Fig. 2). The base iron must contain balanced quantities of carbon and silicon to simultaneously provide castability, white iron in even the thickest sections of the castings, and annealability; therefore, precise metallurgical control is necessary for quality production. Thick metal sections cool slowly during solidification and tend to graphitize, producing mottled or gray iron. This is undesirable, because the graphite formed in mottled iron or rapidly cooled gray iron is generally of the type D configuration, a flake form in a dense, lacy structure, which is particularly damaging to the strength, ductility, and stiffness characteristics of both ferritic and pearlitic malleable iron. Control of Nodule Count. Proper annealing in short-term cycles and the attainment of high levels of casting quality require that controlled distribution of graphite particles be obtained during first-stage heat treatment. With low

Microstructure Malleable iron is characterized by microstructures consisting of uniformly dispersed fine particles of free carbon in a matrix of ferrite or tempered martensite. These microstructures can be produced in base metal of essentially the same composition. Structural differences between ferritic malleable iron and the various grades of pearlitic or martensitic malleable iron are achieved through variations in heat treatment. The microstructure of a casting of any type of malleable iron is derived by controlled annealing of white iron of suitable composition. During the annealing cycle, carbon that exists in combined form, either as massive carbides or as a microconstituent in pearlite, is converted to a form of free graphite known as temper carbon. Ferritic malleable iron requires a two-stage annealing cycle. The first stage converts primary carbides to temper carbon, and the second stage converts the carbon dissolved in austenite at the first-stage annealing temperature to temper carbon and ferrite. The microstructure of ferritic malleable iron is shown in Fig. 1. A satisfactory structure consists of temper carbon in a matrix of ferrite. There should be no flake graphite and essentially no combined carbon in ferritic malleable iron. Pearlitic and martensitic malleable irons contain a controlled quantity of combined carbon, which, depending on heat treatment, may appear in

Fig. 1

Structure of annealed ferritic malleable iron showing temper carbon in ferrite. Original magnification: 200

Fig. 2

Structure of as-cast malleable white iron showing a mixture of pearlite and eutectic carbides. Original magnification: 200

886 / Cast Irons nodule count (few graphite particles per unit area or volume), mechanical properties are reduced from optimum, and second-stage annealing time is unnecessarily long because of long diffusion distances. Excessive nodule count is also undesirable, because graphite particles may become aligned in a configuration corresponding to the boundaries of the original primary cementite. In martensitic malleable iron, very high nodule counts are sometimes associated with low hardenability and nonuniform tempering. Generally, a nodule count of 80 to 150 discrete graphite particles per square millimeter (80 to 150 in 15.5 in.2 of a micrograph at 100) appears to be optimum. This produces random particle distribution, with short distances between particles. Temper carbon is formed predominantly at the interface between primary carbide and saturated austenite at the first-stage annealing temperature, with growth around the nuclei taking place by a reaction involving diffusion and carbide decomposition. Although new nuclei undoubtedly form at the interfaces during holding at the first-stage annealing temperature, nucleation and graphitization are accelerated by the presence of nuclei that are created by appropriate melting practice. High silicon and carbon contents promote nucleation and graphitization, but these elements must be restricted to certain maximum levels because of the necessity that the iron solidify white. Control of Annealing. The rate of annealing of a hard iron casting depends on chemical composition, nucleation tendency (as discussed previously), and annealing temperature. With the proper balance of boron content and graphitic materials in the charge, optimum number and distribution of graphite nuclei are developed in the early part of first-stage annealing, and growth of the temper carbon particles proceeds rapidly at any annealing temperature. An optimum iron will anneal completely through the first-stage reaction in approximately 3 h at 940  C (1720  F). Irons with lower silicon contents or less-than-optimum nodule counts may require as much as 20 h for completion of first-stage annealing. The temperature of first-stage annealing exercises considerable influence on the rate of annealing and the number of graphite particles produced. Increasing the annealing temperature accelerates the rate of decomposition of primary carbide and produces more graphite particles per unit volume. However, high first-stage annealing temperatures can result in excessive distortion of castings during annealing, which leads to straightening of the casting after heat treatment. Annealing temperatures are adjusted to provide maximum practical annealing rates and minimum distortion and are therefore controlled within the range of 900 to 970  C (1650 to 1780  F). Lower temperatures result in excessively long annealing times, while higher temperatures produce excessive distortion. After first- stage annealing, the castings are cooled as rapidly as practical to 740 to

760  C (1360 to 1400  F) in preparation for second-stage annealing. The fast cooling step requires 1 to 6 h, depending on the equipment being employed. Castings are then cooled slowly at a rate of approximately 3 to 11  C (5 to 20  F) per hour. During cooling, the carbon dissolved in the austenite is converted to graphite and deposited on the existing particles of temper carbon. This results in a fully ferritic matrix. In the production of pearlitic malleable iron, the first-stage heat treatment is identical to that used for ferritic malleable iron. However, some foundries then slowly cool the castings to approximately 870  C (1600  F). During cooling, the combined carbon content of the austenite is reduced to approximately 0.75%, and the castings are then air cooled. Air cooling is accelerated by an air blast to avoid the formation of ferrite envelopes around the temper carbon particles (bull’seye structure) and to produce a fine pearlitic matrix (Fig. 3). The castings are then tempered to specification, or they are reheated to reaustenitize at approximately 870  C (1600  F), oil quenched, and tempered to specification. Large foundries usually eliminate the reaustenitizing step and quench the castings in oil directly from the first-stage annealing furnace after stabilizing the temperature at 845 to 870  C (1550 to 1600  F). The furnace atmosphere for producing malleable iron in continuous furnaces is controlled so that the ratio of CO to CO2 is between 1:1 and 20:1. In addition, any sources of water vapor or hydrogen are eliminated; the presence of hydrogen is thought to retard annealing, and it produces excessive decarburization of casting surfaces. Proper control of the gas atmosphere is important for avoiding an undesirable surface structure. A high ratio of CO to CO2 retains a high level of combined carbon on the surface of the casting and produces a pearlitic rim, or picture frame, on a ferritic malleable iron part. A low ratio of CO to CO2 permits excessive decarburization, which forms a ferritic skin on the casting with an underlying rim of pearlite. The latter condition is produced when a significant portion of the subsurface metal is decarburized to the degree that no temper carbon nodules can be developed during first-stage annealing. When this occurs, the dissolved carbon cannot precipitate from the austenite, except as the cementite plates in pearlite. An alternative is the use of nitrogen to control the furnace atmosphere, which is common in European foundries. The atmosphere is controlled by measuring the amount of CO2, which is normally held below 0.25%. The advantage of using nitrogen is the extreme small decarburization rim of just 0.2 mm (0.008 in.). The rate of cooling after first-stage annealing is important in the formation of a uniform pearlitic matrix in the air-cooled casting, because slow rates permit partial decomposition of carbon in the immediate vicinity of the temper carbon nodules, which results in the formation of films of ferrite around the temper carbon (bull’s-eye structure). When the extent of these films becomes excessive, a carbon gradient is

developed in the matrix. Air cooling is usually done at a rate not less than approximately 80  C (150  F) per minute. Air-quenched malleable iron castings have hardnesses ranging from 269 to 321 HB, depending on casting size and cooling rate. Such castings can be tempered immediately after air cooling to obtain pearlitic malleable iron with a hardness of 241 HB or less. High-strength malleable iron castings of uniformly high quality are usually produced by liquid quenching and tempering, using any of the three procedures. The most economical procedure is direct quenching after first-stage annealing. In this procedure, castings are cooled in the furnace to the quenching temperature of 845 to 870  C (1550 to 1600  F) and held for 15 to 30 min to homogenize the matrix. The castings are then quenched in agitated oil to develop a matrix microstructure of martensite having a hardness of 415 to 601 HB. Finally, the castings are tempered at an appropriate temperature between 590 and 725  C (1100 and 1340  F) to develop the specified mechanical properties. The final microstructure consists of tempered martensite plus temper carbon, as shown in Fig. 4. In heavy sections, higher-temperature

Fig. 3

Structure of air-cooled pearlitic malleable iron. Cooled in an air blast. Original magnification:

200

Fig. 4 200

Structure of oil-quenched and tempered martensitic malleable iron. Original magnification:

Malleable Iron Castings / 887 transformation products such as fine pearlite are usually present. Some foundries produce high-strength malleable iron by an alternative procedure in which the castings are forced-air cooled after first-stage annealing, retaining approximately 0.75% C as pearlite, and then reheated to 840 to 870  C (1545 to 1600  F) for 15 to 30 min, followed by quenching and tempering as described previously for the direct-quench process. Rehardened-and-tempered malleable iron can also be produced from fully annealed ferritic malleable iron with a slight variation from the heat treatment used for arrested-annealed (airquenched) malleable. The matrix of fully annealed ferritic malleable iron is essentially carbon free but can be recarburized by heating at 840 to 870  C (1545 to 1600  F) for 1 h. In general, the combined carbon content of the matrix produced by this procedure is slightly lower than that of arrested-annealed pearlitic malleable iron, and the final tempering temperatures required for the development of specific hardnesses are lower. Rehardened malleable iron made from ferritic malleable may not be capable of meeting certain specifications. Tempering times of 2 h or more are needed for uniformity. In general, control of final hardness of the castings is precise, with process limitations approximately the same as those encountered in the heat treatment of mediumor high-carbon steels. This is particularly true when specifications require hardnesses of 241 to 321 HB where control limits of  þ0.2 mm Brinell diameter can be maintained with ease. At lower hardnesses, a wider process control limit is required because of certain unique characteristics of the pearlitic malleable iron microstructure. The relationships between tempering conditions and yield strength and hardness properties are given in Fig. 5.

liquidus arrest. Curve A is the normal liquidus arrest of a cooling curve. The solidification starts at 1285  C (2345  F) at approximately 18 to 20 s after pouring the metal into the cup. Curve B shows a little arrest up front, the (normal) liquidus arrest. It seems that the solidification starts at 1310  C (2390  F) and approximately 12 s. This is a 25  C (45  F) higher solidification temperature and approximately 6 s earlier. Because of the solidification simulation technology used today (2008), it is essential to know how solidification behaves. This information is the basis for updating simulation technology. Simulation Technologies. Many foundries use MagmaSoft (Magma Gmbtt) software for the simulation of the casting process. The recent information from the cooling curve is transferred into the solidification simulation process. This means that the solidification simulation is not based with a calculated solidification temperature out of a chemical analysis but with a real solidification temperature, determined by the cooling curve. This step made the simulation process accurate and reliable. The simulation of a casting is done in three steps. The first step is the simulation of the part itself. This illustrates the temperature range, modules, and the shrinkage or the expected porosity areas. The second step is the simulation of the mold filling (Fig. 7). This displays the effect of the mold filling on the simulation result done previously. It shows changes in temperatures, modules, and shrinkage areas. At last, the complete gating and feeding system is simulated, which leads to the most accurate result. After a positive result, the tools are built, and the prototype parts are produced and checked with destructive testing and x-ray.

Fig. 5

Hardness and minimum yield strength of pearlitic malleable iron. Relationships of tempering time and temperature to hardness and minimum yield strength are given.

Current Production Technologies Digital Solidification Analysis Technology. Already in the 1970s, the American Foundry Society reported the importance of the use of thermal analysis during solidification. Leon, Ekpoon, and Heine (Ref 2) showed that the difference in carbon equivalent as determined by the cooling curve and by chemical analysis indicated a relation to the degree of oxidation of the liquid metal. Jacobs reported a relationship between the results of the thermal analysis and the amount of foundry scrap (Ref 3). Recent research shows the same relationship. However, the deviation in the liquidus arrest of the cooling curve was not found, but the digital cooling curve shows a first-arrest up front, the so called liquidus arrest, that comes and goes under the different melting circumstances. This phenomenon is used to control the method of charging and melting. The cooling curve can also be helpful for predicting the type of solidification structure (Ref 4). Figure 6 shows the difference in the cooling curve around the

Fig. 6

Difference in cooling curve around the liquidus arrest of a white cast iron (base malleable iron) caused by melting conditions. Curve A is a typical normal cooling curve of a solid-solution alloy with a liquidus temperature of approximately 1285  C (2345  F). Curve B shows there is a transformation occurring earlier, at approximately 1310  C (2390  F), due to different melting circumstances. Source: Ref 4

888 / Cast Irons Compared to malleable iron, which is treated in terms of tons per hour, ADI is treated in low volume. The costs are also increased by the additional transport costs to and from the heat treatment company, while malleable iron is treated in-house. The total costs are close for standard ductile iron and malleable iron, according to this new concept. Table 2 shows the high costs of the old malleable iron but also the costs of ADI, which is more than double that of standard ductile. The costs for ADI are based on the present (2008) costs of nickel of $37/kg and copper of $6/kg at an addition rate of 0.75%.

Applications

Fig. 7

Simulation of mold filling of a piston dome

Table 2 Comparison of costs Process cost(a), % Item

Alloy Melting Heat treatment Sum Casting yield(d)

Ductile standard

Malleable old

Malleable new concept(b)

Ductile ADI(c)

84 16 0 100 65

103 22 48 173 40

71 14 28 113 65

136 16 68 220 65

(a) Standard dutile iron total (100%) is base. (b) Simulation and smart engineering practices as discussed in text. (c) ADI, austempered ductile iron. (d) Mass of casting(s) divided by the total mass of metal poured into the mold, expressed as a percentage. Source: Ref 4

Smart engineering is nothing more than positive and simple thinking. Smart engineering focuses on the casting. It is not thought of as part of the complete gating system. It is important to judge the value of existing knowledge when using new technology. Knowing the results of simulation provides information about what is useful and what is useless knowledge of gating and risering. In the past, engineers made dozens of calculations and also used their past experiences to develop a gating and risering system. The results of a simulation are sometimes 100% against all the existing thinking. Now, it is important for the engineer to abandon his routine and experience and follow the simulation. Do not make it difficult or look back too often. Of course, there are some basic rules that should be obeyed, such as maximum metal velocity, pouring times, and pouring temperatures. However, all of these parameters are used in the simulation process. Smart engineering can also answer the question, “How can we use the simulation technology?” In other words,

the goal is to achieve the maximum amount of castings on a pattern and to use the simulation process to fix the gating and risering problems. Effect on Production Costs. Using new technologies improves the production costs of malleable iron. Two major developments influence the cost reduction significantly: the reduction in heat treatment time and the improvement in casting yield. Table 2 gives a review of the production costs among ductile and malleable iron castings. The costs are presented as a percentage where the base is standard ductile iron at 100%. The casting yield is absorbed in the alloy costs. Alloy costs for ductile iron are more expensive than that of malleable iron. Alloy costs are the highest for austempered ductile iron (ADI), due to the current high costs of nickel. The comparison of the costs are not complete for producing a casting because the cost for coremaking, molding, and other operations are similar for both malleable and ductile iron. The cost for the heat treatment of ADI is very high.

The requirement that any iron produced for conversion to malleable iron must solidify white places definite section thickness limitations on the malleable iron industry. Thick metal sections can be produced by melting a base iron of low carbon and silicon contents or by alloying the molten iron with a carbide stabilizer. However, when carbon and silicon are maintained at low levels, difficulty is invariably encountered in annealing, and the time required to convert primary and pearlitic carbides to temper carbon becomes excessively long. High-production foundries are usually reluctant to produce castings more than approximately 40 mm (1.5 in.) thick. Some foundries, however, routinely produce castings as thick as 100 mm (4 in.). Automotive and associated applications of ferritic and pearlitic malleable irons include many essential parts in vehicle power trains, frames, suspensions, and wheels. A partial list includes differential carriers, differential cases, bearing caps, steering-gear housings, spring hangers, universal-joint yokes, automatic-transmission parts, rocker arms, disc brake calipers, and wheel hubs. Examples are shown in Fig. 8 for small parts and in Fig. 9 to 11 for typical automotive parts. Ferritic and pearlitic malleable irons are also used in the railroad industry and in agricultural equipment, chain links, ordnance material, electrical pole line hardware, hand tools, and other parts requiring section thicknesses and properties obtainable in these materials. Malleable iron castings are often selected because the material has excellent machinability in addition to significant ductility. In other applications, malleable iron is chosen because it combines castability with good toughness and machinability. Malleable iron is often chosen because of shock resistance alone. Applications are listed by ASTM specifications in Table 3. Mechanical properties are given in Table 4. Similar standards are grouped. The ASTM standards have specified metric values. A 47 (Ref 5) is A 47/A 47M with a metric grade 22010 (tensile minimum, 340 MPa; tensile yield, 220 MPa). These have not been included for clarity of the table.

Malleable Iron Castings / 889

Fig. 8

Small parts of malleable iron. (a) Nut made of pearlitic malleable iron, grade EN-GJMB-650-2, 600 g. The insert shows the core used for the as-cast internal screw thread. (b) Balancing weight made of ferritic malleable iron, grade EN-GJMB-350-10, 1500 g. The large picture shows the application, and the insert is the actual casting. Courtesy of VS GUSS AG, Germany

Fig. 11 Fig. 9

Various gears in martensitic malleable iron. Courtesy of Gieterij Doesburg, Netherlands

Fig. 10

Wide range of pistons made in pearlitic malleable iron. Courtesy of Gieterij Doesburg, Netherlands

Ferritic Malleable Iron Because ferritic malleable iron consists of only ferrite and temper carbon, the properties of ferritic malleable castings depend on the quantity, size, shape, and distribution of the

Various transmission parts made in oilquenched pearlitic malleable iron. Most of the parts are induction hardened after machining. Courtesy of Gieterij Doesburg, Netherlands

temper carbon particles and on the composition of the ferrite. Fully annealed ferritic malleable iron castings contain 2.00 to 2.70% graphitic carbon by weight, which is equivalent to approximately 6 to 8% by volume. Because the graphitic carbon contributes nothing to the strength of the castings, those with the lesser amount of graphite are somewhat stronger and more ductile than those containing the greater amount (assuming equal size and distribution of graphite particles). Elements such as silicon and manganese in solid solution in the ferritic matrix contribute to the strength and reduce the elongation of the ferrite. Therefore, by varying base metal composition, slightly different strength levels can be obtained in the fully annealed ferritic product. The mechanical properties that are most important for design purposes are tensile strength, yield strength, modulus of elasticity, fatigue strength, and impact strength. Hardness can be considered no more than an approximate indicator that the ferritizing anneal was complete, and it is seldom used for any other

purpose. The hardness of ferritic malleable iron almost always ranges from 110 to 156 HB. The tensile properties at room and lower temperatures of ferritic malleable iron (Table 5) are usually measured on unmachined test bars. Machined test bars are normally used for pearlitic malleable iron. Fracture toughness is listed as well. The fatigue limit of unnotched ferritic malleable iron is approximately 50% of the tensile strength, or from 170 to 205 MPa (25 to 30 ksi). Fig. 12 summarizes the effects of notches on fatigue strength. Notch radius generally has little effect on fatigue strength, but fatigue strength decreases with increasing notch depth. Modulus of elasticity in tension is approximately 170 GPa (25  106 psi). The modulus in compression ranges from 150 to 170 GPa (22  106 to 25  106 psi); in torsion, from 65 to 75 GPa (9.5  106 to 11  106 psi). Because brittle fractures are most likely to occur at high strain rates, at low temperatures, and with a high restraint on metal deformation, notch tests such as the Charpy V-notch test are conducted over a range of test temperatures to establish the toughness behavior and the temperature range of transition from a ductile to a brittle fracture. Figure 13 illustrates the behavior of ferritic malleable iron and several types of pearlitic malleable iron in the Charpy V-notch test. This shows that ferritic malleable iron has a higher upper-shelf energy and a lower transition temperature to a brittle fracture than pearlitic malleable iron does. Additional information on fracture toughness of malleable irons can be found later in the discussion of “Pearlitic and Martensitic Malleable Iron” in this article. Short-term tensile properties show no significant change to 370  C (700  F). Sustained-load stress-rupture data from 425 to 650  C (800 to 1200  F) are given in Fig. 14. The corrosion resistance of ferritic malleable iron is increased by the addition of copper, usually approximately 1%, in certain applications, for example, conveyor buckets, bridge castings, pipe fittings, railroad switch stands, and freight-car hardware. One important use for copper-bearing ferritic malleable iron is chain links. Ferritic malleable iron can be galvanized to provide added protection. The corrosion behavior of cast irons and use of protective coatings is found in “Corrosion of Cast Irons” in Corrosion: Materials, Volume 13B of ASM Handbook (Ref 8). Welding and Brazing. Welding of ferritic malleable iron almost always produces brittle white iron in the weld zone and the portion of the heat-affected zone (HAZ) immediately adjacent to the weld zone. During welding, temper carbon is dissolved, and upon cooling it is reprecipitated as carbide rather than graphite. In some cases, welding with a cast iron electrode may produce a brittle gray iron weld zone. The loss of ductility due to welding may not be serious in some applications. However, welding is usually not recommended unless the castings are subsequently annealed to

890 / Cast Irons convert the carbide to temper carbon and ferrite. Ferritic malleable iron can be fusion welded to steel without subsequent annealing if a completely decarburized zone as deep as the

normal HAZ is produced at the faying surface of the malleable iron part before welding. Silver brazing and tin-lead soldering can be satisfactorily used.

Table 3 Applications of malleable iron castings Mechanical properties are given in Table 4. Specification No.

Ferritic ASTM A 47

Class or grade

Microstructure

32510 35018

Temper carbon and ferrite

ASTM A 338 32510 35018 ASTM A 197 ...

Temper carbon and ferrite

Typical applications

General engineering service at normal and elevated temperatures for good machinability and excellent shock resistance Flanges, pipe fittings, and valve parts for railroad, marine, and other heavy-duty service to 345  C (650  F) Pipe fittings and valve parts for pressure service

Free of primary graphite

Pearlitic and martensitic ASTM A 220 40010 Temper carbon in necessary matrix 45008 without primary cementite or 45006 graphite 50005 60004 70003 80002 90001

General engineering service at normal and elevated temperatures. Dimensional tolerance range for castings is stipulated

Automotive ASTM A 602 M3210 Ferritic

For low-stress parts requiring good machinability: steeringgear housings, carriers, and mounting brackets Compressor crankshafts and hubs For selective hardening: planet carriers, transmission gears, and differential cases For machinability and improved response to induction hardening For high-strength parts: connecting rods and universal-joint yokes For high strength plus good wear resistance: certain gears

M4504 Ferrite and tempered pearlite(a) M5003 Ferrite and tempered pearlite(a) M5503 Tempered martensite M7002 Tempered martensite M8501 Tempered martensite (a) May be all tempered martensite for some applications

Pearlitic and Martensitic Malleable Iron Pearlitic malleable iron is produced either by controlled heat treatment of the same base white iron used to produce ferritic malleable iron or by alloying to prevent the decomposition of carbides dissolved in austenite during cooling from the first- stage annealing temperature. It can be produced by air cooling after first-stage annealing and subsequent tempering to develop specified properties. Martensitic malleable iron is produced by liquid quenching and tempering to develop specified properties (see the section “Control of Annealing” in this article). Variations in heat treatment, coupled with variations in base composition and melting practice, make it possible to obtain a wide range of properties in pearlitic or martensitic malleable iron. The mechanical properties of pearlitic and martensitic malleable iron vary in a substantially linear relationship with Brinell hardness (Fig. 15). In the low hardness ranges, below approximately 207 HB, the properties of airquenched and tempered pearlitic malleable are essentially the same as those of oil-quenched and tempered martensitic malleable. This results from the fact that attaining the low hardnesses requires considerable coarsening of the matrix carbides and partial second-stage graphitization. Either an air-quenched pearlitic structure or an oil-quenched martensitic structure can be coarsened and decarburized to meet this hardness requirement.

Table 4 Properties of malleable iron castings Microstructure and more applications are given in Table 3. Malleable casting specifications

ASTM A 47 SAE J158 GM11M ISO 5922 EN 1562:1997 GMW1 ASTM A 220 SAE J158 GM ISO 5922 EN 1562:1997 GMW1 ASTM A 220 SAE J158 GM85M ISO 5922 EN 1562:1997 GMW1 ASTM A 220 SAE J158 GM84M ISO 5922 EN 1562:1997 GMW1 ASTM A 220 SAE J158 GM88M ISO 5922 EN 1562:1997 GMW1 NA, nonapplicable

Tensile strength Class

32510 M3210 11M JMB/350-10 GJMB-350-10 MB 350-10 50005 M4504 86M JMB/500-5 GJMB-500-5 MB 550-4 60004 M5503 85M JMB/600-3 GJMB-600-3 MB 600-3 70003 M7002 84M JMB/700-2 GJMB-700-2 MB 700-2 80002 M8501 88M JMB/800-1 GJMB-800–1 NA

ksi

50 50 50 ... ... ... 70 65 69.6 ... ... ... 80 75 79.75 ... ... ... 85 90 100 ... ... ... 95 105 104.4 ... ... NA

MPa

... ... ... 350 350 350 ... ... ... 500 500 550 ... ... ... 600 600 600 ... ... ... 700 700 700 ... ... ... 800 800 NA

Yield strength ksi

MPa

Elongation, %

Hardness, Brinell

32.5 32 32.625 ... ... ... 50 45 47.85 ... ... ... 60 55 59.45 ... ... ... 70 70 79.75 ... ... ... 80 85 84.825 ... ... NA

... ... ... 200 200 200 ... ... ... 300 300 340 ... ... ... 390 390 410 ... ... ... 530 530 530 ... ... ... 600 600 NA

10 10 10 10 10 10 5 4 5 5 5 4 4 3 3 3 3 3 3 2 2 2 2 2 2 1 2 1 1 NA

156 max 156 max 156 max 150 max 150 max 149 max 179–228 163–217 163–207 165–215 165–215 179–229 197–241 187–241 197–241 195–245 195–245 195–245 217–269 229–269 241–269 240–260 240–290 241–293 241–285 269–302 269–302 270–320 270–320 NA

Matrix structure

Ferrite

Lamellar spheroidized pearlite and ferrite

Typical applications

General engineering service at normal and elevated temperatures for good machinability and excellent shock resistance. Flanges, pipe fittings, and valve parts for railroad, marine, and other heavy-duty service to 345  C (650  F) Compressor crankshafts and hubs

Lamellar spheroidized pearlite

For good machinability and improved response to induction hardening. Pistons, planet carriers, transmission gears

Tempered martensite

High-strength parts: connecting rods, universal-joint yokes

Tempered martensite

High strength plus good wear resistance: certain gears

Malleable Iron Castings / 891

Fig. 12

Effects of notch radius and notch depth on the fatigue strength of ferritic malleable iron

Fig. 13

Charpy V-notch transition curves for ferritic and pearlitic malleable irons. Source: Ref 7

At higher hardnesses, oil-quenched and tempered malleable iron has higher yield strength and elongation than air-quenched and tempered malleable because of greater uniformity of matrix structure and finer distribution of carbide particles. Oil-quenched and tempered pearlitic malleable iron is produced commercially to hardnesses as high as 321 HB, while the maximum hardness for high-production air-quenched and tempered pearlitic malleable iron is approximately 255 HB. The lower maximum hardness is applied to the air-quenched material for several reasons:
Because hardness on air quenching normally

does not exceed 321 HB and may be as low as 269 HB, attempts to temper to a hardness range above 255 HB produce nonuniform hardness and make the process control limits excessive.
Very little structural alteration occurs during the tempering heat treatment to a higher hardness, and the resulting structure is more

difficult to machine than an oil- quenched and tempered structure at the same hardness.
There is only a slight improvement in other mechanical properties with increased hardness above 255 HB. Because of these considerations, applications for air-quenched and tempered pearlitic malleable iron are usually those requiring moderate strength levels, while the higher-strength applications need the oil-quenched and tempered material. The modulus of elasticity in tension of pearlitic malleable iron is 175 to 195 GPa (25.5  106 to 28.0  106 psi). For automobile crankshafts, the modulus is important and must be determined with greater precision. The results of Charpy V-notch tests on pearlitic malleable iron are presented in Fig. 13. The fracture toughness of ferritic and pearlitic malleable irons has not been widely studied, but one researcher has estimated KIc values for these materials by using a J-integral approach

(Ref 6). Table 5 summarizes the fracture toughness values obtained for the various grades of malleable iron at various temperatures. All of the materials exhibited stable crack extension prior to fracture for 25 mm (1 in.) wide compact-tension specimens. High Yield Strength to Tensile Strength Ratio. The oil- quenched pearlitic and martensitic malleable irons, especially, have an excellent yield strength to tensile strength ratio of up to 0.8. Compared to standard ductile iron, malleable iron is superior with a much higher yield strength. At a tensile strength of 700 MPa (102 ksi), the yield strength of ductile iron is a minimum of 430 MPa (62 ksi), while malleable iron shows a yield strength of 575 MPa (83 ksi), almost 35% higher. Figure 16 shows the relationship between tensile strength and yield strength according to the ASTM standard A 220 (Ref 9) for malleable iron and the EN 1563:1997 for ductile iron. As for ductile irons, fracture toughness testing indicates that malleable irons possess considerably more toughness than is indicated by Charpy impact toughness results. Although the fracture toughness values for pearlitic grades are similar to those obtained for ferritic grades, the higher yield strengths of the pearlitic grades indicate that their critical flaw sizes, which are proportional to (KIc/sy)2, are less than those of the ferritic grades of malleable iron. Detailed information on the principles of fracture toughness and the nomenclature associated with fracture mechanics studies is available in the article “Fracture Toughness and Fracture Mechanics” in Mechanical Testing and Evaluation, Volume 8 of ASM Handbook (Ref 10). The compressive strength of pearlitic and martensitic malleable irons is plotted in Fig. 17. Torsional strength values are presented in Table 6. Sustained-load stress-rupture data for eight different grades of pearlitic malleable iron are shown in Fig. 18. Results of high-temperature Charpy V-notch tests showing the effect of hardness on impact energy are given in Fig. 19. The fatigue limits of pearlitic and martensitic malleable irons are approximately 40 to 50% of tensile strength. Oil-quenched and tempered martensitic iron usually has a higher fatigue ratio than pearlitic iron made by the arrestedanneal method. Wear Resistance. Because of its structure and hardness, pearlitic and martensitic malleable irons have excellent wear resistance. In some moving parts where bushings are normally inserted at pivot points, heat treated malleable iron has proved to be so wear resistant that the bushings have been eliminated. One example of this is the rocker arm for an overhead-valve automotive engine. Welding and Brazing. Welding of pearlitic or martensitic malleable iron is difficult because the high temperatures used can cause the formation of a brittle layer of graphite-free white iron. Pearlitic and martensitic malleable

892 / Cast Irons and clutch hubs. An example of a flame-hardened pearlitic malleable iron part is a pinion spacer used to support the cup of a roller bearing. To preclude service failures, the ends of the pinion spacer are flame hardened to file hardness to a depth of approximately 2.3 mm (0.09). Malleable iron can be carburized, carbonitrided, or nitrided to produce a surface with improved wear resistance. In addition, heat treatments such as austempering have been used in specialized applications.

REFERENCES

Composition, % Group

Grade

C

Si

Mn

P

S

Cr

A-1 B-1 E-1 G-1

35018 32510 35018 35018

2.21 2.50 2.16 2.29

1.14 1.32 1.17 1.01

0.35 0.43 0.38 0.38

0.161 0.024 0.137 0.11

0.081 0.159 0.095 0.086

... 0.029 0.017 ...

Fig. 14

Stress-rupture plot for various grades of ferritic malleable iron. The solid lines are curves determined by the method of least squares from the existing data and are least squares fit to data. The dashed lines define the 90% symmetrical tolerance interval. The lower dashed curve defines time and load for 95% survivors, and the upper dashed curve is the boundary for 5% survivors. Normal distribution is assumed.

iron can be successfully welded if the surface to be welded has been heavily decarburized. Pearlitic or martensitic malleable iron can be brazed by various commercial processes. One application is the induction silver brazing of a pearlitic malleable casting and a steel shaft to form a planetary output shaft for an automotive transmission. In another automotive application, two steel shafts are induction copper brazed to a pearlitic malleable iron shifter shaft plate. Selective Surface Hardening. Pearlitic malleable iron can be surface hardened by either induction heating and quenching or flame heating and quenching to develop high hardness at the heat-affected surface. Considerable research has been done to determine the surface-hardening characteristics of pearlitic malleable and its capability of developing high hardness over relatively narrow surface bands. Generally, little

difficulty is encountered in obtaining hardnesses in the range of 55 to 60 HRC, with the depth of penetration being controlled by the rate of heating and by the temperature at the surface of the part being hardened (Fig. 20). The maximum hardness obtainable in the matrix of a properly hardened pearlitic malleable part is 67 HRC. However, conventional hardness measurements made on castings show less than 67 HRC because of the presence of the graphite particles, which are averaged into the hardness. Generally, a casting with a matrix microhardness of 67 HRC will have approximately 62 HRC average hardness, as measured with the standard Rockwell tester. Similarly, a Rockwell or Brinell hardness test on softer structures will show less than matrix microhardness because of the presence of graphite. Two examples of automobile production parts hardened by induction heating are rocker arms

1. Malleable Iron Castings, Malleable Founders’ Society, The Ann Arbor Press Inc., Ann Arbor, MI, 1960 2. C. Leon, U. Ekpoon, and R.W. Heine, Relationship of Casting Defects to Solidification of Malleable Base White Cast Iron, AFS Trans., Vol 81–74, p 323–344 3. F.W. Jacobs, Practical Application of Liquidus Control for Malleable Iron Melting, AFS Trans., Vol 81–56, p 261–276 4. C.J. van Ettinger and W. Baumgart, “Thermal Analysis, A Unique Fingerprint of a Melt,” Paper 78, 66th World Foundry Congress (Istanbul, Turkey), Sept 6–9, 2004, p 743–757 5. “Standard Specification for Malleable Iron Castings,” A 47/A 47M, Annual Book of ASTM Standards, ASTM International 6. W.L. Bradley, Fracture Toughness Studies of Gray, Malleable and Ductile Cast Iron, Trans. AFS, Vol 89, 1981, p 837–848 7. C.F. Walton and T.J. Opar, Ed., Iron Castings Handbook, Iron Castings Society, 1981, p 297–321 8. T.C. Spence, Corrosion of Cast Irons, Corrosion. Materials, Vol 13B, ASM Handbook, ASM International, 2005, p 43–50 9. “Standard Specification for Pearlitic Malleable Iron Castings,” A 220, Annual Book of ASTM Standards, ASTM International 10. Fracture Toughness and Fracture Mechanics, Mechanical Testing and Evaluation, Vol 8, ASM Handbook, ASM International, 2000, p 563–575

Malleable Iron Castings / 893

Fig. 15

Relationships of tensile properties to Brinell hardness for pearlitic malleable irons from two foundries. The mechanical properties of these irons vary substantially in a linear relationship with Brinell hardness, and in the low hardness ranges (below approximately 207 HB), the properties of air-quenched and tempered material are essentially the same as those produced by oil quenching and tempering.

Table 5

Fracture toughness of malleable irons Test temperature

Malleable iron grade

Ferritic M3210 Pearlitic M4504 (normalized)

M5503 (quenched and tempered)

M7002 (quenched and tempered)

Source: Ref 6

Yield strength

KIc

pffiffiffiffi ksi in.

F

MPa

ksi

pffiffiffiffi MPa m

24 19 59

75 3 74

230 240 250

33 35 36

44 41 44

40 38 40

24 19 57 24 19 58 24 19 58

75 2 70 75 3 73 75 3 72

360 380 390 410 430 440 520 550 570

52 55 57 60 64 66 75 80 83

55 48 30 45 51 30 54 39 39

50 44 27 41 47 27 49 35 36



C



Fig. 17

Fig. 16

Table 6

Compressive strength of martensitic malleable irons

pearlitic

and

Difference in yield strength and ultimate tensile strength ratio (slope of curve) of malleable iron (A 47/A 220/Armasteel GM) and ductile iron (Lineair ductile EN 1563)

Torsion test values of malleable irons Modulus of rupture

Yield strength

Modulus of elasticity

Material(a)

MPa

ksi

MPa

ksi

Ultimate twist, degree

GPa

106 psi

Ferritic 45010(b) 60003(b) 80002(b) High strength(c)

375 570 665 715 850

54 83 97 104 123

130 215 320 410 490

19 31 46 59 71

540 350 280 140 70

65.15 66.75 70.3 69.22 67.5

9.45 9.68 10.20 10.04 9.79

(a) Specimens 20 mm (0.750 in.) in diameter. (b) Pearlitic malleable iron. (c) High-strength malleable iron with yield strength of approximately 620 MPa (90 ksi), tensile strength of 830 MPa (120 ksi), and 2% elongation in 50 mm (2 in.)

Composition, % Material

C

Pearlitic (low carbon-high phosphorus) Group E-2 2.27 Group G-2 2.29 Pearlitic (high carbon-low phosphorus) Group C-2 2.65 Group W-1 2.45 Alloyed pearlitic (low carbon-high phosphorus) Group E-3 2.21 Group L-1 2.16 Group L-2 2.16 Group L-3 2.32

Fig. 18

Si

Mn

S

P

Cr

Others

1.15 1.01

0.89 0.75

0.098 0.086

0.135 0.11

0.019 ...

... ...

1.35 1.38

0.41 0.41

0.15 0.12

... 0.04

0.018 0.032

0.0020 B ...

1.13 1.18 1.18 1.14

0.88 0.72 0.80 0.82

0.110 0.120 0.123 0.117

0.122 0.128 0.128 0.128

0.021 ... ... ...

0.47Mo, 1.03Cu 0.34Mo, 0.83 Ni 0.40Mo, 0.62 Ni 0.38Mo, 0.65 Ni

Stress-rupture plot for (a) pearlitic malleable iron and (b) alloyed pearlitic malleable iron. The solid lines are curves determined by the method of least squares from the existing data. The dashed lines define the 90% symmetrical tolerance interval. The lower dashed curve defines time and load for 95% survivors, and the upper dashed curve is the boundary for 5% survivors. Normal distribution is assumed.

Malleable Iron Castings / 895

Fig. 19

Charpy V-notch impact energy of one heat of air-quenched and tempered pearlitic malleable iron

Fig. 20

Hardness versus depth for surface-hardened pearlitic malleable irons. Curves labeled “Matrix” show hardness of the matrix, converted from microhardness tests. O, oil-quenched and tempered to 207 HB before surface hardening; A, air-cooled and tempered to 207 HB before surface hardening

ASM Handbook, Volume 15: Casting ASM Handbook Committee, p 896-903 DOI: 10.1361/asmhba0005327

Copyright © 2008 ASM International® All rights reserved. www.asminternational.org

White Iron and High-Alloyed Iron Castings Richard B. Gundlach, Stork Climax Research Services

HIGH-ALLOYED WHITE CAST IRONS are an important group of materials whose production must be considered separately from that of ordinary types of cast irons. In these cast iron alloys, the alloy content is well above 4%, and consequently they cannot be produced by ladle additions to irons of otherwise standard compositions. They are usually produced in foundries specially equipped to produce highly alloyed irons. These iron alloys are most often melted in electric furnaces, specifically electric arc furnaces and induction furnaces, in which the precise control of composition and temperature can be achieved. The foundries usually have the equipment needed to handle the heat treatment and other thermal processing unique to the production of these alloys. The high-alloyed white irons are primarily used for abrasion-resistant applications and are readily cast into the parts needed in machinery for crushing, grinding, and handling of abrasive materials. The chromium content of high-alloyed white irons also enhances their corrosion-resistant properties. The large volume fraction of primary and/or eutectic carbides in their microstructures provides the high hardness needed for crushing and grinding other materials. The metallic matrix supporting the carbide phase in these irons can be adjusted by alloy content and heat treatment to develop the proper balance between the resistance to abrasion and the toughness needed to withstand repeated impact. All high-alloyed white irons contain chromium to prevent the formation of graphite upon solidification and to ensure the stability of the carbide phase. Most also contain nickel, molybdenum, copper, or combinations of these alloying elements to prevent the formation of pearlite in the microstructure. While lowalloyed white iron castings, which have an alloy content below 4%, develop hardnesses in the range of 350 to 550 HB, the high-alloyed irons range in hardness from 450 to 800 HB. In addition, several grades contain alloy eutectic carbides (M7C3 chromium carbides), which are substantially harder than the M3C iron carbides


Nickel-chromium white irons, which are low-

in low-alloyed irons. For many applications, the increased abrasion resistance of the more expensive high-alloyed white irons adds significantly to wear life, enabling them to provide the most cost-effective performance. Alloy Grades. Specification ASTM A 532/A 532M, covers the composition and hardness of the abrasion-resistant white iron grades (Table 1). Many castings are ordered according to these specifications. However, a large number of castings are produced with composition modifications for specific applications. It is most desirable that the designer, metallurgist, and foundryman work together to specify the composition, heat treatment, and foundry practice to develop the most suitable alloy and casting design for a specific application. The high-alloyed white cast irons fall into two major groups:

chromium alloys containing 3 to 5% Ni and 1 to 4% Cr, with one alloy modification that contains 7 to 11% Cr. The nickel-chromium irons are also commonly identified as Ni-Hard types 1 to 4.
The chromium-molybdenum irons containing 11 to 23% Cr, with up to 3% Mo and often additionally alloyed with nickel or copper A third group comprises the 25 or 28% Cr white irons, which may contain other alloying additions of molybdenum and/or nickel up to 1.5%.

Nickel-Chromium White Irons The oldest group of high-alloyed irons of industrial importance, the nickel-chromium white irons, or Ni-Hard irons, have been

Table 1 Composition and mechanical requirements of abrasion-resistant cast irons in accordance with ASTM A 532/A 532M Composition(a),% Class

Type

I I I I II II II III

A B C D A B D A

Designation

Ni-Cr-HC Ni-Cr-LC Ni-Cr-GB Ni-Hi Cr 12% Cr 15% Cr-Mo 20% Cr-Mo 25% Cr

C

Si

Ni

Cr

2.8–3.6 2.4–3.0 2.5–3.7 2.5–3.6 2.0–3.3 2.0–3.3 2.0–3.3 2.0–3.3

0.8 max 0.8 max 0.8 max 2.0 max 1.5 max 1.5 max 1.0–2.2 1.5 max

3.3–5.0 3.3–5.0 4.0 max 4.5–7.0 2.5 max 2.5 max 2.5 max 2.5 max

1.4–4.0 1.4–4.0 1.0–2.5 7.0–11.0 11.0–14.0 14.0–18.0 18.0–23.0 23.0–30.0

Mo

1.0 1.0 1.0 1.5 3.0 3.0 3.0 3.0

max(b) max(b) max(b) max(c) max(d) max(d) max(d) max(d)

Mechanical requirements(e) Hardness, HB Class

I I I I II II II III

Type

A B C D A B D A

Designation

Ni-Cr-HC Ni-Cr-LC Ni-Cr-GB Ni-Hi Cr 12% Cr 15% Cr-Mo 20% Cr-Mo 25% Cr

Typical section thickness, max

Sand cast, min

Chill cast, min

Hardened, min

Softened, max

in

mm

550 550 550 550 550 450 450 450

600 600 600 550 550 ... ... ...

600 600 600 600 600 600 600 600

... ... 400 ... 400 400 400 400

8 8 3 diam ball 12 1 diam ball 4 8 8

200 200 75 diam ball 300 25 diam ball 100 200 200

(a) Maximum: 2.0% Mn (b) Maximum: 0.30% P, 0.15% S. (c) Maximum: 0.10% P, 0.15% S. (d) Maximum: 0.10% P, 0.06% S, 1.2% Cu. (e) Other hardness levels and methods given; see standard. Typical section thickness is not part of standard.

White Iron and High-Alloyed Iron Castings / 897 produced for more than 70 years and are very cost- effective materials for crushing and grinding. In these martensitic white irons, nickel is the primary alloying element because, at levels of 3 to 5%, it is effective in suppressing the transformation of the austenite matrix to pearlite, thus ensuring that a hard, martensitic structure (usually containing significant amounts of retained austenite) will develop upon cooling in the mold. Chromium is included in these alloys, at levels from 1.4 to 4%, to ensure that the irons solidify carbidic, that is, to counteract the graphitizing effect of nickel. A typical microstructure is shown in Fig. 1. The optimum composition of a nickel-chromium white iron alloy depends on the properties required for the service conditions and the dimensions and weight of the casting. Abrasion resistance is generally a function of the bulk hardness and the volume of carbide in the microstructure. When abrasion resistance is the principal requirement and resistance to impact loading is secondary, alloys having high carbon contents, ASTM A 532/A 532M class I type A (Ni-Hard 1), are recommended. When conditions of repeated impact are anticipated, the lower-carbon alloys, class I type B (Ni-Hard 2), are recommended because they have less carbide and therefore greater toughness. A special grade, class I type C, has been developed for producing grinding balls and slugs. Here, the nickel-chromium alloy composition has been adapted for chill casting and specialized sand casting processes. The class I type D (Ni-Hard 4) alloy is a modified nickel-chromium iron that contains higher levels of chromium, ranging from 7 to 11%, and increased levels of nickel, ranging from 5 to 7%. Whereas the eutectic carbide phase in the lower-alloyed nickel-chromium irons is M3C (iron carbide), which forms as a continuous network in these irons, the higher chromium in the type D alloy promotes M7C3 chromium carbides, which form a relatively discontinuous eutectic carbide distribution (Fig. 2). This modification in the eutectic carbide pattern provides an appreciable improvement in resistance to fracture by impact. The higher alloy content of this iron grade also results in improved corrosion resistance, which has proved to be useful in the handling of corrosive slurries.

Melting Practice A principal advantage of the nickel-chromium irons is that they can be melted in a cupola; however, better control of composition and temperature are achieved with electric furnace melting. Consequently, electric arc furnace melting and induction furnace melting are most common. The nickel-chromium irons are readily made in either acid-, neutral-, or basic-lined electric furnaces. They are normally dead melted, and there is usually no reason to use oxygen lancing except as a means of

slightly reducing carbon content. Acid linings are generally more economical than basic linings. The type of lining affects silicon and chromium losses. Very little adjustment to slag composition is necessary in acid melting, which is used for the majority of current production. Normal charge materials are various kinds of steel scrap, foundry returns, or returns of similar alloy from service. For foundries that lack the facilities for rapid melt analysis, it may be necessary to select steel scrap rather carefully to ensure that the residual levels of alloying elements, such as manganese and copper, which have a potent effect on austenite retention, are consistent and under control. Chromium is obtained in the form of highcarbon ferrochrome and is generally added near the end of the heat to avoid excessive oxidation losses. Carbon is obtained from electrode graphite, petroleum coke, and other sources. Pig iron is also used to carburize the melt and can be an additional source of silicon. Silicon and manganese are added as ferroalloys. Molybdenum is usually added as ferromolybdenum; however, with arc furnaces, molybdenum oxide briquettes may be used. Sulfur is limited to 0.06%, and phosphorus is kept to 0.10%, as specified by ASTM A 532/A 532M. High superheating temperatures have not been necessary when melting in induction furnaces, which have good stirring action, and a final temperature of 1480  C (2700  F) is usually adequate for thick-section castings. Final temperatures of up to 1565  C (2850  F) are commonly used in arc furnace practice to ensure homogeneity of the bath composition and to accelerate the solution of carbon, added after meltdown, and late alloy additions. No particular problems have been associated with high superheat temperatures, but it is more difficult to predict melting losses. As in steel melting, deoxidation of the bath with aluminum was common, but it is no longer practiced by most producers, with no adverse effects on the soundness or mechanical properties of the casting. A late addition/inoculation of foundry-grade ferro-silicon is often made to improve toughness.

Composition Control Carbon is varied according to the properties needed for the intended service. Carbon contents in the range of 3.2 to 3.6% are prescribed when maximum abrasion resistance is desired. When impact loading is expected, carbon content should be held in the range of 2.7 to 3.2%. Nickel content increases with section size or cooling time of the casting to inhibit pearlitic transformation. For castings of 38 to 50 mm (1.5 to 2 in.) thick, 3.4 to 4.2% Ni is sufficient to suppress pearlite formation upon mold cooling. Heavier sections may require nickel levels up to 5.5% to avoid the formation of pearlite. It is important to limit nickel content to the level needed for control of pearlite; excess

Fig. 1

Typical microstructure of class I type A nickel-chromium white cast iron. Original magnification: 340

Fig. 2

Typical microstructure of class I type D nickel-chromium white cast iron. Original magnification: 340

nickel increases the amount of retained austenite and lowers hardness. Silicon is needed for two reasons. A minimum amount of silicon is necessary to improve fluidity of the melt and to produce a fluid slag, but of equal importance is its effect on as-cast hardness. Increased levels of silicon, in the range of 1 to 1.5%, have been found to increase the amount of martensite and the resulting hardness. Late additions of ferrosilicon (0.2% as 75% Si-grade ferrosilicon) have been reported to increase toughness. It is important to note that higher silicon contents can promote pearlite and may increase the nickel requirement. Chromium, primarily added to offset the graphitizing effects of nickel and silicon in

898 / Cast Irons types A, B, and C alloys, ranges from 1.4 to 4.0%. Chromium content must increase with increasing section size. In type D alloy, chromium levels range from 7 to 11% (typically 9%) for the purpose of producing eutectic carbides of the M7C3 chromium carbide type, which are harder and less deleterious to toughness. Manganese is typically held to a maximum of 0.8% even though 2.0% maximum is allowed according to ASTM A 532/A 532M specification. While it provides increased hardenability to avoid pearlite formation, it is a more potent austenite stabilizer than nickel and promotes increased amounts of retained austenite and lower as-cast hardness. For this reason, higher manganese levels are undesirable. When considering the nickel content required to avoid pearlite in a given casting, the level of manganese present should be a factor. Copper increases both hardenability and the retention of austenite and therefore must be controlled for the same reason that manganese must be limited. Copper should be treated as a nickel substitute and, when properly included in the calculation of the amount of nickel required to inhibit pearlite, it reduces the nickel requirement. Molybdenum is a potent hardenability agent in these alloys and is used in heavy-section castings to augment hardenability and inhibit pearlite.

Pouring Practices High pouring temperatures aggravate shrinkage under feeder heads and other hot spots and can lead to microshrinkage and coarse dendritic structures. Careful control of pouring temperature is particularly important in the consistent production of thick-section castings. Low pouring temperatures are necessary not only to avoid shrinkage defects but also to avoid problems such as metal penetration and burnedon sand. Low pouring temperatures are also effective in controlling dendrite size and the coarseness of the eutectic carbide structure. The eutectic temperature for the various nickelchromium iron grades is approximately 1200  C (2190  F), and solidification generally begins at temperatures of 1200 to 1280  C (2190 to 2340  F), depending on composition. When selecting a pouring temperature, common rules apply. Pouring temperatures are seldom lower than 100  C (180  F) above the liquidus temperature. Higher pouring temperatures may be necessary when pouring smaller castings. Of course, casting configurations must be considered when selecting optimum pouring temperatures.

Molds, Patterns, and Casting Design These alloys may be sand cast or chill cast in permanent molds. The chill cast microstructures develop higher hardness, strength, and impact toughness over the sand cast structures, because of the finer carbides. It is recommended that,

whenever practical, the wearing faces of the casting be cast against a chill to improve abrasion resistance and toughness. Sand molds may be made either of green sand or of dry sand, oil sand, or steel casting sand. Air-setting and thermal-setting resin sands are also commonly used. Molds must be rigid to minimize shrinkage defects. White irons are subject to hot tearing, unless suitable precautions are taken. Occasionally when castings are extracted from the mold, they are found to be cracked. Stresses large enough to cause fracture can arise when either the mold or the cores are excessively strong and there is inadequate mold or core breakdown to allow normal thermal contraction with cooling. Large cores, such as those used in pump volutes, and even small cores used to form bolt holes, can cause cracking of this type. Shrinkage allowance is 13 to 21 mm/m (5/32 to ¼ in./ft); a commonly used figure is 16 mm/m (3/16 in./ft). The actual pattern allowance depends on the choice of alloy and the geometry of the casting. Patterns should employ generous radii and avoid sharp section changes to avoid initiating cracking upon solidification and subsequent cooling. These alloys exhibit relatively high liquid-tosolid shrinkage (5%); therefore, large gates and risers are required for feeding. Special consideration should be given to the design of end gates and the positioning of risers so that they can be readily removed. Risers cannot be removed by burning and are often notched or necked down to facilitate removal by impact. They are most easily removed after the castings have cooled to room temperature, that is, when martensite is present. Martensitic transformation begins at temperatures below 230  C (450  F).

properly made, they have a martensitic matrix structure, as-cast. Tempering is performed between 205 and 260  C (400 and 500  F) for at least 4 h. This tempers the martensite, relieves some of the transformation stresses, and increases the strength and impact toughness by 50 to 80%. Some additional martensite may form upon cooling from the tempering temperature. This heat treatment does not reduce hardness or abrasion resistance. High-Temperature Heat Treatment. In the past, hardening of the class I type D, Ni-Hard 4, was performed by supercritical heat treatment when as-cast hardness was insufficient. An austenitizing heat treatment usually comprised heating the casting at temperatures between 750 and 790  C (1380 and 1450  F) with a soak time of 8 h. Air or furnace cooling, not over 30  C/h (50  F/h), was conducted, followed by a tempering/stress-relief heat treatment. Refrigeration treatment is the more commonly practiced remedy for low hardness today (2008). To achieve a hardness of 550 HB, it is necessary for the as-cast austenite-martensite microstructure to have at least 60% martensite present. When martensite content is increased to 80 to 90%, however, hardness values exceed 650 HB. To reduce the amount of retained austenite, that is, form more martensite, deep freeze treatments are commonly applied. Refrigeration to temperatures between 70 and 185  C (90 and 300  F) for ½ to 1 h will usually raise the hardness level 100 HB units. A tempering/ stress-relief heat treatment usually follows. The typical refrigerated nickel-chromium iron microstructure is shown in Fig. 3. In the heat treatment of any white cast iron, care must be taken to avoid cracking by thermal shock; the castings must never be placed in a hot furnace or otherwise subjected to rapid

Shakeout Practices The shakeout practice is a critical step in the successful production of nickel-chromium iron castings. High residual stresses and cracking can result from extracting castings from a mold at too high a temperature. Cooling all the way to room temperature in the mold is desirable because, as stated previously, transformation to martensite occurs below 230  C (450  F) during the last stages of cooling. This precaution is mandatory in heavy-section castings; the volume increase that occurs with martensite formation results in the development of high transformation stresses in castings having steep temperature gradients due to rapid cooling. Thin-section castings and castings having simple shapes are less susceptible to stresses and are often removed from the mold once the castings have reached black heat. With these alloys, slow cooling favors higher as-cast hardness.

Heat Treatment All nickel-chromium white iron castings are given a stress-relief heat treatment because,

Fig. 3

Microstructure of class I type D nickelchromium white cast iron after refrigeration. Original magnification: 340

White Iron and High-Alloyed Iron Castings / 899 heating or cooling. The risk of cracking increases with the complexity of the casting shape and section thickness.

Machining Alloyed white irons are generally ground to finish size; care must be taken to avoid cracking by overheating. Single-point turning and boring can be done using heavy tooling, rigid setups, sufficient power, and carbide- or boride-tipped cutters or special ceramic inserts. Electrical discharge machining and electrochemical machining are also feasible.

Applications Because of their low cost, the martensitic nickel-chromium white irons are consumed in large tonnages in mining operations as ball mill liners and grinding balls. Class I type A castings are used in applications requiring maximum abrasion resistance, such as ash pipes, slurry pumps, roll heads, muller tires, augers, coke-crusher segments, classifier shoes, brick molds, pipe elbows carrying abrasive slurries, and grizzly disks. Type B is recommended for applications requiring more strength and exerting moderate impact, such as crusher plates, crusher concaves, and pulverizer pegs. Class I type D, Ni-Hard type 4, has a higher level of strength and toughness and is therefore used for the more severe applications that justify its added alloy costs. It is commonly used for pump volutes handling abrasive slurries and coal pulverizer table segments and tires. The class I type C alloy (Ni-Hard 3) is specifically designed for the production of grinding balls. This grade is both sand cast and chill cast. Chill casting has the advantage of lower alloy cost, and, more important, provides a 15 to 30% improvement in life. All grinding balls require tempering for 8 h at 260 to 315  C (500 to 600  F) to develop adequate impact toughness.

Special Nickel-Chromium White Iron Alloys Certain proprietary grades of type A alloy have been developed by the rolling mill industry. The compositions of these alloys have been modified to produce mottled structures, containing some graphite. The graphite inclusions are reported to improve resistance to thermal cracking. These indefinite chill rolls are cast in thickwall gray iron chiller molds in roll diameters of up to 1015 mm (40 in.) or more. The silicon-tochromium ratios and inoculation with ferro-silicon are carefully controlled to control the amount and distribution of the graphite particles. The rolls can be double poured with a gray iron core. With molybdenum modification, the matrix of the chill cast shell becomes martensitic. Some roll alloys are designed to be heat treated, that is, by a modified normalizing heat treatment, to obtain a bainitic microstructure.

High-Chromium White Irons The high-chromium white irons have excellent abrasion resistance and are used effectively in slurry pumps, brick molds, coal-grinding mills, shot-blasting equipment, and components for quarrying, hard-rock mining, and milling. In some applications, they must also be able to withstand heavy impact loading. These alloyed white irons are recognized as providing the best combination of toughness and abrasion resistance attainable among the white cast irons. In the high-chromium irons, as with most abrasion-resistant materials, there is a trade-off between wear resistance and toughness. By varying composition and heat treatment, these properties can be adjusted to meet the needs of most abrasive applications. As a class of alloyed irons, the high-chromium irons are distinguished by the hard, relatively discontinuous M7C3 eutectic carbides present in the microstructure, as opposed to the softer, more continuous M3C eutectic carbides present in the alloyed irons containing less chromium. These alloys are usually produced as hypoeutectic compositions.

improved by increasing the amount of carbide (increasing the carbon content), while toughness is improved by increasing the proportion of metallic matrix (reducing the carbon content). The influence of carbon content on the shape and distribution of the carbide phase in these alloys is illustrated in the micrographs of Fig. 4. Large hexagonal carbide rods occur when carbon contents exceed the eutectic carbon content (Fig. 4c). These primary chromium carbides, which precipitate from the melt before eutectic solidification, are quite deleterious to impact toughness and should be avoided in castings subjected to any impact in service. The eutectic carbon content changes inversely with the chromium content in these alloys. The relationship between eutectic carbon content and chromium content is shown in Fig. 5. To avoid primary M7C3 chromium carbides in high-chromium irons, the composition should follow the formula: %C þ %Cr=15  4:5

(Eq 1)

Classes of High-Chromium Irons Specification ASTM A 532/A 532M covers the compositions and hardnesses of two general classes of the high-chromium irons (Table 1). The chromium-molybdenum irons (class II of ASTM A 532/A 532M) contain 11 to 23% Cr and up to 3.0% Mo and can be supplied either as-cast with an austenitic or austenitic-martensitic matrix, or heat treated with a martensitic matrix microstructure for maximum abrasion resistance and toughness. They are usually considered the hardest of all grades of white cast irons. Compared to the lower-alloy nickel-chromium white irons, the eutectic carbides are harder and can be heat treated to achieve castings of higher hardness. Molybdenum, as well as nickel and copper when needed, is added to prevent pearlite and to ensure maximum hardness. The high-chromium irons (class III of ASTM A 532/A 532M) represent the oldest grade of high-chromium irons, with the earliest patents dating back to 1917. These general-purpose irons, also called 25% Cr and 28% Cr irons, contain 23 to 28% Cr with up to 3.0% Mo. To prevent pearlite and attain maximum hardness, molybdenum is added in all but the lightest-cast sections. Alloying with nickel and copper up to 1% is also practiced. Although the maximum attainable hardness is not as high as in the class II chromium-molybdenum white irons, these alloys are selected when resistance to corrosion is also desired.

Microstructures Carbide. The carbides in high-chromium irons are very hard and wear resistant but are also brittle. In general, wear resistance is

Fig. 4

Microstructures of high-chromium white iron compositions. (a) Low carbon (hypoeutectic). (b) Eutectic. (c) High-carbon (hypereutectic). Original magnification: all 75. Courtesy of Climax Molybdenum Company

900 / Cast Irons

Fig. 5

Relationship between the chromium and carbon contents and the eutectic composition in high-chromium irons

Fig. 6

High-chromium iron with an as-cast austenitic matrix microstructure. Original microstructure: 500. Courtesy of Climax Molybdenum Company

Table 2 Minimum alloy content to avoid pearlite in mold-cooled castings for indicated effective section size (plate thickness or radius of rounds) ASTM A 532/A 532M class

Plate thickness or radius of rounds Cr(a), %

IIB

14–18

IID

18–23

IIIA

23–28

C(a), %

2.0 3.5 2.0 3.2 2.0 3.0

25 mm (1 in.)

1.0 2.0 0.5 1.5 ... 1.0

Mo Mo Mo Mo Mo

50 mm (2 in.)

1.5 2.5 1.0 2.0 0.5 1.5

Mo Mo Mo Mo Mo Mo

100 mm (4 in.)

1.5 2.5 1.0 2.0 1.0 1.5

Mo Mo Mo Mo Mo Mo

+ + + +

1.0 1.0 1.0 1.0

(Ni (Ni (Ni (Ni

+ + + +

Cu) Cu) Cu) Cu)

+ 1.0 (Ni + Cu)

(a) In base irons containing 0.6% Si and 0.8% Mn

As-Cast Austenitic. Solidification in the hypoeutectic alloys occurs by the formation of austenite dendrites followed by the eutectic formation of austenite and M7C3 chromium carbides. Under equilibrium conditions, additional chromium carbide precipitates from the austenite matrix upon cooling from the eutectic to the critical temperature, that is, approximately 760  C (1400  F), and transformation to ferrite and carbide occurs with subsequent cooling. However, when cooling under nonequilibrium conditions such as those encountered in most commercial castings, the austenite becomes supersaturated in carbon and chromium. Because of elevated carbon and chromium contents, a metastable austenitic matrix microstructure normally develops, provided pearlitic and bainitic transformations have been inhibited (Fig. 6). With sufficient alloying with molybdenum, manganese, nickel, and copper, pearlitic transformation can be avoided in virtually any cast section. As-Cast Martensitic Martensitic structures can be obtained as-cast in heavy-section castings that cool slowly in the mold. With slow cooling rates, austenite stabilization is incomplete, and partial transformation to martensite occurs. However, in these castings, martensite is mixed

with large amounts of retained austenite, and therefore, hardness levels are lower than can be achieved in heat treated martensitic castings. These castings must contain sufficient alloy to suppress pearlite upon cooling. Some compositions (higher silicon) have been developed to assist martensite formation in refrigeration treatments. Subcritical heat treatment has also been used to reduce austenite content and, at the same time, increase hardness and toughness. Heat Treated Martensitic. To obtain maximum hardness and abrasion resistance, martensitic matrix structures must be produced by full heat treatment (Fig. 7). The casting must contain sufficient alloy to avoid pearlite formation upon cooling from the heat treatment temperature.

Selecting Compositions to Obtain Desired Structures Many complex sections, such as slurry pump components, are often used in the as-cast austenitic/martensitic condition to avoid the possibility of cracking and distortion when heat treated. To prevent pearlite in mold cooling,

Fig. 7

High-chromium iron with a heat-treatedmartensitic matrix microstructure. Original mag-nification: 500. Courtesy of Climax Molybdenum Company

alloying additions are usually required. As the carbon content is increased, more chromium is consumed, forming additional carbide, and therefore, larger alloying additions are required. Table 2 presents a guide for appropriate alloying to prevent pearlite in the various classes of as-cast irons. Optimum performance is usually achieved with heat treated martensitic structures. Again, alloying must be sufficient to ensure a pearlite-free microstructure, with heat treatment. Of necessity, the heat treatment, discussed later in detail (see the section “Quenching” in this article), requires an air quench from the austenitizing temperature. Faster cooling rates should not be used, because the casting can develop cracks due to high thermal and/or transformation stresses. Thus, the alloy must have sufficient hardenability to allow air hardening. However, overalloying with manganese, nickel, and copper promotes retained austenite, which detracts from resistance to abrasion and spalling. Therefore, it is best to obtain adequate hardenability primarily with molybdenum. Table 3 is offered as a guide to alloying for air quenching heat treated castings of various sections.

Melting Practice The high level of carbon pickup with the high chromium contents in these alloys precludes the use of cupola melting. Consequently, electric arc and induction furnaces are normally used. The high-chromium irons are readily made in acid-, neutral-, or basic-lined electric furnaces. They are normally dead melted, and there is usually no reason to use oxygen lancing except as a means of slightly reducing carbon content. Despite rapid refractory wear due to chromium-bearing slag, acid linings are generally more economical than basic linings. Very

White Iron and High-Alloyed Iron Castings / 901 Table 3 Minimum alloy content to avoid pearlite in heat treatment for indicated effective section size (plate thickness or radius of rounds) ASTM A 532/A 532M class

Plate thickness or radius of rounds Cr(a), %

IIB

14–18

IID

18–23

IIIA

23–28

C(a), %

2.0 3.5 2.0 3.2 2.0 3.0

50 mm (2 in.)

1.5 3.0 1.0 1.5 0.5 1.5

Mo Mo Mo Mo Mo Mo

125 mm (5 in.)

1.5 2.0 2.0 2.0 1.5 1.5

Mo Mo Mo Mo Mo Mo

+ 0.5 (Ni + Cu) + 1.0 (Ni + Cu) + 0.7 (Ni + Cu) + 0.6 (Ni + Cu)

150–255 mm (6–10 in.)

2.0 2.5 2.0 2.0 1.5 1.5

Mo Mo Mo Mo Mo Mo

+ + + + + +

1.0 1.2 0.5 1.2 0.5 1.2

(Ni (Ni (Ni (Ni (Ni (Ni

+ + + + + +

Cu) Cu)(b) Cu) Cu)(b) Cu) Cu)(b)

(a) In base irons containing 0.6% Si and 0.8% Mn. (b) Nickel and copper promote retained austenite and should be restricted to combined levels of 1.2% maximum; manganese behaves similarly and should be restricted to 1.0% maximum.

little adjustment to slag composition is necessary in acid melting, which is used for the majority of current production. Viscous or nonfluid slags should be avoided. Normal charge materials are various kinds of steel scrap, foundry returns, or returns of similar alloy from service. For foundries that lack the facilities for rapid melt analysis, it may be necessary to select steel scrap rather carefully to ensure that the total content of alloying elements that have a potent effect on hardenability and austenite retention is consistent and under control. While some furnace operators prefer to add a portion of high-carbon ferrochrome to the charge, it is generally added near the end of the heat to avoid excessive oxidation losses. Carbon is obtained from electrode graphite, petroleum coke, and other sources. If pig iron is used to carburize the melt, it should have a low silicon content. The ideal silicon content is 0.6%. Less than 0.4% Si in the bath can cause difficulties with viscous slag, while higher silicon contents can promote pearlite. Selection of an incorrect grade of ferrochromium with high silicon is a common cause for an excessively high silicon content. Manganese can range up to 2.0%. Manganese on the high side of this range can erode acid furnace linings. Therefore, manganese content should be limited to 1% in meltdown; the remainder can be added to the ladle as crushed ferroalloy. Molybdenum is usually added as ferromolybdenum; however, with arc furnaces, molybdenum oxide briquettes may be used. Sulfur is limited to 0.06%, and phosphorus is kept to 0.10%. High superheating temperatures have not been necessary when melting in induction furnaces, which have good stirring action, and a final temperature of 1480  C (2700  F) is usually adequate for thick-section castings. Final temperatures of up to 1565  C (2850  F) are commonly used in arc furnace practice to ensure homogeneity of the bath composition and to accelerate the solution of carbon, added after meltdown, and late alloy additions. No particular problems have been associated with high superheat temperatures, but it is more difficult to predict melting losses. As in steel melting, deoxidation of the bath with aluminum was common in the past, but this practice has been discontinued by most producers, with no adverse effects on the

soundness or mechanical properties of the casting. Titanium is sometimes added to limit dendrite size. There have been reports that bath additions of aluminum and titanium aggravate feeding problems. For heats in which hardenability is marginal, there is evidence that these elements tend to promote pearlite with mold cooling or in heat treated castings.

Pouring Practices High pouring temperatures aggravate shrinkage under feeder heads and other hot spots and can lead to microshrinkage and coarse dendritic structures. Careful control of pouring temperature is particularly important in the production of thick-section castings. Low pouring temperatures are necessary to avoid shrinkage defects and prevent problems such as metal penetration and burned-on sand. Low pouring temperatures are also effective in controlling dendrite size and the coarseness of the eutectic carbide structure. The eutectic temperature for the various high-chromium iron grades ranges from 1230 to 1270  C (2245 to 2315  F), and solidification generally begins at temperatures up to 1350  C (2460  F), depending on composition. Pouring temperatures are seldom lower than 100  C (180  F) above the liquidus temperature. Castings thicker than 102 mm (4 in.) are generally poured between 1345 and 1400  C (2450 and 2550  F). Higher pouring temperatures may be necessary for smaller castings. Casting configurations also must be considered when selecting optimum pouring temperatures. In the ladle or during pouring, the metal appears cold and sluggish because an oxide film readily forms on the surface; but the metal is actually quite fluid and may be poured into intricate shapes. This rather viscous surface oxide is less liable to cause surface defects on castings that are poured quickly. Even when the metal is poured on the cold side, flasks have to be clamped, or weighted, to prevent the metal from running out at the parting line.

Molds, Patterns, and Casting Design Molds must be rigid to minimize shrinkage defects. Molds may be made of either green sand or of dry sand, oil sand, or steel casting sand. Air-

setting and thermal-setting resin sands are also commonly used. Because white irons are subject to hot tearing, cores should break down readily. Unless suitable precautions are taken, the high-chromium irons are somewhat more prone to crack in the foundry than are white irons of lower alloy content. Castings are occasionally found to be cracked when extracted from the mold. Stresses large enough in magnitude to cause fracture can arise when either the mold or cores are excessively strong and there is inadequate mold or core breakdown to allow normal thermal contractions upon cooling. Large cores such as those used in pump volutes, and even small cores used to form bolt holes, can cause cracking of this type. Shrinkage allowance is 18 to 21 mm/m (7/32 to ¼ in./ft). The actual pattern allowance depends on the choice of subsequent heat treatment procedures, if any, and the geometry of the casting. Patterns should employ generous radii and avoid sharp section changes to avoid initiating cracking on solidification and subsequent cooling. These alloys exhibit relatively high liquid-tosolid shrinkage; therefore, large gates and risers are required for feeding. Special consideration should be given to the design of end gates and to the positioning of risers so that they can be readily removed. Risers are often notched or necked down to facilitate removal by impact.

Shakeout Practices The shakeout practice is probably the most critical step in the successful production of high-chromium iron castings. A frequent cause of high residual stresses and of cracking is the common practice of extracting castings from the mold at too high a temperature. Cooling all the way to room temperature in the mold is desirable and can be a requirement to avoid cracking, especially if martensite forms during the last stages of cooling. This precaution is mandatory in heavy-section castings to be used in the as-cast condition where the desired, mold-cooled structure is a mixture of austenite and martensite. For irons that are heat treated, the desired mold-cooled structure is often pearlite. This softer structure facilitates removal of gates and risers, minimizes the transformational and thermal stresses that cause cracking, and shortens the response to heat treatment. Careful design of alloy composition ensures the development of a substantially pearlitic structure in the casting after mold cooling, yet provides enough hardenability to prevent pearlite formation during subsequent heat treatment. Heavy-section castings made pearlitic by such an alloy content can often be removed from the mold once the castings have reached black heat.

Heat Treatment Toughness and abrasion resistance are improved by heat treatment to a martensitic

902 / Cast Irons

Fig. 8

Heat treatment schedule high-chromium irons

for

hardening

Fig. 11

Influence of chromium on the optimum hardening temperature in high-chromium

white iron

Fig. 9

Influence of austenitizing temperature on hardness (H) and retained austenite (g) in high-chromium irons

Fig. 10

Microstructure of heat treated martensitic high-chromium iron illustrating fine secondary M7C3 carbides. Original magnification: 680

microstructure. Figure 8, which illustrates the heat treatment process, shows the importance of slow heating to 650  C (1200  F) to avoid cracking. For complex shapes, a maximum rate of 30  C/h (50  F/h) is recommended. Simple

shapes and fully pearlitic castings can be heated at faster rates. The heating rate can be accelerated above red heat. Austenitization. There is an optimum austenitizing temperature for achieving maximum hardness (Fig. 9). The austenitizing temperature determines the amount of carbon that remains in solution in the austenite matrix. Too high a temperature increases the stability of the austenite, while the higher retained austenite content reduces hardness. Low temperatures result in low-carbon martensite, reducing both hardness and abrasion resistance. Because of this sensitivity to temperature, furnaces that can produce accurate and uniform temperatures are most desirable. The successful heat treatment produces austenite destabilization by precipitation of fine, secondary M7C3 carbides within the austenite matrix (Fig. 10). The optimum austenitizing temperature varies with composition (Fig. 11). Class II irons containing 12 to 20% Cr are austenitized in the temperature range of 955 to 1010  C (1750 to 1850  F). Class III irons containing 23 to 28% Cr are austenitized in the temperature range of 1010 to 1095  C (1850 to 2000  F). Heavy sections usually require temperatures on the upper end of the range. Castings should be held at temperature long enough to accomplish equilibrium dissolution of chromium carbides to ensure a proper hardening response. A minimum of 4 h at temperature is necessary. For heavy sections, the rule of 1 h/in. of section thickness is usually adequate. For castings that are fully pearlitic before heat treatment, the holding time at temperature can be reduced. Quenching. Air quenching (vigorous fan cooling) the castings from the austenitizing temperature to below the pearlite temperature range, that is, to between 550 and 600  C (1020 and 1110  F), is highly recommended. The subsequent cooling rate should be substantially reduced to minimize stresses; still-air or even

furnace cooling to ambient temperature is common. Complex and heavy-section castings are often reinserted in the furnace, which is at 550 to 600  C (1020 to 1110  F), and allowed sufficient time to reach a uniform temperature within the casting. After the temperature is equalized, the castings are either furnace or still-air cooled to ambient temperature. Tempering. Castings can be put into service in the hardened (as-cooled) condition without further tempering or subcritical heat treatments; however, tempering in the range of 205 to 230 C (400 to 450  F) for 2 to 4 h is recommended to restore some toughness in the martensitic matrix and to further relieve residual stresses. The microstructure after hardening always contains retained austenite in the range of 10 to 30%. Some retained austenite is transformed following tempering at low temperatures, but if spalling is a problem, highertemperature tempering can be used to further reduce austenite content. Subcritical Heat Treatment. Subcritical heat treatment (tempering) is sometimes performed, particularly in large heat treated martensitic castings, to reduce retained austenite content and increase resistance to spalling. The tempering parameters necessary to eliminate retained austenite are very sensitive to time and temperature and depend on the casting composition and prior thermal history. Typical tempering temperatures range from 480 to 540  C (900 to 1000  F), and times range from 8 to 12 h. Excess time or temperature results in softening and a drastic reduction in abrasion resistance. Insufficient tempering results in incomplete elimination of austenite. The amount of retained austenite cannot be determined metallographically; those experienced with this heat treatment practice have developed techniques using specialized magnetic instruments to determine the level of retained austenite after tempering. Annealing Castings can be annealed to make them more machinable, either by sub-critical annealing or by a full anneal. Sub-critical annealing is accomplished by pearlitizing by means of soaking in the narrow range between 695 and 705  C (1280 and 1300  F) for 4 to 12 h, which produces hardnesses ranging from 400 to 450 HB. Lower hardnesses often can be achieved with full annealing, in which castings are heated in the range of 955 to 1010  C (1750 to 1850  F), followed by slow cooling to 760  C (1400  F), and holding at this temperature for 10 to 50 h, depending on composition. Annealing affects neither the primary carbides nor the potential for subsequent hardening; guidelines for hardening as-cast castings also apply to annealed castings.

Machining These alloyed white irons are generally ground to finish size; care must be exercised to avoid cracking by overheating. Single-point turning and boring can be done using heavy

White Iron and High-Alloyed Iron Castings / 903 tooling, rigid setups, sufficient power, and carbide- or boride-tipped cutters or special ceramic inserts. Electrical discharge machining and electrochemical machining are also feasible.

Applications The high-chromium white irons are superior in abrasion resistance and are used effectively in impellers and volutes in slurry pumps, classifier wear shoes, brick molds, impeller blades and liners for shot blasting equipment, and refiner disks in pulp refiners. In many applications, they withstand heavy impact loading, such as from impact hammers, roller segments and ring segments in coal-grinding mills, feedend lifter bars and mill liners in ball mills for hard-rock mining, pulverizer rolls, and rolling mill rolls.

Special High-Chromium Iron Alloys Irons for Corrosion Resistance. Alloys with improved resistance to corrosion, for applications such as pumps for handling fly ash, are produced with high chromium content (26 to 28% Cr) and low carbon content (1.6 to 2.0% C). These high-chromium, low-carbon irons provide the maximum chromium content in the matrix. The addition of 2% Mo is recommended for improving resistance to chloridecontaining environments. For this application, fully austenitic matrix structures provide the best resistance to corrosion, but some reduction

in abrasion resistance must be expected. Castings are normally supplied in the as-cast condition. Irons for High-Temperature Service. Because of castability and cost, high-chromium white iron castings can often be used for complex and intricate parts in high-temperature applications at considerable savings compared to stainless steel. These cast iron grades are alloyed with 12 to 39% Cr and are used at temperatures up to 1040  C (1900  F) for scaling resistance. Chromium causes the formation of an adherent, complex, chromium-rich oxide film at high temperatures. The high-chromium irons designated for use at elevated temperatures fall into one of three categories, depending on the matrix structure:
Martensitic irons alloyed with 12 to 28% Cr
Ferritic irons alloyed with 30 to 34% Cr
Austenitic irons, which, in addition to con-

taining 15 to 30% Cr, also contain 10 to 15% Ni to stabilize the austenite phase The carbon contents of these alloys range from 1 to 2%. The choice of an exact composition is critical to the prevention of s-phase formation at intermediate temperatures and, at the same time, avoids the ferrite-to-austenite transformation during thermal cycling, which leads to distortion and cracking. Typical applications include recuperator tubes, breaker bars and trays in sinter furnaces, grates, burner nozzles and other furnace parts, glass bottle molds, and valve seats for combustion engines.

ACKNOWLEDGMENTS The author wishes to thank Warren Spear of the Nickel Development Institute and Ralph Nelson of Thomas Foundries Inc. for helpful discussions. SELECTED REFERENCES
“Abrasion-Resistant Cast Iron. Classifica










tion,” BS ISO 21988:2006, British Standards Institute (ISO Standard), 2007 Abrasion-Resistant Cast Iron Handbook, American Foundry Society, Schaumberg, IL, 2000 Abrasion Resistant Castings for Handling Coal: Properties and Uses of the Ni-Hard Irons (1978), Nickel Institute, www.nickelinstitute.org, accessed June 2008 W. Fairhurst and K. Rohrig, Abrasion-Resistant High-Chromium White Cast Irons, Foundry Trade J., May 1974 “Founding. Abrasion-Resistant Cast Irons,” BS EN 12513:2000, British Standards Institute (EN Standard), 2005 F. Maratray, Choice of Appropriate Compositions for Chromium-Molybdenum White Irons, Trans. AFS, Vol 79, 1971, p 121–124 Ni-Hard Material Data and Applications, Reference Book, Series 11017, Nickel Development Institute, 1996 “Standard Specification for Abrasion-Resistant Cast Irons,” A 532/A 532M, Annual Book of ASTM Standards, Vol 01.02, ASTM International, 2005

ASM Handbook, Volume 15: Casting ASM Handbook Committee, p 904-908 DOI: 10.1361/asmhba0005328

Copyright © 2008 ASM International® All rights reserved. www.asminternational.org

High-Alloy Graphitic Irons Richard B. Gundlach, Stork Climax Research Services

HIGH-ALLOY GRAPHITIC IRONS have found special use primarily in applications requiring corrosion resistance or strength and oxidation resistance in high-temperature service. They are commonly produced in both flake graphite and nodular graphite versions. Those alloys used in applications requiring corrosion resistance are the nickel-alloyed (13 to 36% Ni) gray and ductile irons (also called Ni-Resist irons) and the high-silicon (14.5% Si) gray irons. The alloyed irons produced for high-temperature service are the nickel-alloyed gray and ductile irons, the high-silicon (4 to 6% Si) gray and ductile irons, and the aluminum-alloyed gray and ductile irons. Two groups of aluminum-alloyed irons are recognized: the 1 to 7% Al irons and the 18 to 25% Al irons. Neither the 4 to 6% Si irons nor the aluminum-alloyed irons are covered by ASTM International standards. Although the oxidation resistance of the aluminum-alloyed irons is exceptional, problems in melting and casting the alloys are considerable, and commercial production of the aluminum-alloyed irons is uncommon. In general, the molten-metal processing of high-alloy graphitic irons follows that of conventional gray and ductile iron production. The same attention to chill control and inoculation practice is necessary, and the conditions for successful magnesium treatment for producing nodular graphite are similar or identical to those for conventional gray and ductile irons. The high alloy contents influence the eutectic carbon content; therefore, carbon levels in these alloys are lower. In many cases, hypereutectic compositions (high carbon contents) in the ductile irons are avoided because of a greater tendency toward the formation of graphite flotation and degenerate graphite forms in sections larger than 25 mm (1 in.). The higher alloy contents affect the constitution of the irons, creating conditions that favor the formation of third phases and/or secondary eutectics during solidification. Therefore, many of the alloys commonly contain interdendritic carbides or silicocarbides in the as-cast

structure. These constituents remain after heat treatment and are an accepted part of the microstructure.

(500
F). They have been successfully used for furnace and stoker parts, burner nozzles, and heat treatment trays.

High-Silicon Irons for High-Temperature Service

High-Silicon Ductile Irons

Graphitic irons alloyed with 4 to 6% Si provide good service and low cost in many elevated-temperature applications. These irons, whether gray or ductile, provide good oxidation resistance and stable ferritic matrix structures that will not go through a phase change at temperatures up to 900
C (1650
F). The elevated silicon contents of these otherwise normal cast iron alloys reduce the rate of oxidation at elevated temperatures because silicon promotes the formation of a dense, adherent oxide at the surface, which consists of iron silicate rather than iron oxide. This oxide layer is much more resistant to oxygen penetration, and its effectiveness improves with increasing silicon content. Detailed information on the corrosion-resistant properties of high-silicon cast irons is provided in the article “Corrosion of Cast Irons” in Corrosion: Materials, Volume 13B of ASM Handbook (Ref 1).

High-Silicon Gray Irons The high-silicon gray irons were developed in the 1930s by the British Cast Iron Research Association and are commonly called Silal. In Silal, the advantages of a high critical (A1) temperature, a stable ferritic matrix, and a fine, undercooled type D graphite structure are combined to provide good growth and oxidation resistance. Oxidation resistance is further improved with additions of chromium, which in these grades can approach levels of 2% Cr. An austenitic grade called Nicrosilal was also developed, but the Ni-Resist irons have replaced this alloy. Applications. Although quite brittle at room temperature, the high-silicon gray irons are reasonably tough at temperatures above 260
C

The advent of ductile iron led to the development of high-silicon ductile irons, which currently constitute the greatest tonnage of these types of iron being produced. Converting the eutectic flake graphite network into isolated graphite nodules further improves resistance to oxidation and growth. The higher strength and ductility of the ductile iron version of these alloys qualify them for more rigorous service. The high-silicon ductile iron alloys are designed to extend the upper end of the range of service temperatures for ferritic ductile irons. These irons are used to temperatures of 900
C (1650
F). Raising the silicon content to 4% raises the A1 temperature to 815
C (1500
F), and at 5% Si the A1 is above 870
C (1600
F). The mechanical properties of these alloyed irons at the lower end of the range (4 to 4.5% Si) are similar to those of standard ferritic ductile irons. At 5 to 6% Si, oxidation resistance is improved and critical temperature increased, but the iron can be very brittle at room temperature. At higher silicon levels, the impact transition temperature rises well above room temperature, and the upper shelf energy is dramatically reduced. Ductility is restored when temperatures exceed 425
C (800
F). SAE standard J2582 (Ref 2) defines the three grades of silicon-molybdenum ductile iron most commonly used in automotive applications (Table 1). The three grades are differentiated by molybdenum concentration. For most applications, alloying with 0.5 to 1% Mo provides adequate elevated-temperature strength and creep resistance (Fig. 1). Higher molybdenum additions are used when maximum elevated-temperature strength is needed. High molybdenum additions (>1%) tend to generate interdendritic carbides of the Mo2C type, which persist even through annealing, and tend to reduce toughness

High-Alloy Graphitic Irons / 905 Table 1

Grades of high-silicon-molybdenum ductile iron (SAE J2582) Composition, wt%

Grade

1 2 3

C

Si

Mn

Mo

S

P

Mg

3.3–3.8 3.3–3.8 3.3–3.8

3.5–4.5 3.5–4.5 3.5–4.5

0.1–0.5 0.1–0.5 0.1–0.5

0.5 max 0.51–0.70 0.71–1.00

0.035 max 0.035 max 0.035 max

0.05 max 0.05 max 0.05 max

0.025–0.060 0.025–0.060 0.025–0.060

Source: Ref 2

Fig. 2

Micrograph of a 4Si-Mo ductile iron showing nodular graphite structure. Original magnification: 400

Fig. 1

Influence of molybdenum on (a) tensile properties and (b) creep resistance of 4% Si ductile iron at 705
C (1300
F)

and ductility at room temperature. Figure 2 shows the typical microstructure obtained in high-silicon-molybdenum ductile iron. Silicon lowers the eutectic carbon content, which must be controlled to avoid graphite flotation. For 4% Si irons, carbon content should range from 3.2 to 3.5% C, depending on section size, and at 5% Si it should be approximately 2.9% C. The unique properties of each element and its contribution to the total alloy system of

a nodular graphite cast iron are described in the article “Ductile Iron Castings” in this Volume. Melting Practice. For the high-silicon ductile irons, standard ductile iron melting practices apply. Cupola melting is acceptable, but these irons are commonly electric melted. Acid, neutral, or basic linings are used. Conventional ductile iron charge materials are also used; however, care should be taken to minimize chromium, manganese, and phosphorus. Manganese content should not exceed 0.5%, and it is preferable to keep it below 0.3%. Chromium should be no more than 0.1% and preferably below 0.06%. These elements promote as-cast embrittlement due to carbides and pearlite, which form networks in the interdendritic regions. These microstructural constituents are difficult to break down through subsequent heat treatment, and they degrade toughness and machinability. They are a particular problem in turbocharger applications, in which minimum ductility requirements must be met. Consequently, a larger proportion of pig iron in the charge is common. Silicon additions can be made in the ladle, but it is highly recommended that all silicon be added to the furnace, except for that required for magnesium treatment and postinoculation.

The same is true for molybdenum additions; either ferromolybdenum or molybdic oxide briquettes can be used. Conventional magnesium treatment techniques are used, and the same rules for balancing sulfur and setting residual magnesium levels apply to the high-silicon irons. These irons are more susceptible to the formation of intercellular carbides when magnesium levels exceed those needed to balance sulfur and to nodularize the graphite. Although these irons respond like standard ductile irons to most inoculants, foundry-grade ferrosilicon, containing 75 to 85% Si, is most often used for postinoculation. Pouring Practice. Due to increased silicon contents, higher pouring temperatures (>1425
C, or 2600
F) are required to develop a clean melt surface in the ladle. Therefore, pouring temperatures somewhat higher than those for conventional ductile irons should be used to minimize dross and slag defects. Molds, Patterns, and Casting Design. These alloys are normally sand cast; sand molds can be made of green sand, shell mold, and air-set chemically bonded sands suitable for gray and ductile irons. As with any graphitic iron, high mold rigidity is necessary to minimize mold wall movement and shrinkage porosity. Softer molds will produce bigger castings, but in general, little or no allowance for shrinkage is required in the pattern. As with conventional ductile irons, the highly reactive magnesium dissolved in these irons, along with the high levels of silicon, gives rise to the formation of dross and inclusions due to the reaction with oxygen in the air during pouring and filling of the mold. Consequently, clean ladles with teapot designs or skimmers are recommended. Furthermore, measures must be taken in designing the runners and gating system to minimize turbulence and to trap dross before it enters the mold cavity. Shakeout Practice. As mentioned previously, room-temperature impact resistance is low; therefore, riser and gate removal is somewhat easier with these alloys than with standard ductile iron grades. These irons are quite ductile at elevated temperatures, and they should be allowed to cool before cleaning. They do exhibit a ductility trough in the temperature range of 315 to 425
C (600 to 800
F), and riser and gate removal is aided when performed upon cooling through this temperature range. Heat Treatment. The high-silicon ductile irons are predominantly ferritic as-cast, but the presence of carbide-stabilizing elements will result in a certain amount of pearlite and often intercellular carbides. These alloys are inherently more brittle than standard grades of ductile iron and may have higher levels of internal stress due to lower thermal conductivity and higher elevated-temperature strength. These factors should be taken into account when selecting heat treatment requirements. High-temperature heat treatment is often advised to anneal any pearlite and to stabilize the casting against growth in service. A normal

906 / Cast Irons graphitizing (full) anneal in the austenitic temperature range is recommended when undesirable amounts of carbide are present. For the 4 to 5% Si irons, this will require heating to at least 900
C (1650
F) for several hours, followed by slow cooling to below 705
C (1300
F). At higher silicon contents (>5%), in which carbides readily break down, and in castings that are relatively carbide free, subcritical annealing in the temperature range of 720 to 790
C (1325 to 1450
F) for 4 h is effective in ferritizing the matrix. Compared to full annealing, the subcritically annealed material will have somewhat higher strength, but ductility and toughness will be somewhat reduced. Applications. The high-silicon and siliconmolybdenum ductile irons are currently produced as manifolds and turbocharger housings for trucks and some automobiles. They are also used in heat treating racks.

Austenitic Nickel-Alloyed Gray and Ductile Irons The nickel-alloyed austenitic irons are produced in both gray and ductile cast iron versions for high-temperature service. Austenitic gray irons date back to the 1930s, when they were specialized materials of minor importance. After the invention of ductile iron, austenitic grades of ductile iron were also developed. These nickel-alloyed austenitic irons have been used in applications requiring corrosion resistance, wear resistance, and high-temperature stability and strength (Ref 1). Additional advantages include low thermal expansion coefficients,

nonmagnetic properties, and cast iron materials having good toughness at low temperatures. When compared with corrosion and heat-resistant steels, nickel-alloyed irons have excellent castability and machinability. Applications of Nickel-Alloyed Irons. The nickel-alloyed (Ni-Resist) irons have found wide application in chemical process-related equipment such as compressors and blowers; condenser parts; phosphate furnace parts; pipe, valves, and fittings; pots and retorts; and pump casings and impellers. Similarly, in foodhandling equipment, the various alloy components include bottling and brewing equipment, canning machinery, distillery equipment, feed screws, metal grinders, and salt filters. In hightemperature applications, nickel-alloyed irons are used as cylinder liners, exhaust manifolds, valve guides, gas turbine housings, turbocharger housings, nozzle rings, water pump bodies, and piston rings in aluminum pistons. Figure 3 shows the typical microstructure obtained in Ni-Resist irons.

Austenitic Gray Irons Specification ASTM A 436 (Ref 3) defines eight types of austenitic gray iron alloys, four of which are designed to be used in elevatedtemperature applications and four in applications requiring corrosion resistance (Table 2). Nickel additions produce a stable austenitic microstructure with good corrosion resistance and strength at elevated temperatures. The nickel-alloyed irons are also alloyed with chromium and silicon for wear resistance and

oxidation resistance at elevated temperatures. Types 1 and 1b, which are designed for corrosion-resistant applications, are alloyed with 13.5 to 17.5% Ni and 6.5% Cu. Type 1 alloys are used to produce ring carriers used in conjunction with aluminum pistons in diesel engines. Types 2b, 3, and 5, which are principally used for high-temperature service, contain 18 to 36% Ni, 1 to 2.8% Si, and 0 to 6% Cr. With the development of ductile iron, most high-temperature applications shifted to similar Ni-Resist ductile iron grades. Type 4 alloys are alloyed with 29 to 32% Ni, 5 to 6% Si, and 4.5 to 5.5% Cr and are recommended for their stain-resistant properties.

Austenitic Ductile Irons Specification ASTM A 439 (Ref 4) defines the group of austenitic ductile irons (Table 3). The austenitic ductile iron alloys have similar compositions to the austenitic gray iron alloys but have been treated with magnesium to produce nodular graphite structures. Ductile iron alloys are available in every gray iron alloy type but type 1; this is because the high copper content of type 1 is not compatible with the production of nodular graphite iron. The ductile iron alloys have high strength and ductility, combined with the same desirable properties of the gray iron alloys. They provide frictional wear resistance, corrosion resistance, strength and oxidation resistance at high temperatures, nonmagnetic characteristics, and, in some alloys, low thermal expansivity at ambient temperatures.

Table 2 Compositions of austenitic gray iron alloys Composition, % Alloy

Type Type Type Type Type Type Type Type

C

1 1b 2 2b 3 4 5 6

3.00 3.00 3.00 3.00 2.60 2.60 2.40 3.00

max max max max max max max max

Si

Mn

Ni

Cu

Cr

1.00–2.80 1.00–2.80 1.00–2.80 1.00–2.80 1.00–2.00 5.00–6.00 1.00–2.00 1.50–2.50

0.5–1.5 0.5–1.5 0.5–1.5 0.5–1.5 0.5–1.5 0.5–1.5 0.5–1.5 0.5–1.5

13.50–17.50 13.50–17.50 18.00–22.00 18.00–22.00 28.00–32.00 29.00–32.00 34.00–36.00 18.00–22.00

5.50–7.50 5.50–7.50 0.50 max 0.50 max 0.50 max 0.50 max 0.50 max 3.50–5.50

1.5–2.5 2.50–3.50 1.5–2.5 3.00–6.00(a) 2.50–3.50 4.50–5.50 0.10 max 1.00–2.00

S

0.12 0.12 0.12 0.12 0.12 0.12 0.12 0.12

Mo

max max max max max max max max

... ... ... ... ... ... ... 1.00 max

(a) Where some machining is required, the 3.00–4.00% Cr range is recommended. Source: Ref 3

Table 3 Compositions of austenitic nodular irons Composition, %

Fig. 3

Micrograph of a D-5S Ni-Resist ductile iron showing nodular graphite structure. Original magnification: 400

Alloy type

C

Si

Mn

D-2(a) D-2B D-2C D-2M(b) D-3(a) D3-A D-4 D-5 D-5B D-5S

3.00 max 3.00 max 2.90 max 2.2–2.7 2.60 max 2.60 max 2.60 max 2.40 max 2.40 max 2.30 max

1.50–3.00 1.50–3.00 1.00–3.00 1.5–2.50 1.00–2.80 1.00–2.80 5.00–6.00 1.00–2.80 1.00–2.80 4.90–5.50

0.70–1.25 0.70–1.25 1.80–2.40 3.75–4.5 1.00 max(c) 1.00 max(c) 1.00 max(c) 1.00 max(c) 1.00 max(c) 1.00 max(c)

P

0.08 0.08 0.08 0.08 0.08 0.08 0.08 0.08 0.08 0.08

max max max max max max max max max max

Ni

Cr

18.00–22.00 18.00–22.00 21.00–24.00 21.0–24.0 28.00–32.00 28.00–32.00 28.00–32.00 34.00–36.00 34.00–36.00 34.00–37.00

1.75–2.75 2.75–4.00 0.50 max 0.20 max(c) 2.50–3.50 1.00–1.50 4.50–5.50 0.10 max 2.00–3.00 1.75–2.25

(a) Additions of 0.7–1.0% Mo will increase the mechanical properties above 425
C (800
F). (b) D-2M source: Ref 5; all others: Ref 4. (c) Not intentionally added

High-Alloy Graphitic Irons / 907 Austenitic ductile irons suitable for low-temperature service are specified by ASTM A 571/ A 571M (Ref 5). These are designated D-2M, with two strength classes in customary units and two classes in metric units. They are intended for pressure-containing parts in cryogenic applications. Melting Practice. In the past, these iron alloys were generally cupola melted, but today (2008), melting is almost exclusively done in electric furnaces. Choice of furnace linings is usually based on other alloys being melted in the shop; acid, neutral, or basic linings are used. Selection of charge materials is more critical in melting the higher-nickel alloys of the ductile iron versions, because they are more sensitive to the tramp elements that affect graphite structure. The high nickel content causes the materials to be more prone to hydrogen gas defects; therefore, charge materials should be thoroughly dry, and melting times should be short. The molten iron should only be superheated to the temperature necessary to treat and pour, and for as short a time as possible. Pouring is generally done above 1400
C (2550
F) to keep the molten iron surface clean and free of oxide. Magnesium treatment of the ductile iron alloys is normally accomplished with nickelmagnesium alloys. The nickel-magnesium treatment alloy is often added in the furnace; there is no concern for pyrotechnics. The same rules for balancing sulfur and setting residual magnesium levels apply to these nickel-alloyed irons. These irons are more susceptible to the formation of intercellular carbides when magnesium levels exceed those needed to balance sulfur and to nodularize the graphite. Although these irons respond like standard ductile irons to most inoculants, foundry-grade ferrosilicon, containing 75 to 85% Si, is most often used for postinoculation. Postinoculation is most often performed when tapping the furnace. Stream inoculation, where feasible, is also recommended for improved machinability. Carbon content for the type D-5S alloy must be monitored carefully, because section sensitivity is high. Although the ASTM specification allows up to 2.4% C, sections over 25 mm (1 in.) are susceptible to exploded and chunky graphite formation. Carbon levels of 1.6 to 1.8% are recommended for heavy-section castings. Molding and Casting Design. Conventional ferrous molding sands are used, including green sand, shell mold, and chemically bonded sands. The same precautions taken for high-strength irons apply to these alloys. Solidification should progress from thin to thick sections without interruption. Abrupt changes in section thickness should be avoided. Riser location should allow convenient access for riser removal. The shrinkage allowance is generally 21 mm/m ð1=4 in:=ftÞ, or 2.1%. Shakeout. These alloys can develop large thermal stresses upon cooling because of a

relatively low thermal conductivity, combined with high elevated-temperature strength and high thermal expansion rates. Consequently, care should be taken in shaking out too hot, and mold cooling is recommended for intricately shaped castings and castings of widely varying section thickness. Heat treatment of nickel-alloyed ductile irons serves to strengthen the casting and to stabilize the microstructure of the casting for increased durability. Stress-relief heat treatments are typically conducted at temperatures between 620 and 675
C (1150 and 1250
F) to remove residual casting stresses. Mold cooling to 315
C (600
F) is a satisfactory alternative to furnace stress relief. Annealing of some castings may be necessary to reduce hardness. Annealing is performed at 955 to 1040
C (1750 to 1900
F) for 30 min to 5 h, and this treatment will normally break down some of the carbides and spheroidize the rest. Heat treatment for stability of the microstructure for service at temperatures of 480
C (900
F) and above is performed by heating at 760
C (1400
F) for a minimum of 4 h and furnace cooling to 540
C (1000
F), followed by air cooling. An alternative treatment is to heat at 900
C (1650
F) for 2 h and furnace cool to 540
C (1000
F). These treatments stabilize the microstructure and minimize growth and warpage in service. The treatments are designed to reduce carbon levels in the matrix, and some growth and distortion often accompany heat treatment. Type 1 alloys are not generally amenable to this stabilizing treatment. Dimensional stability, when truly critical, can be ensured by heat treating at 870
C (1600
F) or higher for 2 h plus an additional hour per 25 mm (1 in.) of cross section, furnace cooling not faster than 55
C/h (100
F/h) to 540
C (1000
F), and then holding for 1 h per 25 mm (1 in.) of cross section and cooling uniformly. After rough machining, the casting should be reheated to 455 to 480
C (850 to 900
F) and held 1 h per 25 mm (1 in.) of cross section. Refrigeration and reaustenitization heat treatments are applied to type D-2 alloys to increase yield strength. Solution heat treating at 925
C (1700
F), refrigerating at 195
C (320
F), and then reheating between 650 and 760
C (1200 and 1400
F) will increase yield strength considerably without materially affecting magnetic properties or corrosion resistance in seawater or dilute sulfuric acid.

Detailed information on the heat treatment of ductile iron is available in the article “Ductile Iron Castings” in this Volume.

Aluminum-Alloyed Irons The aluminum-alloyed irons consist of two groups of gray and ductile irons. The lowalloyed group contains 1 to 7% Al, and the aluminum essentially replaces silicon as the graphitizing element in these alloys. The highalloyed group contains 18 to 22% Al. Irons alloyed with aluminum in between these two ranges will be white iron as-cast and have no commercial importance. The aluminum greatly enhances oxidation resistance at elevated temperatures and also strongly stabilizes the ferrite phase to very high temperatures—up to and beyond 980
C (1800
F). Like the silicon-alloyed irons, the aluminum irons form a tight, adherent oxide on the surface of the casting that is very resistant to further oxygen penetration. Unfortunately, the aluminum-alloyed irons are very difficult to cast without dross inclusions and laps (cold shuts). The aluminum in the iron is very reactive at the temperatures of the molten iron, and contact with air and moisture must be negligible. Care must be taken not to draw the oxide skin, which forms during pouring, into the mold in order to avoid dross inclusions. Methods for overcoming these problems in commercial practice are under development. At present, there is no ASTM standard covering the chemistry and expected properties of these alloys, and commercial production is very limited. In the past, the 1.5 to 2.0% Al irons have been used in the production of truck exhaust manifolds.

High-Silicon Irons for Corrosion Resistance Irons with high silicon contents (14.5% Si) constitute a unique corrosion-resistant ferritic cast iron group. These alloys are widely used in the chemical industry for processing and for transporting highly corrosive liquids. They are particularly suited to handling sulfuric and nitric acids (Ref 6). The three most common high-silicon iron alloys are covered in ASTM A 518/A 518M (Ref 7). These alloys (Table 4) contain 14.2 to 14.75% Si and 0.7 to 1.15% C.

Table 4 Compositions of high-silicon iron alloys Composition, % Alloy

Grade 1 Grade 2 Grade 3 Source: Ref 7

C

Mn

Si

Cr

Mo

Cu

0.65–1.10 0.75–1.15 0.70–1.10

1.50 max 1.50 max 1.50 max

14.20–14.75 14.20–14.75 14.20–14.75

0.50 max 3.25–5.00 3.25–5.00

0.50 max 0.40–0.60 0.20 max

0.50 max 0.50 max 0.50 max

908 / Cast Irons Grades 2 and 3 are also alloyed with 3.25 to 5% Cr, and grade 2 contains 0.4 to 0.6% Mo. Other compositions are also commercially produced with up to 17% Si. Melting Practice. Induction melting is the preferred method for these alloys. Induction melting permits the very tight control of chemistry needed in melting these materials in order to minimize scrap losses due to cracking. The alloys have a very low tolerance for hydrogen and nitrogen; thus, charge materials must be carefully controlled to minimize the levels of these gases. Vacuum treatment can be used to increase strength and density. The melting point of the eutectic 14.3% Si cast iron is approximately 1180
C (2160
F), and the alloy is generally poured at approximately 1345
C (2450
F). Because of the very brittle nature of high-silicon cast iron, castings are usually shaken out after cooling to ambient temperature. However, some casting geometries demand hot shakeout while the castings are still red hot, so that the castings can be immediately stress relieved and furnace cooled to prevent cracking. Molding and Casting Design. These alloys are routinely cast in sand molds, investment molds, and permanent steel molds. Cores must have good collapsibility to prevent fracture during solidification. Flash must be minimized because it will be chilled iron and can readily become an ideal nucleation site for cracks that propagate into the casting. Casting design for these high-silicon irons requires special considerations. To avoid cracking, sharp corners and abrupt changes in section size must be avoided. Casting designs should have tapered ingates to permit easy removal of gates and risers by impact and to minimize the amount of grinding required. Unless vacuum

degassing is employed, conventional risers are not generally used or desired. High-silicon cast irons are generally cast in sections ranging from 4.8 to 38 mm (3=16 to 1.5 in.). Thin sections fill well because of the excellent fluidity of these cast irons, but tend to chill and become extremely brittle white iron. Sections over 38 mm (1.5 in.) are prone to segregation and porosity, which reduce strength and corrosion resistance. Castings are generally designed with a patternmaker’s shrink rule of 18 mm/m (7=32 in./ft), or 1.8%. Heat Treatment. High-silicon irons can be stress-relieved by heating in the range of 870 to 900
C (1600 to 1650
F), followed by slow cooling to ambient temperatures to minimize the likelihood of cracking. Heat treatments have no significant effect on corrosion resistance. Machinability. High-silicon cast irons have hardnesses of approximately 500 HB and are normally considered machinable only by grinding. Machinability can be improved with higher additions of carbon and/or phosphorus, but only at the expense of other properties, such as strength and corrosion resistance. Applications. High-silicon irons are extensively used in equipment for the production of sulfuric and nitric acids, for sewage disposal and water treatment, for handling mineral acids in petroleum refining, and in the manufacture of fertilizer, textiles, and explosives. Specific components include pump rotors, agitators, crucibles, and pipe fittings in chemical laboratories.

REFERENCES 1. T.C. Spence, Corrosion of Cast Irons, Corrosion: Materials, Vol 13B, ASM Handbook, ASM International, 2005, p 43–50 2. “Automotive Ductile Iron Castings for High Temperature Applications,” J2582, Vol 1, SAE, Dec 2001, p 6.17 3. “Standard Specification for Austenitic Gray Iron Castings,” A 436, Annual Book of ASTM Standards, Vol 01.02, ASTM International, 2005 4. “Standard Specification for Austenitic Ductile Iron Castings,” A 439, Annual Book of ASTM Standards, Vol 01.02, ASTM International, 2005 5. “Standard Specification for Austenitic Ductile Iron Castings for Pressure-Containing Parts Suitable for Low-Temperature Service,” A 571/A 571M, Annual Book of ASTM Standards, Vol 01.02, ASTM International, 2005 6. S.K. Brubaker, Corrosion by Sulfuric Acid, Corrosion: Environments and Industries, Vol 13C, ASM Handbook, ASM International, 2006, p 659–667 7. “Standard Specification for Corrosion-Resistant High-Silicon Iron Castings,” A 518/A 518M, Annual Book of ASTM Standards, Vol 01.02, ASTM International, 2005 SELECTED REFERENCES  “High Temperature Materials for Exhaust

ACKNOWLEDGMENTS The author wishes to thank Warren Spear of The Nickel Development Institute and Don Stickle of the Duriron Company for their valuable assistance.

Manifolds,” J2515, Aug 1999, SAE Ferrous Materials Standards Manual, 2004 ed., HS30, SAE, p 365–374  “Properties and Applications of Ni-Resist and Ductile Ni-Resist Alloys,” Nickel Development Institute, Toronto, Ontario, Canada, 1998

ASM Handbook, Volume 15: Casting ASM Handbook Committee, p 911-917 DOI: 10.1361/asmhba0005295

Copyright © 2008 ASM International® All rights reserved. www.asminternational.org

Steel Ingot Casting Donald J. Hurtuk

WHEN A HEAT OF STEEL is melted and refined, it is necessary to solidify it into useful forms for further processing or final use. Although most steel is now solidified in continuous casting operations, the original method for casting steel involved solidifying into ingots. This is still the preferred method for certain specialty, tool, forging, and remelted steels. The molten steel is transferred from the steelmaking vessel to a refractory-lined ladle, from which it is teemed into containers, called ingot molds, for solidification. Some of the methods, equipment, and theory for pouring and solidifying ingots are presented in this article.

Pouring Method

through a nozzle at the bottom of the ladle, but lip pouring by tilting the ladle is also done in small operations. The flow of steel from the bottom of a ladle is controlled by a stopper rod or, more commonly, a slide gate nozzle. Lip pouring from the ladle is controlled by the skill of the crane operator. The pouring rate is quite fast in top pouring, achieving a rate of rise of the steel in the mold in the range of 125 to 255 cm/min (50 to 100 in./ min). The steel stream from the ladle to the mold will oxidize during pouring, and considerable turbulence is generated in the mold. This method is damaging to the ingot mold, and mold iron consumption rates can be quite high. This method also produces a lower-quality ingot because of the oxidation and turbulence. However, toppoured steel can be poured quite rapidly. When

ingot molds are placed on a string of railroad cars, called a drag, this can be a very efficient way of pouring steel. Bottom pouring (or uphill teeming) involves teeming through a runner system that fills the ingot molds from the bottom until they are full, as shown in Fig. 2. This method generates much less turbulence than top pouring and produces better surface and internal quality in the ingots. It is also less damaging to ingot molds but does require additional equipment to direct the steel into the molds. A fountain or trumpet is required to funnel the steel from the ladle to a sprue plate or stool that has separate channels, or runners, to direct the steel to the molds. The trumpet and runners are refractory lined, and refractory ingates or outlets direct the steel into the ingot mold cavity. Bottom pouring has the

It has long been recognized that the way in which steel is poured to produce ingots is important. Of critical importance is the way in which steel fills the mold. Two basic types of pouring methods have been used for transferring steel from the ladle to the ingot mold. They are top pouring and bottom pouring. Top pouring (or direct teeming) involves teeming from the ladle directly into the top opening of an ingot mold until the mold is full, as shown in Fig. 1. Pouring is normally done

Fig. 1

Method for top pouring, or direct teeming, of steel ingots. Source: Ref 1

Fig. 2

Schematic arrangement and nomenclature for bottom pouring steel ingots. Source: Ref 2

912 / Steel Castings added advantage of filling multiple ingot molds at one time, as shown in Fig. 3. A flux is used in bottom pouring for several reasons. These include:
Prevent oxidation of the steel meniscus
Insulate the rising metal surface
Lubricate the flow of the steel meniscus

against the mold wall


Trap and absorb nonmetallic inclusions (Ref 1)

The flux is normally suspended in bags from the top of the mold and is designed to spread over the surface and begin melting as the molten steel fills the mold. The melted flux fills the gap between the steel and the mold, which improves thermal conductivity and lubricates the surface to generate excellent surface quality. The melting flux at the top helps remove inclusions from the steel to improve steel cleanliness. This is depicted in the drawings in Fig. 4. The application rate of bottom-poured fluxes is generally in the range of 1.0 to 2.3 kg/ton (2 to 5 lb/ton). Removal of nonmetallic inclusions can occur when the inclusions reach the flux/steel interface and become absorbed and removed from the steel. This absorption is controlled by surface tension according to the following relationship: sm  ss  si 0 > 0 where sm is the surface tension of the steel, ss is the surface tension of the slag, and si is the surface tension of the slag/steel interface. If the surface tension of the inclusions is similar to or greater than that of the steel, the inclusions will be absorbed into the slag (Ref 1). With good steelmaking and steel pouring practices, bottom pouring has been able to achieve steel cleanliness levels rivaling remelt quality. The grade of steel, pouring rate (or rate of rise in the mold), and flux properties must be carefully matched to achieve the best results. The pouring rate is much slower than in top pouring. Sometimes, the rate of rise of the steel in the mold must be kept as low as 50 to 75 mm/min (2 to 3 in./min) for particularly crack-sensitive grades. Normal bottom-pouring rates are in the 200 to 300 mm/min (8 to 12 in./min) range.

include mold designs that are open top, bottle top, open bottom, closed bottom, and plug bottom. Electrode ingots are tall, round ingots intended for remelting in electroslag remelting and vacuum arc remelting operations. These steels are very high quality and are applied to some of the most critical applications, such as aerospace and medical. Tool steel ingots are usually small versions of BEU designs for high-quality specialty applications. Ingots are often shaped on the top or bottom to maximize yields by optimizing deformation during rolling. These designs are incorporated into the design of the ingot molds and stools that are used to contain the molten steel during casting. Traditionally, ingots were categorized by the amount of gases released during solidification. The main categories are killed, semikilled, capped, and rimmed; examples are shown in Fig. 6. In this figure, ingot 1 is an example of

Fig. 3

Schematic for bottom pouring multiple ingots. Source: Ref 1

Fig. 4

Drawings depicting the role of bottom-poured flux in bottom pouring. Source: Ref 1

Fig. 5

Cross sections of various types of big-end-down and big-end-up mold designs. Source: Ref 2

Ingot Design The design of the ingot is dictated by the application and type of steel involved. A big-end-up (BEU) design is used to promote ingot soundness and ingot quality by establishing a solidification profile that promotes moving segregation and shrinkage out of the ingot to the hot top, or head, of the ingot. Some of the highest-quality steels are cast in this design. A big-end-down (BED) design is used to produce solidification structures that can be reduced in rolling for plate and sheet applications. With this design, centerline quality is not critical, since it will undergo significant reduction prior to achieving final size. The BEU and BED mold designs are shown in Fig. 5. These

Steel Ingot Casting / 913 a fully killed ingot where no gas is evolved during solidification. Ingots 2 to 4 are examples of semikilled ingots with increasing amounts of gas evolution. Ingot 5 is an example of a capped ingot where gas evolution is sufficient to alter solidification structures at the surface in the upper portions of the ingot. Ingots 6 to 8 are rimmed ingots with increasing amounts of gas evolution. Many of these types of ingots have now been replaced by continuous casting, where yield and quality are considerably higher. Fully killed steel ingots now represent the bulk of steel ingots cast in the steel industry.

Hot Tops Hot tops (or heads) are used on steel ingots to control the heat flow at the top of the ingot during solidification. This is done to improve the quality and yield of the ingot by providing a pool of liquid steel to feed the shrinkage as the ingot solidifies and to provide a region for segregates and nonmetallic inclusions to gather. During ingot solidification, heat flows from the hot top region, or head, in three ways:
Up to the atmosphere
Down to the ingot
Out through the sides

Hot top systems are designed to minimize the first two avenues and maximize the third. The more heat flow that can be directed down to the ingot and out through the mold walls, the better the resulting ingot quality and yield. Hot tops consist of sideboards on the sidewalls of the very top of the ingot and a topping compound or board on top of the ingot. Sideboards can be pure insulating or exothermic and are attached to the ingot mold walls or to a separate casting that sits on top of the mold. Topping compounds are generally powders placed on top of the ingots immediately after pouring. Sometimes, the topping compound can be in a preformed board much like the sideboards. Sometimes, hot tops are not used, which will change the way the ingot solidifies. Examples of hot-topped and non-hot-topped ingots are shown in Fig. 7. It is evident that the use of

hot tops confines the shrinkage that occurs during solidification to the very top of the ingot. This helps improve quality and yield. When high-temperature molten steel comes in contact with exothermic hot top materials, the following reaction occurs: FeO þ Al ! Fe þAl2 O3 þ Heat The generation of heat locally retards solidification of the steel, and the resulting properties of the exothermic materials following the exothermic reaction are highly insulating. This keeps the top of the ingot liquid while the body of the ingot solidifies. If the hot top is properly designed, it will keep the top of the ingot liquid until all of the body of the ingot has solidified. This will allow all shrinkage in the ingot to be fed from the hot top, so little or no shrinkage forms in the ingot itself. It also allows segregation and impurities to be pushed out of the ingot and into the head. By balancing the thermal properties of the sideboards and compound, the solidification of the ingot can be controlled to optimize quality and yield. Pure insulating hot top sideboards and compound produce similar effects to exothermic hot tops. However, without the heat generated by the exothermic reaction, insulating hot tops are less effective and are rarely used anymore, except in very large ingots. Insulating hot tops can be sufficient in very large ingots, where solidification times are long and heat retention in the ingot is extended. The hot top sideboards are generally secured in the mold or hot top casting prior to teeming. They begin to function as soon as teeming is complete. However, the topping compound must be added immediately after teeming is complete. It is critical that this be done as quickly as possible to avoid the start of solidification in the head of the ingot. The same is true if the topping compound is in a preformed board. If the topping compound is added immediately after teeming, solidification in the head will be retarded, and the ingot will solidify correctly.

Ingot Molds and Stools Ingot molds are containers into which molten steel is poured to solidify into workable ingots. Stools are plates on which the ingot molds are

sometimes placed. Ingot molds must be able to withstand the high temperature of liquid steel yet be able to conduct and radiate away the intense heat so the steel can solidify. For this reason, molds and stools are generally made of some form of cast iron that has the properties to satisfy these requirements. The most common ingot mold material is gray iron, comprised of large graphite flakes in a matrix of various portions of ferrite and pearlite. The graphite flakes provide excellent thermal conductivity to remove the intense heat of the liquid steel. Ferrite provides thermal shock resistance to avoid catastrophic failure when the steel is teemed into the mold. Pearlite gives the iron erosion resistance from the damaging effects of the hot steel stream. Compacted graphite iron and nodular or ductile iron have also been successfully used in ingot molds. The modified shape of the graphite in these types of irons reduces the thermal conductivity, but the erosion resistance and toughness of the iron increase significantly. Best results with these irons have been obtained when the ingot mold is of a symmetrical design, such as round or square and not too tall. Other designs tend to warp over time and cause significant operating problems. When these irons are used correctly, the life of ingot molds is generally extended over gray iron molds. These irons have not been as successfully used in stool applications but have had limited success in specialty applications. Ingot molds are recycled for use, and the cycle time between uses is a critical factor in the casting process. The cycle time must be appropriate so that the initial mold temperature at teeming is within an acceptable range. Molds that are too hot or too cold can cause serious operating and quality problems. An inventory of ingot molds must be sufficient to allow proper cycling between uses. The same is true for stools. The design of ingot molds is critical for ingot quality and mold life. The weight of the mold is generally 1 to 1.4 times the weight of the ingot, depending on the design. The BEU and closedbottom molds are at the higher end of this range, while BED and open-bottom molds are

Fig. 7

Fig. 6

Series of structures in various types of ingots. See the text for explanation. Source: Ref 2

Depiction of hot-topped and non-hot-topped killed steel ingots. 1, big-end-up, hot topped; 2, big-end-down, hot topped; 3, big-end-up, not hot topped; 4, big-end-down, not hot topped. Source: Ref 2

914 / Steel Castings at the lower end of this range. To attain uniform heat flow during solidification of the ingot, the mold midface is usually thicker than the corners. A normal ratio of corner-to-midface thickness is 0.87 to 0.92, while lower ratios are necessary when using modified flake irons, such as compacted graphite or ductile iron. The corner-to-midface ratio can have a significant impact on solidification structures, as shown in Fig. 8. When the mold is designed properly, heat flow during ingot solidification is uniform, resulting in uniform ingot structures. Considerable work has been performed to optimize mold design relative to ingot solidification. Figure 9 shows designs by A.P. Banks in a 1967 British patent. Factors such as grade

of steel, pouring temperature, and type of mold iron must be considered to fully optimize mold design. Some molds have corrugated inside surfaces to improve surface quality and crack resistance in the ingot. However, these usually shorten the life of ingot molds, since the greater surface area is prone to deteriorate faster over time than smooth-wall molds. Hot top castings that create the head in ingot casting are usually made of gray iron. Modified graphite irons have had some success in this application as well. Sprue plates are bottom plates that contain the runners in bottom pouring that feed steel to the ingot molds. Trumpets are refractory-lined castings that funnel the steel from the ladle to the sprue plate and ingot molds. Sprue plates and trumpets are usually made of gray iron and experience long life.

Ingot Solidification Ingot solidification takes place as heat is removed from the steel through the ingot mold to the surrounding atmosphere. Ingot solidification patterns follow the configuration of the inside cavity of the ingot mold and are further influenced by the hot top. Typical solidification of an 81 cm (32 in.) square BEU hot-topped ingot is shown in Fig. 10. The rate of heat removal controls the resulting solidification structures that form in the ingot. Normal structures usually include a chill zone, a columnar dendritic zone, and an equiaxed zone, as shown in Fig. 11. The chill zone forms rapidly as the hot liquid steel contacts the relatively cold mold wall, creating a

Fig. 8

Influence of mold corner-to-midface ratio on ingot solidification. Structures in ductile iron molds. Source: Ref 1

Fig. 9

Mold designs to optimize heat flow. Source: Ref 3

very high thermal gradient at the ingot surface. This zone of planar solidification develops just before an air gap forms between the steel and the mold wall as the solidified steel shrinks away from the mold wall. The structures quickly transition to columnar dendritic structures as the thermal gradient drops and stabilizes somewhat. This still-high thermal gradient is the driving force for the columnar dendritic structures, according to the principle of constitutional supercooling, as depicted in Fig. 12. Segregates ahead of the advancing interface in the region of high thermal gradient keep the liquid steel supercooled, which promotes the columnar growth. As the thermal gradient decreases as solidification progresses, this driving force disappears, and the structures become equiaxed toward the center of the ingot. The structures in continuous casting are identical to those in ingot casting, as shown in the macroetched cross section in Fig. 13. Considerable work has been done in recent years to study continuously cast solidification structures. Most of what has been learned is directly transferable to statically cast ingots. Extensive columnar structures promote segregation and bridging along the centerline. It is generally preferred to control these structures to minimize centerline segregation and shrinkage. The temperature of the liquid steel has an effect on these structures. It is known that superheat directly affects the length of the columnar dendrites, as shown in Fig. 14. Reduced superheat can be used as a way to control columnar structures. Steel chemistry also affects columnar structures. Carbon content is the most notable element and influences

Steel Ingot Casting / 915

Time after pour (minutes)

66 20

100 60

140

120 80 40

160

140

20

60 100

40 80 120

72

60

54

48

42

36

120

Height (inches)

140

30

24

100

18

80

12

Fig. 12

Constitutional supercooling in alloy solidification. Source: Ref 5

60 40 20

6

0

Fig. 10

Solidification patterns in an 81 cm (32 in.) square bottom-end-up hot-topped ingot of killed steel. Source: Ref 2

Fig. 14

Fig. 13

Macroetched cross section of a 22.2 by 25 cm (8.75 by 10 in.) continuously cast bloom in the as-cast condition. Source: Ref 6

Fig. 11 Ref 4

Sketch of ingot structures showing chill zone, columnar zone, and equiaxed zone. Source:

columnar dendrites according to the carbon content, as shown in Fig. 15. This effect is controlled by the peritectic reaction that steel undergoes upon solidification, which is: Lþd!g where L is the liquid, d is delta iron, and g is austenite. The high-temperature portion of the iron-carbon equilibrium phase diagram, where this reaction occurs, is shown in Fig. 16. This

Influence of superheat on columnar structures. Source: Ref 6

reaction is important because significant shrinkage occurs as the body-centered cubic crystal structure of d-ferrite transforms to the more dense face-centered cubic crystal structure of austenite. This is the driving force behind much of the shrinkage that occurs during solidification and the formation of the air gap between the mold and ingot. The shrinkage is largest at approximately 0.10% C, at the onset of the peritectic reaction. Columnar structures are minimized, but surface quality suffers at this carbon content. Shrinkage is least at approximately 0.60% C, beyond the limits of the peritectic reaction. Columnar structures are maximized

916 / Steel Castings

Fig. 15

Plot showing influence of carbon content on columnar structures. Source: Ref 6

at this carbon content, since better contact between the ingot and mold is maintained. This keeps the thermal gradient high and provides the driving force for columnar structures. Surface quality is usually better at this carbon content. Beyond 0.60% C, columnar structures are reduced, but, of course, there is more carbon to segregate. Centerline carbon segregation can be quite troublesome in steels above 1.0% C. Nickel content also influences columnar dendrites, much as carbon does, as seen in Fig. 17. This effect is also controlled by a peritectic reaction, and the limits of the reaction in ironnickel at 4.0 and 5.4% Ni represent the concentrations of minimized and maximized columnar

structures, respectively, as shown in Fig. 18. The mechanism involved is the same as in the case of carbon. Most other solute elements have the effect of refining solidification structures by promoting equiaxed structures with increasing concentration. Columnar dendritic solidification structures are depicted in Fig. 19. These dendrites have primary and secondary arms. New primary arms can form by branching from secondary arms. Sometimes, tertiary arms form on secondary arms. All of this generates considerable segregation during solidification and creates a structure that can easily entrap inclusions. Columnar structures will increase centerline segregation and can even bridge at the centerline to create shrinkage. Equiaxed structures reduce these problems. Examples are shown in Fig. 20. It is generally considered best to produce as many equiaxed structures as possible to maximize ingot quality.

Ingot Stripping

Fig. 16

High-temperature portion of the iron-carbon equilibrium phase diagram. Source: Ref 7

Once ingot solidification is complete, and in some cases even earlier, the ingot is removed from the mold by a process called stripping. For BED designs, the ingot mold is usually lifted from the ingot by an overhead crane. For BEU designs, the ingot can be extracted from the ingot mold by an overhead crane. Sometimes, it is necessary to invert the ingot mold and dump the ingot out. Ingots can be stripped before solidification is complete to shorten the cycle time for mold reuse. This is normally done with BED ingots, where center quality is less important. Movement of an ingot with liquid centers will certainly disrupt the solidification front and create segregation and possibly shrinkage problems. However, if the ingot will undergo significant reduction into certain flat rolled products, this will not adversely affect the quality of the final product. Ingots are then generally charged into a soaking pit for reheating before further processing by rolling. Ingots for remelt or forging are slow cooled until they reach ambient temperature and can then be further processed.

Solidification Simulation

Fig. 17

Plot showing influence of nickel content on columnar structures. Source: Ref 6

Ingot producers have found value in simulating the solidification of ingots by various available simulation packages when considering new ingots or designs. Some simulations have remarkable accuracy in predicting solidification patterns and times. It has been found that seemingly small changes to hot top design, ingot design, pouring temperature, and so on can have disastrous effects on ingot quality and yield. A consideration of these possible effects by simulation before application in the plant can avoid many problems and expense.

Steel Ingot Casting / 917

Fig. 19

Columnar dendritic solidification in alloys. Source: Ref 5

REFERENCES

Fig. 18

Iron-nickel equilibrium phase diagram. Source: Ref 8

Fig. 20

Schematic illustrations of columnar and equiaxed ingot structures. Source: Ref 5

1. R.L. Harvey and A.P. Banks, “Bottom Pouring of Steel Ingots,” Foseco Inc. Seminar (Cleveland, OH), 1983 2. H.E. McGannon, Ed., The Making, Shaping and Treating of Steel, 8th ed., United States Steel Corporation, Pittsburgh, PA, 1964 3. A.P. Banks, British Patent 1,086,946, 1967 4. T.F. Bower and M.C. Flemings, Trans. AIME, Vol 239, 1967, p 1620 5. M.C. Flemings, Solidification Processing, McGraw-Hill, Inc., 1974 6. D.J. Hurtuk, “Factors Influencing Steel Solidification Structures,” Ph.D. thesis, Case Western Reserve University, Cleveland, OH, 1981 7. J. Chipman, Thermodynamics and Phase Diagram of the Fe-C System, Met. Trans., Vol 3, 1972, p 55–64 8. J.I. Goldstein and R.E. Ogilvie, A Re-Evaluation of the Iron Rich Portion of the Fe-Ni System, Trans. Met. Soc. AIME, Vol 233, 1965, p 2083

ASM Handbook, Volume 15: Casting ASM Handbook Committee, p 918-925 DOI: 10.1361/asmhba0005296

Copyright © 2008 ASM International® All rights reserved. www.asminternational.org

Steel Continuous Casting Robert D. Pehlke, University of Michigan

THE ADVANTAGES OF CONTINUOUS CASTING in primary metals production have been recognized for more than a century. In recent decades, a dramatic growth of this processing technology has been realized in ferrous as well as nonferrous metal production. The principal advantages of continuous casting are a substantial increase in yield, a more uniform product, energy savings, and higher manpower productivity. These advantages and the ease of integration into metals production systems have led to the wide application of continuous casting processes.

Historical Aspects of Continuous Casting of Steel One of the earliest references to continuous casting is a patent granted in 1840 to George Sellers, who had developed a machine for continuously casting lead pipe (Ref 1). There is some indication that this process had been underway before Sellers’s patent, which was directed toward improvement of this continuous casting process. The first work on continuous casting of steel was by Sir Henry Bessemer, who patented a process for “manufacture of continuous sheets of iron and steel” in 1846 and made plant trials on continuous casting of steel in the 1890s (Ref 2). Although continuous casting had its start before the beginning of the 20th century, it was not until the mid-1930s in Germany that commercial production of continuously cast brass billets was introduced. Sigfried Junghans, an active inventor of casting technology, provided many improvements in the process, in particular the introduction of the oscillatingmold system to prevent the casting from sticking to the mold. Further development of the process for the casting of nonferrous metals continued, including the installation of processing units in North America. Mold lubrication in the form of oil, or, more recently, lowmelting slag powders, was introduced. Taper of the mold to compensate for metal shrinkage on solidification provided improved heat transfer and, more importantly, fewer cracks. In 1935, a plant with casting rolls for continuous production of brass plates was operated

at Scovill Manufacturing in the United States, and the Vereinigte Leichtmetallwerke in 1936 started a semicontinuous casting machine for aluminum alloys. Immediately after World War II, commercial development of continuous casting of steel began in earnest, with pilot plants at Babcock and Wilcox Company (United States), Low Moor (Great Britain), Amagasaki (Japan), Eisenwerk Breitenfeld (Austria), BISRA (Great Britain), and Allegheny Ludlum Corporation (United States). These were followed by production plants for casting billets in the West and stainless slabs in the Soviet Union and Canada (the latter at Atlas Steels) (Ref 3). The Schneckenburger and Kung patent on the curved strand was filed in Switzerland in 1956, and production was commercialized with a billet machine at Von Moosche Eisenwerke (Switzerland) in 1963. In 1961, at Dillingen Steelworks (West Germany), the first verticaltype large slab machine with bending of the strand to horizontal discharge was started up. In 1964, Shelton Iron and Steel (Great Britain) was the first new steelworks to turn out its entire production by continuous casting, consisting of four machines with eleven strands for medium to very large bloom sizes, and operating in connection with Kaldo converters. That same year, the first Concast S-type curved-mold machine for large slabs was started up at Dillingen Steelworks (West Germany). The height of this type of machine was less than 50% of the corresponding height of a vertical type of machine. In the same year, a bow-type slab caster was presented by Mannesmann (West Germany). In 1968, McLouth Steel (United States) started up four curved-strand slab casting machines immediately after the first four-strand low-head caster for large slabs in the West was started up at the Weirton Steel Division of National Steel (United States). Subsequently, National pioneered the casting of slabs for tinplate applications (Ref 3). Continuous casting of steel is entering a new era of development, not only with respect to its increasing application in the production process but also in its own evolution as a process and its interaction with other processes in steel manufacture. Continuous casting output has shown an accelerating growth curve. More than 80%

of current world steel production is continuously cast, and continuous casting in Japan exceeds 90%. The advantages of the process, along with its developments and current challenges for improvement, are outlined in the following sections.

General Description of the Process The purpose of continuous casting is to bypass conventional ingot casting and to cast to a form that is directly rollable on finishing mills. The use of this process has resulted in improvement in yield, surface condition, and internal quality of product when compared to the previous ingot-made material. Continuous casting involves the following sequence of operations:  Delivery of liquid metal to the casting strand  Flow of metal through a distributor (tundish)

into the casting mold

 Formation of the cast section in a water-

cooled copper mold

 Continuous withdrawal of the casting from

the mold

 Further heat removal to solidify the liquid

core of the casting by water spraying beyond the mold  Cutting to length and removing the cast sections The main components of continuous casting machines are presented in Fig. 1. Molten steel in a ladle is delivered to a reservoir above the continuous casting machine, called a tundish. The flow of steel from the tundish into one or more open-ended, water-cooled copper molds is controlled by a stopper rod-nozzle or a slide gate valve arrangement. To initiate a cast, a starter, or dummy bar, is inserted into the mold and sealed so that the initial flow of steel is contained in the mold and a solid skin is formed. After the mold has been filled to the desired height, the dummy bar is gradually withdrawn at the same rate that molten steel is added to the mold. The initial liquid steel freezes onto a suitable attachment of the dummy bar so that the cast strand can be

Steel Continuous Casting / 919

Fig. 1

Main components of a curved continuous casting strand. Source: Ref 4

Fig. 2

Principal types of continuous casting. V, vertical; VB, vertical with bending; VPB, vertical with progressive bending; CAS, circular arc with straight mold; CAC, circular arc with curved mold; PBC, progressive bending with curved mold; H, horizontal. Source: Ref 5

withdrawn down through the machine. Solidification of a shell begins immediately at the surface of the copper mold. The length of the mold and the casting speed are such that the shell thickness is capable of withstanding the pressures of the molten metal core upon exiting from the copper mold. To prevent sticking of the frozen shell to the copper mold, the mold is normally oscillated during the casting operation, and a lubricant is added to the mold. The steel strand is mechanically supported by rolls below the mold, where secondary cooling is achieved by spraying cooling water onto the strand surface to complete the solidification process. After the strand has fully solidified, it is sectioned into desired lengths by a cutoff torch or shear. This final portion of the continuous casting machine also has provision for disengagement and storage of the dummy bar. Several arrangements are now in commercial use for the continuous casting of steel. The types of continuous casting machines in use include vertical, vertical with bending, curved or S-strand with either straight or curved mold, curved strand with continuous bending, and horizontal. Examples of the principal types of

machines currently producing slabs are shown in Fig. 2. Most of the original continuous casting machines for steel were vertical machines. Vertical machines with bending and curved strand machines, although more complicated in their construction, were developed to minimize the height of the machine and allow installation in existing plants without modification of crane height. Four basic caster designs for slabs are shown in Fig. 3, with an indication of the required installation height and the corresponding solidification distance or metallurgical length (mL).

Plant Layout The design and layout of a steelmaking facility often focuses initially on continuous casting. The optimal plant layout varies markedly from one installation to another. One major factor in the configuration is whether or not the steelmaking complex is being constructed on a greenfield site or being added (“shoehorned”) into an existing works. Many of the major

integrated steel works in Japan were constructed as greenfield installations during the period from 1960 to 1975. Most of the minimills constructed throughout the world, and in particular in the United States, were also built on greenfield sites. In building a greenfield site, the plant layout should incorporate two major features: a smooth and well-organized arrangement for materials handling and flow, and the capacity for future expansion. Generally, these plants are designed for 100% continuous casting, and no ingot facilities are included (Ref 7). Nearly all of the recently built minimills incorporate one or more electric furnaces and provide for 100% continuous casting of billets. Chaparral Steel Company at Midlothian, TX, is an excellent example (Ref 8). A profile of a proposed large-scale electric furnace billet casting plant is shown in Fig. 4. A twin-strand slab caster was shoehorned into the Number 4 basic oxygen furnace shop at the previous Inland Steel Company. The addition of this caster substantially increased the output of this facility, where ingot casting was the rate-limiting step in production. The arrangement of this installation, showing the caster and ingot facilities, is presented in Fig. 5. An important characteristic of the plant layout, and in particular of the materials handling facilities, is the concept that the continuous casting machine cannot wait. This design and operating philosophy has had a dramatic impact on steelmaking operations, which have now become a synergism to the continuous casting facilities. Because of this shift in priorities, marked improvements in productivity have been developed for continuous casting, as outlined subsequently. Dramatic increases in energy costs, as well as the desire for higher productivity, led to the development of the “hot connection.” Substantial energy savings can be achieved by directly charging the hot continuously cast slab or billet to the reheating furnaces of the rolling mill. The latest installations have included direct in-line hot rolling of the cast product.

Near-Net Shape Casting A current trend in the steel industry is to develop processes for casting steel close to final product shape, that is, near-net shape casting. A conventional caster produces slabs 150 to 350 mm (6 to 14 in.) thick, which are used to produce hot rolled strip 2.5 to 25 mm (0.10 to 1.0 in.) thick. The amount of energy to reduce the thick slabs to 50 mm (2 in.) thick represents a significant part of the total production cost. Consequently, processes for production of thinner primary thicknesses are actively being pursued. It is anticipated that following thin-slab production will come processes for strip (5 to 20 mm, or 0.2 to 0.8 in.) and thin strip (0.6%) and silicon (0.3 to 0.8%) additions are limited by other alloy effects and are normally inadequate alone to prevent pinholes. Aluminum is the most common supplemental deoxidizer used to prevent pinholes. As little as 0.01% Al will prevent pinholes. Aluminum is normally added at the tap—approximately 1 kg/Mg (2 lb/ton) (0.10% added) with a recovery of 30 to 50% for a final content of

Shape Casting of Steel / 933

Fig. 13

Fig. 14

Effect of slag basicity (V-ratio) and bath silicon on the retention of chromium in a reduced slag. Source: Ref 5

Chromium and silicon contents of liquid iron in equilibrium with silica-saturated slags

approximately 0.03 to 0.05%. This is normally supplemented at the pour ladle with additional deoxidation, which could be more aluminum or calcium, barium, silicon, manganese, rare earths, titanium, or zirconium. More than the minimum amount of deoxidizer required for preventing porosity is needed to maintain sulfide inclusion shape control, but excessive amounts can cause intergranular failures or dirty metal. The second addition at the pour ladle is normally 1 to 3 kg/Mg (2 to 7 lb/ton). The commonly used elements for deoxidation are (in order of decreasing power) zirconium, aluminum, titanium, silicon, carbon, and manganese. Argon-Oxygen Decarburization (AOD). Some foundries have recently installed AOD units to achieve some of the results that vacuum melting can produce. These units look very much like Bessemer converters with tuyeres in the lower sidewalls for the injection of argon or nitrogen and oxygen. They are processing units that must be charged with molten metal from an arc or induction furnace. Up to approximately 20% cold charge can be added to an

AOD unit; however, the cold charge is usually less than 20% and consists of solid ferroalloys. The continuous injection of gases causes a violent stirring action and intimate mixing of slag and metal, which can lower sulfur values to below 0.005%. The gas contents approach or may be even lower than those of vacuum induction melted steel. The dilution of oxygen with inert gas, argon, or nitrogen causes the carbon-oxygen reaction to go to completion in favor of the oxidation reaction of iron and the oxidizable elements, notably, chromium in stainless steel. Therefore, superior chromium recoveries from less expensive high-carbon ferrochromium are obtained compared to those of electric arc melting practices. Argon-oxygen decarburization units are used in the production of high-alloy castings, particularly of grades that are prone to defects due to high gas contents. Carbon and low-alloy steels for castings with heavy wall sections may be subject to hydrogen embrittlement and are also processed in these units with good results. Powder Injection/Wire Injection. The powder injection or wire injection of reactive metals (calcium, magnesium, or rare earths) into liquid steel for desulfurization has come to the forefront of technology in the past decade. Ultralow levels of sulfur (1040  C (1900  F), WQ >1040  C (1900  F), WQ >1040  C (1900  F), WQ

531 600 531

77 87 77

248 290 255

36 42 37

60 50 55

... ... ...

140 160 140

>1040  C (1900  F), WQ

Keyhole notch V-notch V-notch Keyhole notch Keyhole notch Keyhole notch V-notch V-notch Keyhole notch Keyhole notch Keyhole notch V-notch Keyhole notch Izod Vnotch Keyhole notch

149.2 110 135.6 100 100.3 74

586

85

310

45

50

...

156

94.9

70





531

77

248

36

50

...

163

81.4

60





>1095 C (2000 F), WQ

CF3M CF3MA CF8M

>1040 C (1900 F), WQ >1040  C (1900  F), WQ >1065  C (1950  F), WQ

552 621 552

80 90 80

262 310 290

38 45 42

55 45 50

... ... ...

150 170 170

CF8C

>1065  C (1950  F), WQ

CF16F

162.7 120 135.6 100 94.9 70

531

77

262

38

39

...

149

40.7

30





531

77

276

40

52

...

150

101.7

75





>1095 C (2000 F), WQ

CG8M CH20

>1040 C (1900 F), WQ >1095  C (2000  F), WQ

565 607

82 88

303 345

44 50

45 38

... ...

176 190

108.5 40.7

80 30

CK20

1150  C (2100  F), WQ

524

76

262

38

37

...

144

67.8

50

476

69

214

31

48

...

130

94.9

70

CN7M

Specimen





1120 C (2050 F), WQ

(a) AC, air cool; FC, furnace cool; OQ, oil quench; WQ, water quench; T, temper; A, age

Fig. 24

Typical yield strength values for austenitic manganese steels. See Fig. 2 for key.

grades receive a high-temperature solution anneal. This treatment consists of holding the casting at a temperature that is high enough to dissolve all chromium carbides, which are damaging to intergranular corrosion resistance, and then cooling them rapidly enough to avoid reprecipitation of the carbides by quenching in water, oil, or air. Although this can be accomplished throughout in the lower-carbon grades (0.15%N) AC from 1040 ⬚C (1900F)(>2%Ni;>0.15%N)

Alloy

CE30 (a) (b) CF8 CF2 CF8M, CF12M CF8 CF16F CH20 CK20 CN7M

Heat treatment

As-cast WQ from 1065–1120 ⬚C(1950–2050 ⬚F) WQ from 1065–1120 ⬚C(1950–2050 ⬚F) WQ from above 1095 ⬚C(2000 ⬚F) WQ from 1065–1150 ⬚C(1950–2100 ⬚F) WQ from 1065–1120 ⬚C(1950–2050 ⬚F) WQ from above 1095 ⬚C(2000 ⬚F) WQ from above 1095 ⬚C(2000 ⬚F) WQ from above 1150 ⬚C(2100 ⬚F) WQ from above 1065–1120 ⬚C(1950–2050 ⬚F)

Fig. 25

Mechanical properties of cast corrosion-resistant steels at room temperature. (a) Tensile strength. (b) 0.2% offset yield strength. (c) Charpy keyhole impact energy. (d) Brinell hardness. (e) Elongation. Also given are the heat treatments used for test materials: AC, air cool; FC, furnace cool; WQ, water quench; A, anneal; T, temper.

CF8C, CF8M, CE30A, and CA15 are currently used in high-pressure service at temperatures to 540  C (1000  F) in sulfurous acid environments in the petrochemical industry. Other uses

are in the power-generating industry at temperatures to 565  C (1050  F). Room-temperature properties in the aged condition, that is, after exposure to elevated-service

Elevated-Temperature Tensile Properties. The short-term elevated-temperature test, in which a standard tension test bar is heated to a designated uniform temperature and then strained to fracture at a standardized rate, identifies the stress due to a short-term overload that will cause fracture in uniaxial loading. The manner in which the values of tensile strength and ductility change with increasing temperature is shown in Fig. 29 for alloy HP50WZ. Representative tensile properties at temperatures between 650 and 1095  C (1200 and 2000  F) are shown in Table 15 for several heat-resistant alloy steel grades. Creep and Stress-Rupture Properties. Creep is defined as the time-dependent strain that occurs under load at elevated temperature. Creep is operative in most applications of heat-resistant high-alloy castings at the normal service temperatures. In time, creep may lead to excessive deformation and even fracture at stresses considerably below those determined in room-temperature and elevated-temperature short-term tension tests. The designer must usually determine whether the serviceability of the component in question is limited by the rate or the degree of deformation. When the rate or degree of deformation is the limiting factor, the design stress is based on the minimum creep

Steel Castings Properties / 967

Transformation of d-ferrite to austenite and sphase upon exposure of a solution-treated CF8 casting to elevated temperature

Fig. 28

Fig. 26

Effects of time at elevated temperature on the tensile properties of static and centrifugal CF8 alloy castings. Parts had a ferrite number of 9 to 11 and contained 0.081% N.

Fig. 29

Fig. 27

Effect of time at elevated temperature on the room-temperature impact strength, ferrite number and stresscorrosion cracking (SCC) resistance of CF8 castings with a ferrite number of 9 to 11 and nitrogen content of 0.081%. Arrows indicate no failure in SCC testing after 336 h.

Tensile properties versus temperature for heatresistant alloy HP50WZ

rate and design life after allowing for initial transient creep. The stress that produces a specified minimum creep rate of an alloy or a specified amount of creep deformation in a given time (for example, 1% total creep in 100,000 h) is referred to as the limiting creep strength or limiting stress. The manner in which the minimum creep rate depends on the applied stress is illustrated in Fig. 30 by data for alloy HP50WZ. When fracture is the limiting factor, stressto-rupture values can be used in design (Fig. 31). The stress-to-rupture values can be combined with those for minimum creep rate, as shown in Fig. 32. It should be recognized that long-term creep and stress-rupture values (for example, 100,000 h) are often extrapolated from shorter-term tests. Whether these property

968 / Steel Castings Table 15 Representative short-term tensile properties of cast heat-resistant alloys at elevated temperatures Property at indicated temperature 650  C (1200  F) Ultimate tensile strength

Yield strength

870  C (1600  F)

MPa

ksi

MPa

... 10 ... 14 ... ... ... ... ... 5 ... ... 8

159 145 127 148 179 161 210 140 179 130 135 131 141

23 21 18.5 21.5 26 23 30.5 20 26 19 19.5 19 20.5

... 107 93 110 ... 101 ... 100 121 103 ... 103 121

MPa

HD HF HH (type I) HH (type II) HI HK HL HN HP HT HU HW HX

... ... ... ... 414 60 217 31.5 ... ... ... ... 417 60.5 222 32 ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... 292 42.5 193 28 ... ... ... ... ... ... ... ... 303 45 138 20

Fig. 30

Minimum creep rate versus temperature for alloy HP50WZ

MPa

ksi

Yield strength

Elongation, %

Alloy

ksi

Ultimate tensile strength

1095  C (2000  F) Ultimate tensile strength

Yield strength

ksi

Elongation, %

MPa

ksi

MPa

ksi

Elongation, %

... 15.5 13.5 16 ... 15 ... 14.5 17.5 15 ... 15 17.5

18 16 30 18 12 16 ... 37 27 24 20 ... 48

... ... ... 38 ... 39 ... 43 52 41 ... ... ...

... ... ... 5.5 ... 5.5 ... 6 7.5 6 ... ... ...

... ... ... ... ... 34 ... 34 43 ... ... ... ...

... ... ... ... ... 5 ... 5 6 ... ... ... ...

... ... ... ... ... 55 ... 55 69 ... ... ... ...

stressed

Fig. 32

Creep-rupture properties of alloy HK40. Scatter bands are þ20% of the central  tendency line. Although such a range usually encompasses data for similar alloy compositions, scatter of values may be much higher, especially at longer times and high temperatures.

Fig. 31

Rupture time versus stress and temperature for alloy HP50WZ

values are extrapolated or determined directly often has little bearing on the operating life of high-temperature parts. The actual material behavior is often difficult to predict accurately because of the complexity of the service stresses relative to the idealized, uniaxial loading conditions in the standardized tests and because of the attenuating factors such as cyclic loading, temperature fluctuations, or metal loss from corrosion. The designer should anticipate the synergistic effects of these variables. Thermal Fatigue Resistance. The design of components that are subject to considerable temperature cycling must also include consideration of thermal fatigue. This is particularly true if the temperature changes are frequent or rapid and nonuniform within or between casting sections. Fatigue is a condition in which failure results from alternating load applications in shorter times or at lower stresses than would

be expected from constant-load properties. Thermal fatigue denotes the conditions in which the stresses are primarily due to hindered thermal expansion or contraction. Good design helps minimize external restraints to expansion and contraction. Rapid heating and cooling may, however, impose temperature gradients within the part, causing the relatively cool elements of the component to restrain the hotter elements. Finite element analysis has shown that, for some industrial applications, these thermally induced stresses may exceed those resulting from mechanical loads. Nevertheless, such test results have been useful in considering alloy selection questions, identifying the superior thermal fatigue resistance of nickel-containing grades, and documenting the good performance of some HH-type compositions. Thermal Shock Resistance. Thermal shock failure may occur as a result of a single, rapid temperature change or as a result of rapid cyclic temperature changes, which induce stresses that are high enough to cause failure. Brittle ceramic materials are subject to such failure in a single temperature cycle, but only very

infrequently are conditions encountered that would cause failure in a single thermal cycle of the cast heat-resistant grades. The nickel-predominating grades are generally the most resistant to thermal shock. Resistance to Hot Gas Corrosion. The corrosion of heat-resistant alloys, that is, their attack by the environment at elevated temperatures, varies significantly with alloy type, temperature, velocity, and the nature of the specific environment to which the part is exposed. Table 16 presents a general ranking of the standard cast heat-resistant grades in various environments.

Ferrite in Cast Stainless Steels The CF alloys constitute the most technologically important and highest-tonnage segment of corrosion-resistant casting production. These 19Cr-9Ni alloys are the cast counterparts of the AISI 300-series wrought stainless steels (Table 2). In general, the cast and wrought alloys possess equivalent resistance to corrosive media, and they are frequently used in conjunction with each other. Important differences do exist, however, between the cast CF alloys and their wrought AISI counterparts. Most significant among these is the difference in alloy microstructure in the end-use condition. The CF-grade cast alloys have duplex structures (Table 2) and usually contain 5 to 40% ferrite, depending on the particular alloy. Their wrought counterparts are fully austenitic. The ferrite in cast stainless with duplex structures is magnetic, a point that is often confusing when cast stainless steels are compared to their wrought counterparts by checking their attraction to a magnet. This difference in microstructures is attributable to the fact that the chemical compositions of the cast and wrought alloys are not identical by intent. Significance of Ferrite. Ferrite is intentionally present in cast CF-grade stainless steels for three principal reasons: to provide strength, to improve weldability, and to maximize resistance to corrosion in specific environments. Strengthening in the cast CF-grade alloys is limited essentially to that which can be gained by incorporating ferrite into the austenitematrix phase. These alloys cannot be strengthened by thermal treatment, as with the cast martensitic alloys, nor by hot or cold working, as with the wrought austenitic alloys. Strengthening by carbide precipitation is also out of the question because of the detrimental effect of carbides on corrosion resistance in most aqueous environments. Thus, the alloys are effectively strengthened by balancing alloy composition to produce a duplex microstructure consisting of ferrite (up to 40% by volume) distributed in an austenite matrix. It has been shown that the incorporation of ferrite into 19Cr-9Ni cast steels improves yield and tensile strengths substantially without loss of ductility or impact toughness at temperatures below

Steel Castings Properties / 969 Table 16 Corrosion resistance of heat-resistant cast steels at 980  C (1800  F) in 100 h tests in various atmospheres Corrosion rating(a) in indicated atmosphere Alloy

Air

Oxidizing flue gas(b)

HA HC HD HE HF HH HI HK HL HN HP HT HU HW HX

U G G G S G G G G G G G G G G

U G G G G G G G G G G G G G G

Reducing flue gas(b)

Reducing flue gas(c)

Reducing flue gas (constant temperature)(d)

Reducing flue gas cooled to 150  C (300  F) every 12 h(d)

U G G G S G G G G G G G G G G

U S S ... U S S U S U G U U U S

U G G G S G G G G S G S S U G

U G G ... S G G G G S ... U U U U

(a) G, good (corrosion rate r < 1.27 mm/yr, or 50 mils/yr); S, satisfactory (r < 2.54 mm/yr, or 100 mils/yr); U, unsatisfactory (r > 2.54 mm/yr, or 100 mils/yr). (b) Contained 2 g of suifur/m3 (5 grains S/100 ft3). (c) Contained 120 g S/m3 (300 grains S/100 ft3). (d) Contained 40 g S/m3 (100 grains S/100 ft3)

resistance is impractical or impossible. In the case of SCC, the presence of ferrite pools in the austenite matrix is thought to block or make more difficult the propagation of cracks. In the case of intergranular corrosion, ferrite is helpful in sensitized castings that may become sensitized. The carbides precipitate at the discontinuous ferrite/austenite boundaries rather than at the continuous austenite/austenite grain boundaries, where they would increase susceptibility to intergranular attack. The presence of ferrite also places additional phase boundaries in the austenite matrix, and there is evidence that intergranular attack is arrested at austenite-ferrite boundaries. The most comprehensive study of the effect of ferrite on the corrosion resistance of cast stainless steels indicates that ferrite:

Fig. 33

Yield strength and tensile strength versus percentage of ferrite for CF8 and CF8M alloys. Curves are mean values for 277 heats of CF8 and 62 heats of CF8M.

425  C (800  F). The magnitude of this strengthening effect for CF8 and CF8M alloys at room temperature is illustrated in Fig. 33. Fully austenitic stainless steels are susceptible to a weldability problem known as hot cracking or microfissuring. The intergranular cracking occurs in the weld deposit and/or in the weld heat-affected zone and can be avoided if the composition of the filler metal is controlled to produce approximately 4% ferrite in the austenitic weld deposit. Duplex CF-grade alloy castings are immune to this problem. The presence of ferrite in duplex CF alloys improves resistance to SCC and generally to intergranular attack. Although failures of highalloy castings due to these two types of corrosion are not common, SCC and intergranular attack are concerns because they can occur unexpectedly, particularly in castings that have been sensitized by welding in the field, where postweld heat treatment to restore corrosion


Improves the resistance of CF alloys to chlo-

ride SCC
Improves resistance of these alloys to intergranular attack
Affords greater operating safety for the CF alloys with respect to both types of attack at ferrite contents exceeding 10% It is important to note, however, that not all studies have shown ferrite to be unconditionally beneficial to the general corrosion resistance of cast stainless steels. Whether or not corrosion resistance is improved by ferrite and to what degree depends on the specific alloy composition and heat treatment and the service conditions (environment and stress state). Ferrite Control. From the preceding discussion, it is apparent that controlled ferrite contents in predominantly austenitic cast chromium-nickel steels, notably, the CF alloys, offer certain property advantages and that the amount of ferrite present will depend primarily on the compositional balance of the alloy. The underlying causes for the dependence of ferrite content on composition are found in the phase equilibria for the Fe-Cr-Ni system. These phase

equilibria have been exhaustively documented and related to commercial stainless steels. The major elemental components of cast stainless steels are in competition to promote austenite or ferrite phases in the alloy microstructure. Chromium, silicon, molybdenum, and niobium promote the presence of ferrite in the alloy microstructure; nickel, carbon, nitrogen, and manganese promote the presence of austenite. By balancing the contents of ferrite-forming and austenite-forming elements within the specified ranges for the elements in a given alloy, it is possible to control the amount of ferrite present in the austenite matrix. The alloy can usually be made fully austenitic or with ferrite contents up to 30% or more in the austenite matrix. The relationship between composition and microstructure in cast stainless steels permits the foundry metallurgist to predict and control the ferrite content of an alloy, as well as its resultant properties, by adjusting the composition of the alloy. This is accomplished with the Schoefer constitution diagram for cast chromium-nickel alloys (Fig. 34). This diagram was derived from an earlier diagram developed by Schaeffler for stainless steel weld metal. Use of Fig. 34 requires that all ferrite-stabilizing elements in the composition be converted into chromium equivalents and that all austenite-stabilizing elements be converted into nickel equivalents by means of empirically derived coefficients representing the ferritizing or austenitizing power of each element. A composition ratio is then obtained from the total chromium equivalent, Cre, and nickel equivalent, Nie, calculated for the alloy composition according to the following: Cre ¼ %Cr þ 1:5ð%SiÞ þ 1:4ð%MoÞ þ %Nb  4:99 Nie ¼ %Ni þ 30ð%CÞ þ 0:5ð%MnÞ þ 26ð%N  0:02Þ þ 2:77

(Eq 1)

(Eq 2)

where the elemental concentrations are given in weight percent. Although similar expressions have been derived that take into account additional alloying elements and different compositional ranges in the Fe-Cr-Ni alloy system, use of the Schoefer diagram has become standard for estimating and controlling ferrite content in stainless steel castings. See ASTM A 800/A 800M for the standard practice of estimating ferrite content in certain ausyenitic Fe-Cr-Ni alloys. The Schoefer diagram possesses obvious utility for casting users and the foundry metallurgists. It is useful for estimating or predicting ferrite content if the alloy composition is known, and it is useful for setting nominal values for individual elements in calculating the furnace charge for an alloy in which a specified ferrite range is desired. Limits of Ferrite Control. Although ferrite content can be estimated and controlled on the basis of alloy composition only, there are limits to the accuracy with which this can be done. The reasons for this are many. First, there is an

970 / Steel Castings

Fig. 34

Schoefer diagram for estimating the ferrite content of steel castings in the composition range 16 to 26% Cr, 6 to 14% Ni, 4% Mo (max), 1% Nb (max), 0.2% C (max), 0.19% N (max), 2% Mn (max), and 2% Si (max). Dashed lines denote scatter bands caused by the uncertainty of the chemical analysis of individual elements. See text for equations used to calculate Cre and Nie.

Fig. 35

Schematic showing heating/cooling cycles used for various heat treatments

unavoidable degree of uncertainty in the chemical analysis of an alloy (note the scatter band in Fig. 34). Second, the ferrite content depends on thermal history in addition to composition, although to a lesser extent. Third, ferrite contents at different locations in individual castings can vary considerably, depending on section size, ferrite orientation, presence of alloying element segregation, and other factors. Measurements of ferrite content in stainless steel castings are also subject to significant limitations. Magnetic measurements of ferrite content are limited to small material volumes

and require simple casting geometries. In addition, careful calibration with primary and secondary standards is required for accuracy of measurement. Quantitative metallographic determinations of ferrite content on polished surfaces are essentially impossible to conduct in a nondestructive fashion with respect to the casting. The metallographic approach is also quite time-consuming, is limited by alloy etching characteristics and microscope resolution, and is complicated by the fact that it is a two-dimensional technique while ferrite pools and colonies in the alloy structure are three-dimensional. Both the foundry metallurgist and user of stainless steel castings should recognize that the aforementioned factors place significant limits on the degree to which ferrite content (either as ferrite number or ferrite percentage) can be specified and controlled in stainless steel castings. In general, the accuracy of ferrite measurement and the precision of ferrite control diminish as the ferrite number increases. As a working rule, it is suggested that  þ6 about the mean or desired ferrite percent be viewed as a limit of ferrite control under ordinary circumstances, with  þ3 possible under ideal circumstances.

Carbon and Low-Alloy Heat Treatment Heat treatment is an important step in the production of steel castings because it develops the mechanical properties of a hardenable steel. Several types of heat treatment are available. The essential elements of any heat treatment are the heating cycle and the cooling cycle. Figure 35 shows schematically a heating cycle and three different cooling cycles. The length of time that a casting is held at temperature and the cooling rate are important factors. The holding time should be long enough to complete the desired microstructural transformation. Annealing is practiced on low-carbon steels to provide a soft, readily machinable structure. Annealed castings have relatively low strength but good ductility. In a full anneal, the castings are heated above the upper critical temperature, held there long enough to complete the transformation to austenite, and then furnace cooled at a controlled rate to obtain a stress-relieved casting with a pearlite-ferrite structure that is ductile and readily machinable. There are variations of the annealing heat treatment for specialized purposes, for example, to achieve a spheroidized pearlite structure. The annealing temperatures of such specialized treatments may differ substantially from those of a full anneal. Full annealing and spheroidizing heat treatments are costly and should not be specified unless maximum ductility is actually required. Cooling is accomplished by reducing or by simply turning off the heat input to the furnace. When the castings have cooled to below the lower critical temperature (approximately 425



C, or 800  F), the transformation of austenite is usually complete. The castings can then be removed from the furnace and air cooled or even quenched. Furnace cooling to lower temperatures merely wastes furnace time and requires additional heat to bring the furnace up to temperature for the next load. Normalizing consists of heating the steel to a suitable temperature above the upper critical transformation temperature, holding long enough to complete the transformation to austenite, removing the work from the furnace, and cooling it in still air. Some castings are tempered after normalizing. The castings must be placed so that the air can circulate freely around every casting. If air flow is restricted, the operation will be more like annealing. On the other hand, accelerated cooling by fans or forced-air flow may produce a result more like quenching. The microstructure that results from normalizing is a mixture of ferrite and pearlite, associated with only low residual stresses and almost no distortion. Tensile strengths up to 655 MPa (95 ksi) can be obtained in this way. The normalizing and tempering operation is used to meet a number of standard specifications in this strength range. Because of the uniform structure obtained upon normalizing, machinability is good. The cost of normalizing makes this heat treatment attractive. It requires less furnace time than annealing, and its cooling cycle is less expensive than quenching. Quenching hardens the steel by heating it above the transition temperature (austenitized), as in annealing or normalizing (Fig. 35). The work is cooled more rapidly than in the other heat treatments—fast enough so that pearlite and ferrite do not have time to form. Water and oil are the media most commonly used for quenching steel castings. Water is used whenever possible. Higher-carbon steels require oil quenching. Some complicated shapes also demand oil quenching to minimize quench cracking. Oil quenches a little more slowly than water at all temperatures. At lower temperatures, the cooling curves for oil taper off; this means even slower rates than for water. Therefore, martensite forms more slowly in oil than in water. Certain organic chemicals can be added to water to produce a quenching solution that resembles oil in its heat-removal characteristics. The main advantage of these solutions is that they have the behavior of oil without the fire hazard. Their greatest disadvantage is that they coat the work and thus change the composition of the bath. The quench severity of these baths varies widely with small changes in composition. Tight control is necessary and sometimes difficult. The quenching of carbon steels is always followed by tempering in order to adjust the mechanical properties of the quenched steel. The higher tensile strength levels of carbon steels can be obtained only by quenching and tempering. The quenching-and-tempering

Steel Castings Properties / 971 operation produces the optimum combination of strength and toughness. Tempering is a heat treatment that follows quenching and sometimes normalizing. One purpose of tempering is to reduce the residual stresses that develop during cooling and transformation. Another objective of tempering is to modify the metallurgical structure of martensite and thus adjust strength and other mechanical properties to specified levels. Tempering consists of heating the work to a temperature below the transformation range, holding for a specified time, and finally cooling. Carbon steels are tempered in the range of 175 to 705  C (350 to 1300  F). The holding time at temperature may vary from 30 min to several hours. A longer time at a given tempering temperature or a higher tempering temperature for a given time produces greater tempering. Martensite softens more than pearlite at a given tempering temperature. Composition also affects the tempering response; carbide-forming elements cause the steel to exhibit greater resistance to tempering. Tempering below 595  C (1100  F) may cause temper embrittlement in certain steels. Tempering is usually not done in this temper embrittlement range, and when higher tempering temperatures are used, the work can be quenched from the tempering temperature to minimize time in the embrittling zone during cooling. Stress Relief. Because tempering is a stressrelief treatment, there is no need for a special stress relief after tempering. Stress relief is obtained by heating to temperatures above 260  C (500  F). A stress-relief treatment is sometimes required when operations are performed after heat treatment that leave residual stresses in the casting. Welding, induction hardening, and grinding are examples of such operations. The maximum temperature for stress relief is generally limited to 30  C (50  F) below the tempering temperature that had been used in heat treating the casting to prevent it from softening. Hydrogen Removal. Hydrogen has been found to cause low elongation and reduction in area in steel. If the steel contains 4 or 5 ppm of hydrogen, the ductility will be approximately 20% of that of a hydrogen-free steel. Hydrogen in steel is a mobile element. Above room temperature, the hydrogen will diffuse from the steel, and ductility will be restored. Hydrogen removal can be accelerated by heating to 205 to 315  C (400 to 600  F). This heat treatment is commonly referred to as aging. The aging time is proportional to section thickness. Generally, 25 mm (1 in.) equals 20 h. For heavy sections (250 mm, or 10 in.), hydrogen removal by aging becomes impractical because of time requirements.

High-Alloy Heat Treatment The heat treatment of stainless steel castings is very similar in purpose and procedure to the thermal processing of comparable wrought

materials. However, the differences in detail warrant separate consideration here. Martensitic alloys CA15 and CA40 do not require subcritical annealing to remove the effects of cold working. However, in workhardenable ferritic alloys, machining and grinding stresses are relieved at temperatures from approximately 260 to 540  C (500 to 1000  F). Casting stresses in the martensitic alloys noted previously should be relieved by subcritical annealing prior to further heat treatment. When these hardened martensitic castings are stress relieved, the stress-relieving temperature must be kept below the final tempering or aging temperature. Alloy CA6NM (UNS J91540) possesses better casting behavior and improved weldability; it equals or exceeds all of the mechanical, corrosion, and cavitation resistance properties of CA15; and it has largely replaced the older alloy. Both CA6NM and CA15 castings are normally supplied in the normalized condition at 955  C (1750  F) minimum and tempered at 595  C (1100  F) minimum. However, when it is necessary or desirable to anneal CA6NM castings, a temperature of 790 to 815  C (1450 to 1500  F) should be used. The alloy should be furnace cooled or otherwise slow cooled to 595  C (1100  F), after which it can be air cooled. When stress relieving is required, CA6NM can be heated to 620  C (1150  F) maximum, followed by slow cooling to prevent martensite formation. Homogenization. Alloy segregation and dendritic structures may occur in castings and may be particularly pronounced in heavy sections. Because castings are not subjected to the high-temperature mechanical reduction and soaking treatments involved in the mill processing of wrought alloys, it is frequently necessary to homogenize some alloys at temperatures above 1095  C (2000  F) to promote uniformity of chemical composition and microstructure. The full annealing of martensitic castings results in recrystallization and maximum softness, but it is less effective than homogenization in eliminating segregation. Homogenization is a common procedure in the heat treatment of precipitationhardening castings. Ferritic and Austenitic Alloys. The ferritic, austenitic, and mixed ferritic-austenitic alloys are not hardenable by heat treatment. They can be heat treated to improve their corrosion resistance and machining characteristics. The ferritic alloys CB30 and CC50 are annealed to relieve stresses and to reduce hardness by heating above 790  C (1450  F). The austenitic alloys achieve maximum resistance to intergranular corrosion by solution annealing. As-cast structures, or castings exposed to temperatures from 425 to 870  C (800 to 1600  F), may contain complex chromium carbides precipitated preferentially along grain boundaries in wholly austenitic alloys. This microstructure is susceptible to intergranular corrosion, especially in oxidizing solutions. (In partially ferritic alloys, carbides tend to

precipitate in the discontinuous ferrite side; thus, these alloys are less susceptible to intergranular attack than wholly austenitic alloys.) The purpose of solution annealing is to ensure the complete solution of carbides in the matrix and to retain these carbides in solid solution. Solution-annealing procedures for all austenitic alloys are similar and consist of heating to a temperature of approximately 1095  C (2000  F), holding for a time sufficient to accomplish complete solution of carbides, and quenching at a rate fast enough to prevent reprecipitation of the carbides—particularly while cooling through the range from 870 to 540  C (1600 to 1000  F). The temperatures to which castings should be heated prior to quenching vary somewhat, depending on the alloy. A two-step heat treating procedure can be applied to the niobium-containing CF8C alloy. The first treatment consists of solution annealing. This is followed by a stabilizing treatment at 870 to 925  C (1600 to 1700  F), which precipitates niobium carbides, prevents formation of the damaging chromium carbides, and provides maximum resistance to intergranular attack. Because of their low carbon contents, CF3 and CF3M as-cast do not contain enough chromium carbides to cause selective intergranular attack; therefore, these alloys can be used in some environments in this condition. However, for maximum corrosion resistance, these grades require solution annealing. Martensitic Alloys. Alloy CA6NM should be hardened by air cooling or oil quenching from a temperature of 1010 to 1065  C (1850 to 1950  F). Even though the carbon content of this alloy is lower than that of CA15, this fact in itself and the addition of molybdenum and nickel enable the alloy to harden completely without significant austenite retention when cooled as suggested. The choice of cooling medium is primarily determined by the maximum section size. Section sizes exceeding 125 mm (5 in.) will harden completely when cooled in air. Generally, alloy CA6NM is not susceptible to cracking during cooling from elevated temperatures. For this reason, no problem should arise in the air cooling or oil quenching of configurations that include thick as well as thin sections. A wide selection of mechanical properties is available through the choice of tempering temperatures. Alloy CA6NM is normally supplied, normalized, and tempered at 595 to 620  C (1100 to 1150  F). Reaustenitizing occurs upon tempering above 620  C (1150  F); the amount of reaustenitization increases with temperature. Depending on the amount of this transformation, cooling from such tempering temperatures may adversely affect both ductility and toughness through the transformation to untempered martensite. Even though the alloy is characterized by a decrease in impact strength when tempered in the range of 370 to 595  C (700 to 1100  F), the minimum reached is significantly higher than that of CA15. This improvement in impact

972 / Steel Castings toughness results from the presence of molybdenum and nickel in the composition and from the lower carbon content. The best combination of strength with toughness is obtained when the alloy is tempered above 510  C (950  F). The minor loss of toughness and ductility that does occur is associated with the lesser degree of tempering that takes place at the lower temperature and not with embrittlement, as may be the situation with other 12% Cr steels that contain no molybdenum. The addition of molybdenum to 12% Cr steels makes them unusually stable thermally and normally not susceptible to embrittlement in the annealed or annealed-and-cold-worked conditions, even when exposed for long periods of time at 370 to 480  C (700 to 900  F). The hardening procedures for CA15 castings are similar to those used for the comparable wrought alloy (type 410). Austenitizing consists of heating to 955 to 1010  C (1750 to 1850  F) and soaking for at least 30 min; the high side of this temperature range is normally employed. Parts are then cooled in air or quenched in oil. To reduce the probability of cracking in the brittle, untempered martensitic condition, tempering should take place immediately after quenching. Tempering is performed in two temperature ranges: up to 370  C (700  F) for maximum strength and corrosion resistance, and from 595 to 760  C (1100 to 1400  F) for improved ductility at lower strength levels. Tempering in the range of 370 to 595  C (700 to 1100  F) is normally avoided because of the resultant low impact strength. In the hardened-and-tempered condition, CA40 provides higher tensile strength and lower ductility than CA15 tempered at the same temperature. Both alloys can be annealed by cooling slowly from the range of 845 to 900  C (1550 to 1650  F). Precipitation-Hardening Alloys. It is desirable to subject precipitation-hardenable castings to a high-temperature homogenization treatment to reduce alloy segregation and to obtain more uniform response to subsequent heat treatment. Even investment castings that are slowly cooled from the pouring temperature exhibit more nearly uniform properties when they have been homogenized.

Corrosion-Resistant Applications Martensitic grades include CA15, CA40, CA15M, and CA6NM. Alloy CA15 contains the minimum amount of chromium necessary to make it essentially rustproof. It has good resistance to atmospheric corrosion as well as to many organic media in relatively mild service. Alloy CA40 is a higher-carbon modification of CA15 that can be heat treated to higher strength and hardness levels. A molybdenum-containing modification of CA15, alloy CA15M, provides improved elevated-temperature strength properties and improved low-temperature properties. Alloy CA6NM is an Fe-Cr-

Ni-Mo alloy of low carbon content. The presence of nickel offsets the ferritizing effect of the low carbon content so that strength and hardness properties are comparable to those of CA15 and impact strength is substantially improved. The molybdenum addition improves the resistance of the alloy in seawater. A wide range of mechanical properties can be obtained in the martensitic alloy group. Tensile strengths from 620 to 1520 MPa (90 to 220 ksi) and hardnesses as high as 500 HB can be obtained through heat treatment. The alloys have fair to good weldability and machinability if proper techniques are employed; CA40 is considered the poorest and CA6NM is the best in this regard. The martensitic alloys are used in pumps, compressors, valves, hydraulic turbines, propellers, and machinery components. Austenitic grades include alloys CH20, CK20, and CN7M. The CH20 and CK20 alloys are high-chromium, high-carbon, wholly austenitic compositions in which the chromium exceeds the nickel content. They have better resistance to dilute sulfuric acid than CF8 and have improved strength at elevated temperatures. These alloys are used for specialized applications in the chemical-processing and pulp and paper industries for handling pulp solutions and nitric acid. The high-nickel CN7M grade containing molybdenum and copper is widely used for handling hot sulfuric acid. This alloy also offers resistance to dilute hydrochloric acid and hot chloride solutions. It is used in steel mills as containers for nitric-hydrofluoric pickling solutions and in many industries for severe-service applications for which the high-chromium CFtype alloys are inadequate. The 6-Mo grades are used to resist marine corrosion and are used in water desalination facilities. Ferritic grades are designated CB30 and CC50. Alloy CB30 is practically nonhardenable by heat treatment. As this alloy is normally made, the balance among the elements in the composition results in a wholly ferritic structure similar to that of wrought type 442 stainless steel. By balancing the composition toward the low end of the chromium and the high ends of the nickel and carbon ranges, however, some martensite can be formed through heat treatment, and the properties of the alloy approach those of the hardenable wrought type 431. Alloy CB30 castings have greater resistance to most corrosives than the CA grades and are used for valve bodies and trim in general chemical production and food processing. Because of its low impact strength, however, CB30 has been supplanted in many applications by the higher-nickel-containing austenitic grades of the CF type. The high-chromium CC50 alloy has good resistance to oxidizing corrosives, mixed nitric and sulfuric acids, and alkaline liquors. It is used for castings in contact with acid mine waters and in nitrocellulose production. For best impact strength, the alloy is made with more than 2% Ni and more than 0.15% N. Austenitic-ferritic grades include CE30, CF3, CF3A, CF8, CF8A, CF20, CF3M,

CF3MA, CF8M, CF8C, CF16F, and CG8M. These alloys usually contain 5 to 40% ferrite, depending on the particular grade and the balance among the ferrite-promoting and austenite-promoting elements in the chemical composition. This ferrite content improves the weldability of the alloys and increases their mechanical strength and resistance to SCC. The amount of ferrite in a corrosion-resistant casting can be estimated from its composition by using the Schoefer diagram (Fig. 34) or from its response to magnetic measuring instruments. Alloy CE30 is a high-carbon, high-chromium alloy that has good resistance to sulfurous acid and can be used in the as-cast condition. It has been extensively used in the pulp and paper industry for castings and welded assemblies that cannot be effectively heat treated. A controlled ferrite grade, designated CE30A, is used in the petroleum industry for its high strength and resistance to SCC in polythionic acid. The CF alloys as a group constitute the major segment of corrosion-resistant casting production. When properly heat treated, the alloys are resistant to a great variety of corrosives and are usually considered the best general-purpose types. They have good castability, machinability, and weldability and are tough and strong at temperatures down to 255  C (425  F). Alloy CF8, the cast equivalent of type 304 stainless steel, can be viewed as the base grade, and all the others as variants of this basic type. The CF8 alloy has excellent resistance to nitric acid and all strongly oxidizing conditions. The higher-carbon CF20 grade is satisfactorily used for less corrosive service than that requiring CF8, and the low-carbon type CF3 is specifically designed for use where castings are to be welded without subsequent heat treatment. The molybdenum-containing grades CF8M and CF3M have improved resistance to reducing chemicals and are used to handle dilute sulfuric and acetic acids, paper mill liquors, and a wide variety of industrial corrosives. Alloy CF8M has become the most frequently used grade for corrosion-resistant pumps and valves because of its versatility in meeting many corrosive service demands. Because CF3M has a low carbon content, it can be used without heat treatment after welding. The niobium-stabilized CF8C alloy is the cast equivalent of type 347. Castings of this alloy, therefore, are used to resist the same corrosives as CF8 but where field welding or service temperatures of 650  C (1200  F) are involved. Higher mechanical properties are specified for grades CF3A, CF8A, and CF3MA than for the CF3, CF8, and CF3M alloys, because the compositions are balanced to provide a controlled amount of ferrite that will ensure the required strength. These alloys are being used in nuclear power plant equipment. The CF16F grade has an addition of selenium to improve the machinability of castings that require extensive drilling, threading, and the like. It is used in service similar to that for which CF20 is used. Type CG8M has a higher

Steel Castings Properties / 973 molybdenum content than CF8M and is preferred to the latter in service where improved resistance to sulfuric and sulfurous acid solutions and to the pitting action of halogen compounds is needed. Unlike CF8M, however, it is not suitable for use in nitric acid or other strongly oxidizing environments. Duplex alloys have higher yield strength than the austenitics. This difference gives the duplex alloys an economic edge in, for example, the chemical process industry; higher process flow rates and operating pressures are possible without a major equipment modification. Cost savings can also be realized when the higher strength allows the downgaging of the wall thickness of piping, heat exchanger tubing, tanks, columns, and pressure vessels. For rotating equipment, such as centrifuges, the mass of equipment can be reduced by using a duplex stainless steel. Further savings in motors and gearing results because of smaller loads. For some time, cast duplex valves and pumps have used the strength of these materials either to allow higher pressures or to lower costs by using thinner walls. Precipitation-hardening alloys include CB7Cu and CD4MCu. Alloy CB7Cu is a lowcarbon martensitic alloy that may contain minor amounts of retained austenite or ferrite. The corrosion resistance of CB7Cu lies between that of the CA types and the nonhardenable CF alloys, so it is used when both high strength and improved corrosion resistance are required. Castings of CB7Cu are machined in the solution-treated condition and then through hardened by a low-temperature aging treatment (480 to 595  C, or 900 to 1000  F). Because of this capability, the CB7Cu grade has found wide application in highly stressed, machined castings in the aircraft and food-processing industries. Type CD4MCu is a two-phase alloy with an austenite-ferrite structure that, because of its high chromium and low carbon contents, does not develop martensite when heat treated. Like the CB7Cu grade, CD4MCu can be hardened by a low-temperature aging treatment, but it is normally used in the solution-annealed condition. In this condition, its strength is double that of the CF grades, and its corrosion resistance is optimized. This alloy has corrosion resistance equal to, or better than, the CF types and has excellent resistance to SCC in chloride-containing media. It is highly resistant to sulfuric and nitric acids and is used for pumps, valves, and stressed components in the marine, chemical, textile, and pulp and paper industries, for which a combination of superior corrosion resistance and high strength is essential.

Heat-Resistant Applications Iron-chromium alloys include grades HA, HC, and HD. Alloy HA has limited application because of its low strength and limited resistance to gaseous corrosion at high temperature.

It has been used in valves, flanges, and fittings where light stresses are encountered. Alloys HC and HD can be used for loadbearing applications up to 650  C (1200  F) and where only light loads are involved up to 1040  C (1900  F). These grades, however, become embrittled by s-phase in the 650 to 870  C (1200 to 1600  F) temperature range. They are similar to each other in corrosion resistance, with HD exhibiting higher strength. Both are used in ore-roasting furnaces for such parts as rabble arms and blades, for salt pots and grate bars, and in high-sulfur applications in which high strength is not required. Iron-chromium-nickel alloys include HE, HF, HH, HI, HK, and HL. They are predominantly or completely austenitic and exhibit greater strength and ductility than the ironchromium alloys. Alloy HE has excellent corrosion resistance at high temperatures coupled with moderate strength. This combination of corrosion resistance and strength makes HE suitable for service to 1095  C (2000  F). Grade HE can be used in high-sulfur applications and is often found in ore-roasting and steel mill furnaces. The alloy is prone to s-phase formation, however, at temperatures of 650 to 870  C (1200 to 1600  F). Alloy HF is essentially immune to s-phase formation and can be used at temperatures to 870  C (1600  F). It is used for tube supports and beams in oil refinery heaters and in cement kilns, ore-roasting ovens, and heat treating furnaces. Grade HH exhibits high strength and excellent resistance to oxidation at temperatures to 1095  C (2000  F). Its composition can be balanced to yield a partially ferritic or a completely austenitic structure and a wide range of properties. Because of this, the composition of the HH alloy should be tailored to the application. The partially ferritic alloy, type I, has a somewhat lower creep strength and a higher ductility at elevated temperature than the wholly austenitic alloy type II. Type I is also more prone to s-phase formation between 650 and 870  C (1200 and 1600  F). Type II is preferred for application in this temperature range. Both types are widely used for furnace parts of many kinds but are not recommended for severe-temperature cycling service, such as that experienced by quenching fixtures. Grade HI is similar to the fully austenitic alloy HH, but its higher chromium content confers sufficient scaling resistance for use up to 1180  C (2150  F). Its major application has been in cast retorts for calcium and magnesium production. Alloy HK has high creep and rupture strengths and can be used in structural applications to 1150  C (2100  F). Its resistance to hot gas corrosion is excellent. It is often used for furnace rolls and parts as well as for steam reformer and ethylene pyrolysis tubing. Grade HL has properties similar to those of HK and exhibits the best resistance to corrosion in high-sulfur environments to 980  C (1800  F) of the alloys in this group. It is typically used in gas dissociation equipment.

Iron-nickel-chromium alloys include grades HN, HP, HT, HU, HW, and HX. These materials employ nickel as a predominant alloying (or base) element and remain austenitic throughout their temperature range of application. They are generally suitable for use to 1150  C (2100  F) and resist thermal fatigue and shock induced by severe temperature cycling. However, the nickel-base grades are not considered suitable for high-sulfur environments. Alloy HN has properties similar to those of HK. It is employed in brazing fixtures, furnace rolls, and parts. Grade HP is extremely resistant to oxidizing and carburizing atmospheres. It has good strength in the temperature range of 900 to 1095  C (1650 to 2000  F) and is often specified for heat treat fixtures, radiant tubes, and coils for ethylene pyrolysis heaters. Alloy HT can withstand oxidizing conditions to 1150  C (2100  F) and reducing conditions to 1095  C (2000  F). It is widely used to heat treat furnace parts subject to cyclic heating, such as rails, rolls, disks, chains, boxes, pots, and fixtures. It has also found application for glass rolls, enameling racks, and radiant tubes. Alloy HU has excellent resistance to hot gas corrosion and thermal fatigue, and it has good high-temperature strength. It is often used for severe applications, such as burner tubes, lead and cyanide pots, retorts, and furnace rolls. Grades HW and HX are extremely resistant to oxidation, thermal shock, and fatigue. Their high electrical resistivities make them suitable for the production of cast electrical heating elements. Both are highly resistant to carburization when in contact with tempering and cyaniding salts. The higher alloy content of HX confers better gas corrosion resistance, particularly in reducing gases containing sulfur, where HW is not recommended. Both grades are typically used for hearths, mufflers, retorts, trays, burner parts, enameling fixtures, quenching fixtures, and containers for molten lead.

ACKNOWLEDGMENT Parts of this article were adapted from “Steel Castings,” revised by Malcolm Blair, Properties and Selection: Irons, Steels, and High-Performance Alloys, Vol 1, ASM Handbook; John M. Svoboda, “Plain Carbon Steels,” “LowAlloy Steels,” and “High-Alloy Steels,” all from Castings, Vol 15, ASM Handbook; and Steel Castings Handbook, 6th ed., edited by Malcolm Blair and Thomas Stevens. REFERENCES 1. M. Blair, Corrosion of Cast Stainless Steels, Corrosion; Materials, Vol 13B, ASM Handbook, ASM International, 2005, p 78–87 2. J.D. Redmond, Selecting Second-Generation Duplex Stainless Steels, Chem. Eng., Oct 27 and Nov 24, 1986

974 / Steel Castings 3. Steel Castings Handbook, 6th ed., Steel Founders’ Society of America and ASM International, 1995 4. Steel Castings Handbook Supplement 2— Summary of Standard Specifications for Steel Castings, Steel Founders’ Society of America, 1999

5. Steel Castings Handbook Supplement 8—High Alloy Data Sheets—Corrosion Series, Steel Founders’ Society of America, 2004 6. Steel Castings Handbook Supplement 9— High Alloy Data Sheets—Heat Series, Steel Founders’ Society of America, 2004

7. R.I. Stephens, “Fatigue and Fracture Toughness of Five Carbon or Low Alloy Cast Steels at Room or Low Climatic Temperatures,” Research Report 94 (A and B), Steel Founders’ Society of America, 1982

ASM Handbook, Volume 15: Casting ASM Handbook Committee, p 975-986 DOI: 10.1361/asmhba0005330

Copyright © 2008 ASM International® All rights reserved. www.asminternational.org

Selection and Evaluation of Steel Castings David Poweleit, Steel Founders Society of America

STEEL CASTINGS are used in safety-critical applications and in harsh, demanding environments to carry significant loads. They are commonly welded into a fabricated structure. Castings give users unlimited potential for steel geometry. Steel castings are used broadly in industrial equipment for end-use applications such as mining and construction. The railroad industry uses steel castings extensively for couplers, trucks, wheels, and corners. Fifth wheels for semitrucks are steel castings. The complex internal passages of pumps and valves are created through steel casting. Steel castings are used in energy production and are relied on for high-performance military applications. Steel castings are used in high-pressure service at nuclear power plants. Finally, steel castings serve many other ends where steel and geometry are important. New markets, such as building construction, are supported every day. The reason steel castings are attractive is their freedom of design. Any shape that can be imagined can be cast. Frequently, a casting cannot be made effectively, not because the design was too aggressive but because it was too timid. Often, a fabrication design that is inadequate is sent to the foundry to see if it can be made as a casting. This normally causes manufacturing problems, and changing the manufacturing process will not overcome inherent performance characteristics of a design. Cast parts are best manufactured when designed for the casting process. One reason that castings are seen as problems is because the foundry is often asked to make poor designs successful, and so the lead time is long and the cost is high. Good casting design allows weight to be reduced, cost to be lowered, and performance to be improved. Steel casting producers routinely test each heat of steel to make sure it meets the mechanical properties required in the material specification. The heat is also analyzed chemically to certify that it meets the standard. Other specialized tests can be required, such as low-temperature impact testing, when service performance requirements dictate. Steel castings are expensive sources of scrap steel if false economy is applied at the design stage. Good applications of steel castings are components with complex details that would

have many parts with high fabrication costs, poor material utilization, and reduced performance if assembled from multiple small wrought parts. Casting size and shape have limits based on section size or geometric limits. The flexibility of casting allows material to be placed where needed and material to be removed where it is not needed. Castings permit big, sweeping curves, nonuniform sections, and complex geometry that serve both functional and aesthetic needs. The properties of cast steel depend on the composition and heat treatment. Because designers often use yield strength as a basic property in design, material is often ordered to higher strength without considering the advantage in castings of using a lowerstrength material with optimum ductility and weldability. Since the load-carrying cross section can be increased to accommodate lower-strength material, the casting can be supplied in the highest ductility with strength levels that are compatible with the rolled structural shapes. This use of cast steel in its optimal condition ensures that the casting will perform safely and reliably and that excessive loads will cause failure to occur first in the rolled section familiar to the designer.

Design Cast steel is used in monstrous mining trucks as part of the chassis, because of durability and resistance to impact loads. Steel castings in truck frames have been leveraged to minimize stress concentrations, handle multidirectional loading, and relocate welds to lower-stressed regions. Steel is commonly thought of as a high-strength but high-weight material. While the density of steel is high, steel can offer equal strength-to-weight ratios compared to either titanium or aluminum, and at a fraction of the cost of titanium or with a superior modulus of elasticity compared to aluminum. Cast steel also offers wear resistance, corrosion resistance, heat resistance, and weldability. Materials such as maraging steels and modified martensitic precipitation-hardening steels can achieve these properties, and additional steel casting alloy development is always looking for means to improve mechanical performance. By taking advantage of creative geometry via casting,

the weight of a cast steel part can be further reduced by removing material where it is not needed on the part and using transitioning section thickness to have them sized for the load carried at each point. Details are found in the articles “Riser Design,” “Gating Design,” and “Shape Casting of Steel” in this Volume. Design Impact. Casting offers freedom of geometry, so casting part design can play a key role in the mechanical performance and function of the component. It is wise to consider the function and loads on the part independent of pre-existing part designs to take advantage of this design freedom. Sections of a cast part subject to higher stress can be enlarged, while lowstress regions can be reduced. This flexibility facilitates a part with optimum performance and reduces weight; both minimize cost. One way to ensure the best casting design is to work with a foundry on design. A foundry will have the technical expertise to partner on the casting, design, and materials selection. General Design Rules. There are several rules of thumb to the development of a good steel casting. First, reduce the number of isolated heavy sections and have smooth-flowing transitions. Smooth transitions between features promote laminar flow of the liquid metal into the casting and also improve the mechanical performance of the part by reducing stress raisers. An additional advantage of these smooth transitions is that designs for castability and for finite-element stress analysis often converge on the same optimal design. It is feasible to cast any geometry, but this may increase cost. Junctions within a casting should be designed to not add mass. Changing section thickness in a casting should be through smooth, easy transitions; adding taper and large radii helps. Reducing undercuts and internal geometry helps to minimize the cost of the pattern or die. The foundry and customer should agree on tolerances. Specifying as-cast tolerances is also important in minimizing cost. Postprocessing, such as machining, surface treatment, and how the part will be held in a fixture, also influences the final cost of the part. Datum points should be stated, and machine stock should be added to required locations. It is important to identify all datum targets in the initial design so that gates, risers, and other finishing operations do not affect

976 / Steel Castings the integrity of these. Draft is the amount of taper or the angle that must be allowed on all vertical faces of a casting tool to permit its removal from the mold without tearing the mold walls. Draft should be added to the design dimensions, but metal thickness must be maintained. The amount of draft recommended under normal conditions is approximately 1.5
. All of these rules can be reviewed with a foundry engineer. Good applications use the performance of steel with the flexible geometry of a casting. Steel castings offer high mechanical properties over a wide range of operating temperatures. Cast steel offers the mechanical properties of wrought steel and can be welded. Material Effects. The chemical composition and microstructure of a steel casting determine its mechanical properties. Heat treatment can change microstructure and provide a wide range of mechanical properties. The response to heat treatment for a given section is hardenability. A steel with a high hardenability will have uniform hardness in thicker sections than ones with low hardenability. In general, adding alloying elements increases cost and improves some properties but may reduce others. Most elements will increase the hardenability of steel. The effects of common alloying elements on steel properties are given in Table 1. When selecting a steel alloy, it is important to first know the required properties of the cast component. Carbon should be kept as low as possible to maximize weldability. Minimizing alloying elements to safely meet the performance requirements of the item will reduce cost. The foundry can provide assistance with materials selection to ensure the appropriate properties are considered when specifying the material, and they will know the availability and cost factors of the purchase. Design Considerations. Design requirements are typically determined in terms of strength or maximum stress. The design is commonly constrained by elastic modulus, fatigue, toughness, or ductility. Increasing the strength of steel normally reduces the ductility, toughness, and weldability. It is often more desirable in steel casting design to use a lower-strength grade and increase the section size or modify the Table 1

shape. However, if weight or size reduction is a factor to reduce operating cost over the life of a part, the initial higher-cost material may be attractive. The design freedom makes castings an attractive way to obtain the best fabrication and material performance and the needed component stiffness and strength. When designing a part, it is important to understand the design limits so that proper materials selection can be made. Stress, strain, fatigue, impact, wear, creep, and corrosion are common service conditions that can impose design limits. Stress results from a mechanical load carried by a component. The strength of a steel is the measure of its load-carrying ability. Loads can be applied and removed without deformation if they are small enough (elasticity). When a large enough load is applied, the material will stretch or compress and deform permanently (plasticity). Plastic deformation starts to occur when the yield strength is exceeded. The maximum load that can be applied before the material fractures is the ultimate tensile strength. Designers need to ensure that the part will not break or permanently deform. Thus, it is important to design a casting for stress levels below the yield strength or in the elastic region of the part. Typically ½ the yield strength is used for safety but 2/3 can be used with a thorough evaluation. Room-temperature properties are usually shown in tables, so the temperature dependence of the strength of steel must be considered for the intended design environment. The strength of steel depends on the composition and heat treatment. Steel is an iron-base alloy, mainly alloyed with carbon. Other alloying elements (Table 1) add strength but are also important in determining how effectively the steel grade will respond to heat treatment. Heat treatment rearranges the crystal structure of the iron and the distribution of carbon. Slow cooling rates produce coarse microstructures, which have lower strength. Cooling slowly in the furnace is called annealing and is not commonly used, except as an intermediate treatment to allow some grades to be machined. Cooling in still air is called normalizing and is the most common treatment to provide good strength and ductility. Rapidly cooling in water or oil is known as quenching. Steel must be reheated

Effects of alloying elements

Element

Carbon (C) Manganese (Mn) Silicon (Si) Nickel (Ni) Chromium (Cr) Molybdenum (Mo) Vanadium (V) Tungsten (W) Aluminum (Al) Titanium (Ti) Zirconium (Zi) Oxygen (O) Nitrogen (N) Hydrogen (H) Phosphorus (P) Sulfur (S)

Effect on Steel Properties

Increases strength but decreases toughness and weldability (most common and important) Similar, although lesser, effect as carbon Similar to carbon but with a lesser effect than manganese (important for castability) Improves toughness Improves oxidation and other corrosion resistance Improves hardenability and high-temperature strength Improves high-temperature strength Improves high-temperature strength Reduces the oxygen or nitrogen in the molten steel Reduces the oxygen or nitrogen in the molten steel Reduces the oxygen or nitrogen in the molten steel Negative effect by forming gas porosity Negative effect by forming gas porosity. Increases resistance to localized corrosion in stainless steels In high quantities, results in poor ductility Can increase strength but drastically reduces toughness and ductility Reduces toughness and ductility

or tempered after quenching to improve ductility. Quenching and tempering give the highest strength available from any grade. Varying the tempering temperature and time allows the production of a wide range of final strength levels with a quenched-and-tempered grade. Strain is the amount of elongation or compression in a loaded component. The increase in length per unit length is the elongation, and the change in area at the point of fracture of the bar is the reduction of area. The ability to absorb strain without fracture is critical to safety and reliability. Steel can be bent, twisted, compressed, or stretched without breaking. The amount of strain in a component will also determine its functionality. The elastic limit is the amount of stress that the material will withstand without permanent deformation. The yield strength is the amount of stress required to produce a small specified plastic deformation (often 0.2% in the United States). The ability to stretch or deform without cracking is called ductility. Brittleness is a characteristic of materials having limited plastic deformation; the yield strength and ultimate tensile strength are close. Ductility and brittleness are temperature-dependent. Low-alloy steels undergo a transition in character from ductile to brittle at low temperature. The nil ductility transition temperature is a measure of where brittle fracture occurs with no observable reduction of area. These properties and characteristics such as resilience and toughness are means of expressing material behavior to predict and judge performance. See the article “Steel Castings Properties” in this Volume. In general, increasing the strength of a steel grade reduces its ductility. While most designers think in terms of the material strength, most of the production of steel is in the lower-strength grades, which have good ductility. Heat treatment and carbon content influence strength and ductility, as shown in Fig. 1 and 2. Carbon contents are typically kept well below 0.30% to avoid problems with cracks in heat treating or welding. The ratio of stress to strain is the elastic modulus. Data are derived from tensile testing. The modulus of elasticity is based solely on the material composition; heat treatment does not affect modulus. The modulus of elasticity depends on the bonding energy between the atoms. Steels have a modulus of approximately 200 GPa (30  106 psi). Steels have the highest modulus of elasticity of commonly used materials. The larger the modulus, the smaller the deflection of a part; steel provides good stiffness. Fatigue is the failure of a component when it is repeatedly loaded, even at levels well below the yield strength of the steel. It is measured by repeated loading and unloading of several test bars at different stress levels and determining the number of cycles to failure. A typical result of stress versus cycles is shown in Fig. 3. Low-cycle fatigue is below 100,000 loadings, where ductility is needed. High-cycle fatigue is normally above 1 million loadings, and high strength is required. Smooth and notched samples are tested. The notched

Selection and Evaluation of Steel Castings / 977

Fig. 1

Room-temperature properties of cast low-alloy steels. QT, quenched and tempered; NT, normalized and tempered. Source: Ref 1

samples simulate surface irregularities in applications, both designed and accidental. Notched samples will fatigue at lower stress levels than unnotched, but cast samples had slightly higher notched fatigue strength than wrought alloys. Impact toughness measures the ability of steel to resist fracture or cracking during service. The ability of steel to resist cracking at low temperatures or during impact loading is known as fracture toughness. Toughness is measured by the amount of energy required to break the material at a certain temperature. A common test that measures toughness is the Charpy impact test. Like ductility, toughness tends to fall as the strength of the material increases. The reduction in toughness at higher carbon contents for two heat treatments is shown in Fig. 4. Wear occurs when materials rub against each other under load, and material is lost from the contacted surfaces. Gouging, abrasion, impact, cavitation, erosion, and corrosion must be considered in different types of applications. Typically, harder materials resist wear better. Since hardness increases with strength, higher-carbon, higher-strength steels are commonly used. It is important to retain adequate toughness at the high hardness levels to avoid cracking and premature failures. Creep occurs at elevated temperatures when the material permanently deforms at loads

Fig. 2

Stress-strain curves for selected steels showing influence of carbon content. Source: Ref 2

below the yield strength. Resisting creep requires complex alloys with relatively high carbon content. Generally, as the temperature or load increases, higher alloy contents are required for adequate performance. Creep rate is the rate of strain at a particular load, temperature, and time. The rate is not constant. Stress rupture is the time to failure under a given load at a particular temperature. Selection of a material for creep service must also take into account the oxidation or other high-temperature corrosion mechanisms that may limit the service life of the component. Corrosion is a chemical or electrochemical reaction between a material and its environment that produces a deterioration in the material or its function. It can be a general loss of material or a localized condition such as pitting, crevice corrosion, or selective attack. There are many standard laboratory tests for corrosion, but these rarely replicate service conditions or the nature of the environment. Materials evaluation should include a combination of laboratory testing, simulated service, and in-service monitoring. Corrosive conditions such as temperature, pH, oxygen concentration, chlorine and other chemicals, biological agents, and other variables

must be considered when selecting a material for service. Alloy Selection. The structure of wrought sections of steel is elongated in the direction of rolling, drawing, or forging. The strength and ductility are improved in the direction of elongation but reduced perpendicular to that direction (Fig. 5). The lack of directionality in steel castings gives them uniform properties in all directions. The cold forming of steel can also strengthen the steel but reduces ductility and toughness. Cast steel grades achieve the same trade-off of strength, ductility, and toughness by alloying and heat treatment. Therefore, casting grades with mechanical properties similar to wrought grades are designated differently (e.g., ASTM A 216 grade WCB is the cast counterpart to wrought 1020). Both ASTM A 915 (Table 2) and A 958 use grade names similar to their wrought counterparts; the grades may be heat treated to a variety of strength, ductility, and toughness levels. Selecting a steel alloy begins with understanding the design limits. Knowing the major mode of failure is key. Materials selection is part of the process of meeting the design

978 / Steel Castings

Fig. 5

Relation between mechanical properties of rolled steel slab and the inclination of the test speciment. Source: Ref 3

Fig. 3 Ref 3

Fatigue characteristic (S-N) curves for cast and wrought 8600-series steels quenched and tempered to the same hardness. Notched (Kt = 2.2) and unnotched specimens tested with R.R. Moore rotating-beam tests. Source:

Fig. 4

Room-temperature Charpy V-notch impact test values versus carbon content of cast steels in normalized-and-tempered condition. Tempering temperature: 650
C (1200
F)

criteria. Alloying elements can be added to improve performance; thus, starting with a carbon steel and building from there is the best practice. A material selection guide for five major design applications is shown in Fig. 6. This chart is meant to provide some initial guidance with reference to ASTM International standards designations. It is important to consult with a foundry to select the right material for each application. This is especially true of

higher-alloyed materials (outer rings in Fig. 6). Other alloys may be better, and the alloy and heat treatment can be tailored for specific conditions. The segments of the chart are discussed as follows. Structures (segment of the Fig. 6 guide) covers general applications governed by strength, deflection, and fatigue. Strength and ductility are still the main properties used to designate available steel grades. Increasing the strength of steel is easily achieved through more severe heat treatments or increases in the alloy content. The addition of alloying elements not only increases the strength that an alloy can achieve but also the section size of the part that can be effectively heat treated. The grades in Table 2 are common alloys of steel available for casting ASTM A 915 or A 958. The first column is the alloy designation, the second is the minimum requirements for the lowest-strength grade commonly available from that alloy, the third column is the minimum requirements for the highest-strength grade, and the last column is the calculated largest section that can be effectively heat treated through the section. It is apparent that higher-strength alloys have lower ductility but can be heat treated more effectively in larger sections. Higher-strength alloys also require more extensive weld procedures and may crack

in heat treatment, especially at high carbon contents. Fatigue resistance depends on the strength and ductility of the material, service conditions, corrosion, wear, residual stresses in the component, design (particularly in high-stress areas), and surface condition. The duty cycle, required service life, and analysis of the effects of failure of the component will determine the importance of fatigue resistance. Fatigue analysis is difficult, and component testing is not unusual to verify the design and part durability. Analytical tools such as computer modeling of service loads, along with the development of useful materials properties databases, reduce the number of design iterations. Properties such as strain-controlled cyclical properties, crack growth rates, integration of inspection standards, and life prediction improve designs by reducing the traditional requirements for factor of safety and allow more aggressive use of material. Castings allow the geometry to be tailored to the service requirements. Steels for structural applications can be found in ASTM A 27, A 148, A 747, A 915, and A 958. Pressure-containing applications have similar characteristics to structural applications. Specifications for pressure-containing applications have been developed to meet ASME Boiler and Pressure Vessel code. These steels are specified in ASTM standards and the related ASME code (SA): A 217/A 217M (SA-217/SA217M) for high-temperature service; A 487/A 487M (SA-487) and A 352/A 352M (SA-352) for low-temperature service; A 389 specially heat treated for high-temperature service; and

Selection and Evaluation of Steel Castings / 979

Fig. 6

Cast steel selection guide referencing ASTM International standards. Circle is segmented by primary application feature. Rings of greater diameter are more highly alloyed. Consult the standard for specific grades and classes of materials.

Table 2

Common cast steel properties

Grades similar to wrought grades Minimum low strength(a) ASTM A 915 and A 958 grade

SC1020 SC1030 SC1040 SC8620 SC8630 SC4130 SC4140 SC4330 SC4340

Tensile—yield

Minimum high strength

Median composition ideal critical diameter(b)

Tensile—yield

MPa

ksi

Elongation, %

MPa

ksi

Elongation, %

mm

in.

450–240 450–240 485–250 550–345 550–345 550–345 620–415 620–415 620–415

65–35 65–35 70–36 80–50 80–50 80–50 90–60 90–60 90–60

24 24 22 22 22 22 20 20 20

485–250 550–345 620–415 795–665 1035–930 1035–930 1105–1035 1450–1240 1450–1240

70–36 80–50 90–60 115–95 150–135 150–135 165–150 210–180 210–180

22 22 20 14 7 7 5 4 4

10 18 23 51 89 76 140 152 203

0.4 0.7 0.9 2.0 3.5 3.0 5.5 6.0 8.0

(a) The minimum tensile requirement of all grades by ASTM A 958 is 450–240 MPa (65–35 ksi). Values given are likely minimums. (b) The section size is not specified in either standard.

A 757/A 757M ferritic and martensitic for lowtemperature service. Similarly, there are European standards, such as BS EN 10213:2007 steel casting, for pressure purposes. Impact resistance, or toughness, is required when a part performs a safety-critical function, is subject to low temperatures, or is impact loaded in service. Impact toughness can be improved

through careful control of composition and heat treatment. Adding nickel is the common way to improve toughness. The toughness of all grades can be improved by lowering carbon, sulfur, and phosphorus and by using a quench-and-temper heat treatment. When toughness is needed, it should be verified by test. The steels specified for low-temperature service (ASTM A 352 and

A 757) require Charpy V-notch impact testing. Other grades can be tested as agreed upon by supplier and purchaser as a supplemental requirement as part of the purchase order or quality-assurance program. Always check specifications. In the case of corrosion-resistant cast steels for general applications, ASTM A 743/A 743M does not specify an impact strength, while ISO 11972 has ISO V-notch impact strength among the mechanical requirements, although verification is only necessary if required by the purchaser. High-temperature resistance and creep strength are required to carry loads at elevated temperatures. As the temperature increases, the required alloy content also increases. Commonly, chromium and molybdenum are added to the steel to improve elevated-temperature properties. Higher carbon content also helps. The preferred heat treatment of carbon and low-alloy steels is to normalize and temper. Weldable carbon steels for high-temperature service are found in ASTM A 216, and martensitic stainless and alloy steels for pressure and high-temperature service are in A 217. When the service temperatures exceed 650
C (1200
F), the alloyed steels are no longer adequate, and the cast heat-resistant grades containing high levels of chromium and nickel are used. These alloys are in ASTM A 297 and A 351. Wear (segment of Fig. 6) applies to mechanical wear and chemical corrosion. Mechanical wear and corrosion often exacerbate each other. The mechanism is called mechanically assisted corrosion or, in the case of fluids, erosioncorrosion. Severe wear or corrosion environments require high-alloy steels. Wear resistance is usually improved through the use of highhardness materials. Strength and hardness are related, so high-strength materials are commonly used when wear is a problem. Increasing carbon content also increases wear resistance. Special materials such as austenitic manganese alloys or high-chromium irons are used to give better wear resistance. Toughness must be adequate to avoid premature catastrophic failure. High-chromium irons offer good wear resistance in abrasion or even in corrosive environments. When impact loading is a part of the wear environment, austenitic manganese alloys work harden, allowing them to resist wear while maintaining high toughness. Wear materials are found in ASTM A 128, A 351, A 532, A 743, A 744, A 890, and A 494. Carbon or alloy steels may be used in less severe environments. Severe environments, such as saltwater and chemical processing, require high-alloy stainless steels or nickel-base alloys. Generally, higher chromium and molybdenum are needed as the environment becomes more severe. The proper selection of corrosion-resistant materials depends on the thorough knowledge of the end-use environment, and factors are often industry-specific; thus, a metallurgist or corrosion expert should be consulted. For more about corrosion and wear, see the Handbooks listed in the Selected References in this article.

980 / Steel Castings With cast steel alloys, one can increase various performance characteristics, such as corrosion resistance and wear resistance, through alloying and heat treatment. Mechanical properties such as strength and elongation can be adjusted. There are three keys to selecting the right steel casting alloy for optimized performance and cost:  Design the geometry of the steel casting to

carry the load uniformly.

 Start with carbon steel for most applications;

modify the heat treatment, and then add alloying elements to attain the required properties.  Know the design limits, normal and extreme conditions, design life, tolerances, and quality requirements for an application, and work with a foundry to design the part, the process, and select a material to exceed the requirements on a statistically acceptable basis that will assure productivity. Drawing Notes. Material should be identified as conforming to a specific specification, with the grade and/or class stipulated. Nondestructive examination, verification testing, inspection method, and criteria for acceptance should be called out in accordance with a national or industry standard or, in some cases, an internal specification. Porosity can be quantified as a maximum size per discontinuity, minimum separation between discontinuities, maximum number of discontinuities per area, and minimum size to count as a discontinuity; these are stipulated in the acceptance. Acceptance can also be based on some form of comparison to a standard. In-process welding is common to remove discontinuities. Generally, the weld would meet the same requirements as the cast material. Type and extent of reheat treatment after welding should be stated. Allowable draft, marking requirements, surface-quality referencing a comparator plate, allowance for projections where the gating system was cut off, incorporation of fillets and radii, and a comment to remove all burrs are common notes for steel casting drawings. Zoning a casting to have more stringent levels of examination in critical areas is optimal for minimizing cost while ensuring performance. Zoning can also be used in limiting weld repair or draft on the part. Most purchase specifications state that requirements for material and testing shall be done based on agreement of the purchaser and supplier. Alloy Designation. Cast steels do not follow the conventional AISI grade designation for wrought steels. Steel castings may have different alloy composition and heat treatment compared to a wrought counterpart. This is to provide similar mechanical properties and manufacturing characteristics while balancing the manufacturability of the steel as a casting in a mold. For instance, cast corrosion-resistant grades have 19% Cr and 9% Ni versus their wrought 300-series counterparts that have 18% Cr and 8% Ni. The wrought microstructure for 300-series stainless is fully austenitic, which makes the material nonmagnetic. The cast

microstructure has ferrite in the austenite; thus, it is actually magnetic. The ferrite provides strength, improves weldability, and maximizes corrosion resistance. The cast and wrought materials have similar resistance to a corrosive medium and can be used with one another. Several ASTM specifications, such as A 958, specify cast grades similar to wrought grades as SCxxxx, where xxxx is the wrought grade, so that designers can quickly reference a cast counterpart to the wrought steel. High-alloy steel grades are based on the Alloy Casting Institute system and the composition of the material. Grades start with either a “C” or an “H” for corrosion-resistant and heat-resistant grades, respectively. The second letter represents the nominal chromium-nickel content (Fig. 7), with later letters representing a higher percentage of nickel in the alloy. After the two letters, a number is provided to represent the carbon content. In corrosion-resistant alloys, the number is the maximum 1/100% carbon. In heat-resistant alloys, the number is the 0.1% midpoint of the range for carbon. Additional letters are then used to represent other major alloying constituents. An example of a corrosion-resistant grade would be CF8M. Breaking this down, the “C” would mean corrosion resistant, the “F” a nominal 19% Cr and 9% Ni, the “8” a maximum of 0.08% C, and the “M” that molybdenum is alloyed into this material. An example of a heat-resistant grade would be HK40. Breaking this down, the “H” would mean heat resistant, the “K” a nominal 26% Cr and 10% Ni, and the “40” a carbon ranging from 0.35 to 0.45%. Note that HY80 and HY100 from MIL-S-23008 are actually high-yield steel castings versus heatresistant grades.

Manufacturing Steel castings are made through the same casting processes used for other materials. However, due to the higher liquidus temperature of steel compared to some nonferrous melts and the mode of solidification, there are several attributes that are specific to the manufacturing of steel castings. Minimum Section Thickness. The rigidity of a section often governs the minimum thickness to which a section can be designed. There are cases, however, when a very thin section will suffice mechanically, and castability becomes the governing factor. In these cases, it is necessary that a limit of minimum section thickness be adopted in order for the liquid steel to completely fill the mold cavity. Liquid steel cools rapidly as it enters a mold. A minimum thickness of 6 mm (0.25 in.) is suggested for design use when conventional steel casting techniques are employed. Wall thickness of 1.5 mm (0.060 in.) and sections tapering down to 0.76 mm (0.030 in.) are common for investment castings. Draft is the amount of taper or the angle that must be allowed on all vertical faces of a pattern

Fig. 7

Chromium and nickel Institute standard corrosion-resistant cast steels. second letter in the designation.

content in Alloy Casting grades of heat-and These letters are the See text for details.

to permit its removal from the mold without tearing the mold walls or causing the tool to wear. Draft should be added to the design dimensions, but metal thickness must be maintained. Regardless of the type of pattern equipment used, draft must be considered in all casting designs. Draft can be eliminated by the use of cores; however, this adds significant costs. In cases where the amount of draft may affect the subsequent use of the casting, the drawing should specify whether this draft is to be added to or subtracted from the casting dimensions as given. The necessary amount of draft depends on the size of the casting, the method of production, and whether molding is by hand or machine. Machine molding will require a minimum amount of draft. Interior surfaces in green sand molding usually require more draft than exterior surfaces. The amount of draft recommended under normal conditions can range from 1 to 2
for sand casting but can be near 0
for investment casting. The draft allowance would normally be added to design dimensions. Mold features such as the parting line and cores are covered in full detail in other articles in this volume. Pouring. Cast steel is typically poured into the mold at a temperature of slightly less than 1650
C (3000
F). This higher pouring temperature limits the mold media that liquid steel can be poured into. It also means that steel needs to be melted with either an electric arc furnace or an induction furnace. An argon-oxygen decarburization vessel can be used when melting steel to lower sulfur content or increase the recovery of chromium; thus, it is sometimes used in conjunction with heavy-walled castings from carbon and low-alloy steels or with highalloy castings. The injection of reactive metals (calcium, magnesium, or rare earths) into the liquid steel is another means of lowering sulfur content. Carbon and low-alloy steels are deoxidized by adding aluminum, titanium, or zirconium. Of these, aluminum is used more frequently because of its effectiveness and low cost. Given that liquid steel has a greater affinity to react with the oxygen in the air, it is important to prevent turbulent flow or entrainment in the stream during pouring. The high temperature of liquid steel also makes venting of the

Selection and Evaluation of Steel Castings / 981 mold valuable, so that gases can escape as the liquid metal fills the cavity. Inclusion Shape Control. Nonmetallic inclusions that form during solidification depend on the oxygen and sulfur contents of the casting. Deoxidation decreases the amount of nonmetallic inclusions by eliminating the oxides, and it affects the shape of the sulfide inclusions. High levels of oxygen (>0.012% in the metal) form FeO, which decreases the solubility of manganese sulfides and causes the manganese sulfides to freeze early in solidification as globules. The use of silicon deoxidation alone normally causes the formation of globular (type I) sulfides. Decreasing the oxygen to between 0.008 and 0.012% in the metal, through the use of aluminum, titanium, or zirconium, increases the solubility of the manganese sulfides so that they solidify last as grain-boundary films. These grain-boundary sulfide films (type II inclusions) are detrimental to the ductility and toughness of the steel and cause increased susceptibility to hot tearing. If the oxygen content is very low ( 1.5) is added, then galaxies of sulfides (type IV inclusions) form; these are only for rare earth additions. The most common inclusion modifier is calcium. Directional Solidification. Steel castings begin to solidify at the mold wall, forming a continuously thickening envelope as heat is dissipated through the mold-metal interface. The volumetric contraction that occurs within a cross section of a solidifying cast member must be compensated by liquid feed metal from an adjoining heavier section or from a riser that serves as a feed metal reservoir and is placed adjacent to, or on top of, the heavier section. The lack of sufficient feed metal to compensate for volumetric contraction at the time of solidification is the cause of shrinkage cavities. They are found in sections that, due to design, must be fed through thinner sections. The thinner sections solidify too quickly to permit liquid feed metal to pass from the riser to the thicker sections. Note that even though voids may lack material, they may not prevent the casting from achieving the final performance characteristics. Developing the means to get the liquid metal into a casting and provide enough feed metal is generally the responsibility of the foundry engineer. The size and location of these features can be analyzed through solidification and fill simulation software. To encourage directional solidification, castings should have well-positioned gates, gradually changing cross sections, and smooth-flowing geometry.

Steel Casting Processes. Steel castings are largely manufactured as sand or investment castings but can also be done with centrifugal casting and other specialized processes. Sand castings can be made via green sand or chemically bonded sand, which would include the shell casting process. Sand-mold-making processes can be automated to make a higher-production volume of parts. Sand typically produces a rougher surface (200 to 300 rms) and requires more open tolerances (þ þ1.5 mm, or 0.030 and 0.75 and  0.060 in., when over 500 mm, or 20 in.). Through the pit molding process, sand molding is capable of producing parts that would easily fit in one’s hand to parts larger than a car. Investment casting typically produces a smoother surface (10 to 80 rms), has fairly tight tolerances (þ þ0.075 0.025 mm for the first 25 mm and  mm for each additional 25 mm, or 0.0010 in. for the first 1 in. and 0.003 in. for each additional inch), and can produce parts over 50 L (2 ft3) in volume. Tooling for sand casting patterns was traditionally made from wood. This has largely been replaced by a composite material. Tooling for investment casting is a metal die for making the investment waxes. Both sand and investment casting processes for steel can take full advantage of the rapid tooling technologies that are being developed and used. Rapid prototyping techniques, such as stereolithography, selective laser sintering, laminated object manufacturing, and three-dimensional printing, can be leveraged to make short-run or prototype steel castings. Heat Treatment. Some components can be used as-cast, such as those made out of lowcarbon stainless steels, but the majority of steel castings are heat treated. Heat treating is employed to at least stress relieve and normalize the casting, since the metal is often slowly cooled within the mold. Thin sections of investment castings will experience rapid cooling. Differential cooling rates within a component will impart varying mechanical properties that may need to be addressed by stress relieving and normalizing. Heat treatment can modify the mechanical properties of the part, but this usually means a trade-off. For instance, strength and hardness can be improved, but ductility, toughness, and weldability will be degraded. Cast steel responds very well to heat treatment. Weldability. Most steels are readily weldable, and it is routine practice to weld castings in production. Therefore, it is common in many specifications to set weld practices between the customer and the supplier. Carbon and Low-Alloy Steels. The weldability of carbon and low-alloy steels is primarily a function of composition and heat treatment. Carbon steels having low manganese and silicon contents and a carbon content below 0.30% can be welded without any special precautions. When the carbon content exceeds 0.30%, preheating of the casting prior to welding may be advisable. The low-temperature preheat (120 to 205
C, or 250 to 400
F) reduces the rate at which heat is

extracted from the heat-affected zone (HAZ) adjacent to the weld. Preheating also helps to relieve mechanical stresses and to prevent underbead cracking, because hydrogen is still relatively mobile and can diffuse away from the last areas to undergo a metallurgical transformation. Preheating minimizes the chances of a hardening transformation occurring in the HAZ and thus reduces the hardness adjacent to the weld. The metallurgical notch that could result from the hardened area can be eliminated in this manner. Generally speaking, if the hardness of the HAZ after welding does not exceed 35 HRC or 327 HB, preheating of the casting for welding is not required. As additional alloying elements are added to the steel, the need for preheating increases. Most of the low-alloy steels, such as the SC8630, SC8730, or SC4130 steels, require some preheating. When properly preheated, such steels are readily welded. A comparison can be made between the weldability of two steels by comparing their carbon equivalents (CE). One such calculation is: CE ¼ %C þ %Mn=6 þ ð%Cr þ %Mo þ %VÞ=5 þ ð%Cu þ %NiÞ=15

Another includes silicon: CE ¼ %C þ ð%Mn þ %SiÞ=6 þ ð%Ni þ %CuÞ=15 þ ð%Cr þ %Mo þ %VÞ=5

Steels having the same CE will have approximately the same weldability and will require similar preheating and other precautions. Other factors (section thickness) will also affect weldability. The alloy steels of higher hardenability, for example, the cast 4300 series, require more care and must be preheated in the range of 205 to 315
C (400 to 600
F). In the case of some low-alloy steels, especially those subject to toughness requirements, the interpass temperature (maximum or minimum temperature of the weld bead before additional beads are laid) must be maintained at a prescribed low level to prevent embrittlement of the parent metal. High-Alloy Steels. Most of the high-alloy steels are readily weldable, especially if their microstructures contain small percentages of d-ferrite. Grades of stainless can become sensitized and lose their corrosion resistance if subjected to temperatures above 425
C (800
F). Great care must therefore be used in welding to be certain that the casting or fabricated component is not heated excessively; thus, a waterjet sprayer may be used between welds to maintain the interpass temperature, and stainless steels are almost never preheated. Control of heat input and interpass temperature would also apply to heat-resistant steels. Complete reheat treatment may be required after welding in stress-corrosion cracking service performance conditions. Heating the casting above 1065
C (1950
F) and then cooling it

982 / Steel Castings rapidly redissolves the carbides precipitated during the welding operation and restores corrosion resistance. If solution annealing cannot be used, alloying elements such as titanium and niobium can be added to form stable carbides. Another option is to lower the carbon content. As the alloy content of the high-alloy grades is increased, welding without cracking becomes more difficult, due to the tendency of forming microfissures adjacent to the weld. This tendency toward microfissuring increases as nickel and silicon contents increase and carbon content decreases. The microfissuring is reduced by extremely low sulfur contents. In welding these grades, low interpass temperatures, low heat inputs, and peening of the weld to relieve mechanical stresses are effective. Postprocessing of some sort is typically required on all steel castings. Machining, plating or coating, inspection, assembly, and packaging are common practices. Additional activities such as straightening can be included when driven by the design application. Cast steel is moderately machinable and would be similar to that of wrought steels. Cast steel is readily weldable. Coatings can be applied to cast steels to improve wear or corrosion protection, as one would do with wrought steels. The most common coating on steel castings is paint, and surface preparation is an integral part of the process.

Nondestructive Examination Nondestructive examination (NDE) is a valuable tool for ensuring manufacturing quality and part performance in steel castings. The existing quality-assurance methods are workmanship standards. The current design standards often require large factors of safety and significant testing of a part to failure, which is expensive. The inspection level is set conservatively to ensure no failures in the field. Understanding the advantages and limitations of NDE methods is important. Research has shown that simulation is a good way to predict part quality and know how the part will perform. Simulation also has the advantage of minimizing testing costs and reducing development time. Similar to welding procedures that rely on the same principles of metal solidification as castings, casting procedures call out testing requirements to ensure mechanical performance. First-article tests typically call out x-ray and magnetic particle inspection to ensure the quality level of the casting. Therefore, ensuring the quality and performance of a casting is very similar to that of a weldment. Since the same NDE practices are used for all metal manufacturing techniques, the challenges with repeatability and reliability are applicable to every process. Traditionally, nondestructive testing has been used to certify casting quality. Soundness is verified through the use of radiographic inspection. Surface quality is evaluated using magnetic particle inspection. More recently, the use of

computer simulation of the casting solidification, integrated with finite-element analysis of its performance, has been used to design optimal casting configurations. The development of these tools allows the designer to ensure that critical areas of the part meet requirements while ensuring the most economical means of manufacturing the whole part. Quality is a primary concern for casting designers and users. The casting process is more poorly understood and variable than many alternate manufacturing methods. Solidification introduces features in manufacturing that may limit component performance. Even with these limits on understanding and confidence, castings are used in the most demanding and critical applications. From the production of prosthetic hip and knee replacements, to single-crystal jet engine turbine blades, to crane hooks for lifting off-shore platforms, castings are used in the most difficult and demanding applications. From the producers’ point of view, many designers and customers are designing and specifying castings incorrectly due to fundamental misconceptions about castings and their performance. For steel castings, many of these misunderstandings can be resolved by relying on industry standards and practices that achieve high performance with high reliability. As designers and customers better grasp casting process technology and specification, better cast components can be developed. According to ASTM E 1316–2005, “Standard Terminology for Nondestructive Testing,” components have defects only when they fail to meet the specification requirements. No matter how large a crack in a part may be, from a specification and technical point of view, it is not considered a defect unless an inspection for cracks is specified and the crack exceeds the required acceptance criterion. No real part has perfectly flat surfaces, perfectly square corners, or perfectly sized and located features. The amount of deviation from the ideal that is tolerated is expressed in the tolerances for geometric design. All real parts have mechanical properties that are measurably different than the typical or specified minimum values for the grade. A conservative design approach ensures adequate performance. All real parts have imperfections and discontinuities. Material production processes deliver materials that contain nonuniform properties and other variations. Forming the part introduces other imperfections and discontinuities. Testing is used to ensure that the part will meet the performance requirements demanded. For each critical performance attribute to be evaluated, a test method should be identified and an acceptance criterion established. A part contains defects only when the part fails to meet the specified requirements. Much of the common discussion of defects in castings treats ordinary phenomena of solidification, such as shrinkage, as a defect. Shrinkage is only a defect if soundness is specified and the

shrinkage exceeds the imposed acceptance requirement, or if a functional requirement such as a pressure test is specified and not met. This inexact terminology confuses the user of castings by implying that castings are full of defects when, in fact, proper design and specification can easily produce a defect-free casting. Defect-free does not mean perfect, whatever perfect would mean. This situation is made worse by poor failure analysis, where the cause of failure is reported as the part feature that initiated the final fracture. Most failures in new designs are due to inadequate design, not poor-quality part production. Most field failures are due to product misuse and not poor-quality component manufacture. When a part is subjected to unsurvivable loads, it will fail through the most heavily loaded section, and the failure will initiate at the largest performance-limiting feature in the heavily loaded section. In both cases, it is possible to find the fracture-initiating feature, but it is fundamentally incorrect to report that it caused the failure and was therefore a defect. A discontinuity or NDE indication of voids does not necessarily cause a part to fail. If the discontinuity is less than the critical size, has a certain orientation, or is in a location of low stress, the feature will not affect the suitability for service of the part. It is helpful to recognize that all real parts have performance limitations of a finite design life. The performance-limiting features of the part may be a material property, physical property, or even a geometric feature. All manufacturing processes can limit part performance. The challenge is to use the characteristics of the manufacturing process to enhance the performance of cleverly designed parts. Quality is not freedom from defects but desired and predictable performance. This is not limited to the fitness for service but may also include aesthetics or manufacturability. Testing of parts can be either destructive or nondestructive. For example, a part can be loaded to failure, giving a measurement of its actual load-carrying capacity, but after testing, the part is no longer usable. Nondestructive testing evaluates parts without destroying their ability to be used. Since the term test often implies the determination of a property value of a specimen of material, many prefer not to use it for nondestructive methods. While the title of ASTM Annual Volume 3.03 uses the term nondestructive testing, the standard indicates that nondestructive examination, nondestructive evaluation, and nondestructive inspection are equivalent. In the standard, under the definition of test and inspection, it is indicated that examination is the preferred term. This article follows that practice, calling the techniques nondestructive examination. Unfortunately for designers and users as well as the producer, NDE standards are workmanship standards. Workmanship standards specify a level of acceptance based on the general capability of the process and not on the performance of the part. They are set not on an engineering

Selection and Evaluation of Steel Castings / 983 evaluation of the impact on part performance but on the subjective judgment of standards writers. Acceptance levels are not chosen based on application requirements but are arbitrary. The inherent subjectivity of NDE standards limits their usefulness. Visual Inspection. Casting specifications may contain ambiguous wording in regard to visual inspection, such as, “Castings are to be clean and free from injurious defects.” There is no definition for “defect” and no basis for judgment as to what was “injurious.” If the purchaser said, “A surface condition was injurious; it had to be removed and welded,” this would also be problematic. Or, if the “injurious defect” had to be “completely removed to sound metal,” there is no definition for “sound metal” and no basis for judging “completely.” Requirements of this type can be easily misunderstood and misapplied; they can cause no end of grief for the designer, user, and producer. ASTM A 703, A 781, A 957, and A 985 specifications have replaced such ambiguous wording with the requirement, “The surface of the casting shall be examined visually and shall be free of adhering sand, scale, cracks, and hot tears.” This preferred wording goes on to clarify the extent of removal required: “Unacceptable visual surface discontinuities shall be removed and their removal verified by visual examination of the resultant cavities.” Unfortunately, visual standards are subjective, and producers commonly overfinish the casting by grinding and welding without improving the surface or adding value for the customer. The designer and user can specify any surface they like, even for aesthetic reasons. Frequently, slight changes in the casting process alter the texture of the surface, and this change triggers concern for the user. Recent studies show that visual inspection in the producer’s operations is costly and variable. In a repeatability and reproducibility study, two small castings were twice given to two inspectors. After the castings were marked for additional grinding and welding, they were cleaned and returned for a second inspection. After each inspection, the marks were covered by white dots, and the dots counted. The same casting could have 45 dots of work or 0, depending on the visual inspection. In one case, an engineer, to avoid any unnecessary finishing, walked a single large casting through the process, and that casting had one-third the normal grinding and welding. In a similar study, the time spent grinding the castings was evaluated. Typically, 80% of the grinding was not required to process the part but to meet the surface-quality requirements. Unnecessary finishing increases cost and stretches out production times. Traditional measures of surface roughness are not normally useful in casting surface inspection, since the texture of the casting is more severe and long range than machined surfaces. New surface inspection techniques using laser topography or other optical means may be useful for visual inspection. It may be possible

to use these data-intensive surface measurements to develop a less subjective standard. Fundamentally, it is not clear what role surface roughness may play in performance. Current industry best practices are to clearly communicate the need and expectation for surface finish. Avoid any finishing or inspection of surfaces to be machined. Do not do any finishing on surfaces used as datum targets. Accept the normal shot-blasted cast surface unless otherwise required. In highly loaded areas of the casting, where bending and fatigue limit performance capability, more stringent standards may be required. In this case, it is common to apply magnetic particle or liquid penetrant inspection techniques. Magnetic particle inspection is used to detect surface features that disrupt the magnetic field in a magnetized part. Nonmagnetic materials cannot be inspected with this technique. For steel castings, magnetic particle inspection is a valuable way of identifying cracks that are so tight or fine that they escape visual inspection. It also picks up surface and slightly subsurface inclusions and porosity that typically are shallow and have little effect on performance. Magnetic particle methods for dry powder and wet inspection are set forth in ASTM E 109 and E 138. Acceptance criteria can be selected from a set of reference photographs in ASTM E 125 that depict the appearance of casting surface conditions revealed by the dry powder magnetic particle technique. Different types of discontinuities do not have equal effects on performance, and an effort should be made to assign different acceptance levels to areas of the casting, based on the stresses to which each area is subjected in service. Purchasers can specify ASTM A 903. Instead of pictures and subjective comparisons, this standard sets out dimensions and frequencies of indications for acceptance criteria when using magnetic particle and liquid penetrant inspections. A misuse of the standard is to tighten the acceptance criteria on a casting when inclusions or porosity are found on a machined surface. Since the inspection is unable to see below the surface, it does not detect porosity or inclusions at the machined surface. Worse, the tighter standard requires the producer to grind and weld a surface that will be removed by machining. This adds cost and delays production. When a producer is responsible for machining and finds a problem at the machined surface, no magnetic particle inspection is imposed. No NDE technique can routinely detect this undesirable condition. Ultrasonic examination may be able to detect this condition, but the casting surface would need to be machined in order to prepare it for examination. The producer makes changes in the casting process or adds a larger machine stock allowance to solve the problem. In fact, many users avoid this problem by purchasing a rough-machined casting from the producer. Magnetic particle inspection is valuable for first-article inspection to ensure that the casting

design and rigging do not allow cracking in production. If cracks are found, magnetic particle inspection is used to identify the castings to be welded. Magnetic particle inspection is also used on heavily loaded casting surfaces to ensure surface integrity. Magnetic particle inspection is used to audit the process by inspecting samples of normal production. Liquid penetrant inspection is another surface-feature detection method. It is not generally used on as-cast or shot-blasted surfaces because of the likelihood of obtaining false indications. Penetrant may be retained in surface roughness and give indications unrelated to actual surface conditions. The penetrant method is best suited for use on machined, ground, or very smooth as-cast surfaces. Liquid penetrant inspection is of particular importance for austenitic alloys because they are nonmagnetic; therefore, their surfaces cannot be examined by magnetic particle inspection. ASTM E 165 describes the standard method for conducting this test. A set of reference photographs for acceptance or rejection is contained in ASTM E 433. Acceptance criteria are found in ASTM E 125 for the dry powder magnetic particle technique. Each of the documents must specify actual dimensions, including maximum length of indications and number of indications per unit area. Liquid penetrant inspection is subject to many of the limits and misuse cited in magnetic particle testing, in addition to false indications. It is included in ASTM A 903, and this provides workmanship levels. All the specifications for surfaces must recognize the scale of the part. Frequently, the highest level with the smallest allowable features is specified for large castings. This is unrealistic, adds cost, and becomes the subjective standard of the customer’s inspector. Liquid penetrant inspection, like magnetic particle, should be used for first-article inspection, process auditing, crack detection, and to monitor heavily loaded areas on critical castings. Radiographic Inspection. Customers commonly see radiography or x-ray inspection as essential to the production of high-quality steel castings. This creates a permanent record for liability defense, especially on critical castings. Unfortunately, radiography as a technique overpromises and underdelivers. The subjective and nonreproducible application of radiography, with limits on resolution and indication location, severely impedes its usefulness in performance evaluation. There are three basic groups of reference radiographs issued by ASTM International for evaluation of steel castings: E 446 applies to castings up to 51 mm (2 in.) in thickness; E 186 is for heavy-walled 51 to 114 mm (2 to 4.5 in.) thick sections; and E 280 is for heavy-walled thickness of 114 to 305 mm (4.5 to 12 in.). Currently, reference radiographs become standards for acceptance and rejection after the purchaser and the producer have agreed, in the purchase order or contract, to the acceptable severity level for each individual type of discontinuity. The choice of discontinuity

984 / Steel Castings severity level should ideally be based on realistic evaluation of design and stress analysis criteria under anticipated service conditions. Generally, low severity levels are specified for pressurecontaining castings with high-pressure rating and wall sections of 25 mm (1 in.) or less. Likewise, low severity levels are specified for machinery or dynamically loaded castings subject to high fatigue and impact stresses and with wall sections of less than 13 mm (0.5 in.). As wall sections increase and the fatigue and impact stresses are reduced, severity levels become somewhat relaxed. For structural castings, which are not dynamically loaded, moderate severity levels are usually specified, and again, for heavier sections (approximately 76 mm, or 3 in.), higher severity levels are usually called for. Unfortunately, the workmanship basis of radiography and the subjective application of the acceptance criteria do not support any particular application of radiography for performance. None of the reference radiographs are based on any kind of test data, and the severity levels are not graded to any basis of acceptability as to service performance. Since the radiograph does not indicate where in the cross section the indication may be, it is inherently unable to predict performance. The radiographic standards are used as a reference point in communicating the purchaser’s requirements. Because the standards are subjective, they are not reproducible. The radiographic images are not to be evaluated based on gray scale, and most importantly, the reference radiographs are to be prorated to the actual casting image size. No standard method or approach is suggested for this prorating, which allows the standard to have the appearance of objectivity while being completely subjective. A recent study applying gage repeatability and reproducibility to only the reading of existing plate-casting radiographs demonstrates the subjective nature of this technique. One hundred twenty-eight x-rays were rated seven times for shrinkage type and level. Unanimous agreement was 37% of the x-rays on shrinkage type, 17% on shrinkage level, and 12.5% on both type and level. Agreement was higher if the castings were completely sound or very unsound. Two of the radiographers rated the x-rays twice. Comparing each radiographer’s second rating to their first, both radiographers gave different type ratings for at least 19% of the x-rays and different level ratings for at least 34% of the x-rays. Both radiographers also reversed accept/reject decisions for 10 to 15% of the x-rays. Thus, this shows the subjectivity of radiographic examination. For this reason, most market sectors select some radiography for the first-article evaluation. This allows the adequacy of the rigging and solidification soundness to be evaluated. In critical parts, radiography may be needed on each part. The casting process is frequently audited with occasional radiography for noncritical parts. For many applications, the imposition of level 3 communicates the desire for a

reasonable soundness that is ordinarily achievable in structural castings. For critical areas, a radiographic level of 1 may be required. Requiring quality levels in excess of those justified by actual service conditions adds needlessly to the cost of the casting. It should also be kept in mind that the entire casting need not necessarily require radiographic inspection, and that the same severity levels need not apply to all areas of the casting. Careful analysis or, at least, good judgment can effect sizable cost-savings. In any case, the areas to be radiographed with the required severity level should be indicated on the casting drawing. Ultrasonic Inspection. Although the ultrasonic method of inspection has not been in common use for as long as radiographic methods, it is nevertheless a valuable tool for examining heavy-wall castings for internal discontinuities. The first ASTM specification for ultrasonic inspection of steel castings was issued in 1970 and is for longitudinal-beam ultrasonic inspection of heat treated carbon and low-alloy steel castings. In general, this inspection method is not useful for austenitic steel castings due to the large grain size of these castings. It is well recognized that ultrasonic inspection and radiography are not directly comparable. However, the technique is invaluable in detecting discontinuities in heavy sections, where radiographic methods would be considerably slower. Since no image of the discontinuity is obtained, considerable judgment must be exercised in the interpretation of results. The addition of computer analysis allows a wider application of ultrasonic inspection. It is more quantitative (digital) than most of the other NDE techniques. This should allow the enhanced use of ultrasonic, especially to develop standards that ensure performance. At the least, ultrasonic inspection can assess the amount of sound wall that exists to carry the design load. It can also pick up areas of microporosity too fine to be detected by radiography. One approach in the examination of large, heavy-wall castings, when ultrasonic evaluation may not be acceptable to the purchaser, is to first inspect with ultrasonic to identify areas with indications and then check these areas with radiography. Another possibility, since radiography does not reveal the depth of a discontinuity, is to follow radiography with ultrasonic in order to determine and evaluate the depth of the discontinuity. Leak Testing. Fluid-handling components are often leak tested to ensure performance. The component manufacturer frequently does this test. If the casting fails, it becomes the responsibility of the producer to replace or rework the casting. Often, leaks are due to porosity too fine to be detected by radiography. It may be possible to use solidification analysis software and a minimum Niyama value (a porosity prediction model), which would evaluate the microporosity, to assess castings for leaks. Hardness testing is an efficient means to ensure that the processing of an individual

casting was done correctly. The hardness of cast steel has a good correlation to strength but it is not a good indicator of quality in austentic and duplex stainless steels. Hardness testing location(s) should be stipulated. Tests and Castings ASTM International requirements and commercial practices normally result in the producer performing mechanical tests on each heat of steel to verify conformance with specification requirements. Many purchasers misunderstand the significance of these tests and believe that the test bar results for each heat will be the same as the properties of the casting. This is not true in most products, certainly not in steel products and clearly not in castings. The routine mechanical tests that are performed for each heat are to verify the capability of the material, not to determine the properties of the product. The properties of bars cut from the product, which are cut from the casting, depend on the location, section size, heat treatment, and shape of the casting. If verification of mechanical properties of the product is required, a check of hardness in a critical area is done. If a large critical casting is produced, a large connected test coupon may be added to the casting for properties verification. In this case, the properties of the test coupon are not given in the material standard but are subject to agreement between purchaser and producer. If a test bar is removed from a casting and it fails to meet the specification requirements for the material grade, the casting is not defective. Only if the test bar from the heat fails to meet the properties is the material unacceptable. Test coupons are typically made from an ASTM International double-legged keel block in accordance with ASTM A 985/A 985M for investment castings and A 703/A 703M for pressure-containing parts (Fig. 8). There is good correlation between tensile test data for attached specimens and keel block specimens (Table 3). Analysis has shown that two tests provide a 95% probability that the test sample and actual strength are within  þ6.9 MPa (þ 1 ksi) for ultimate tensile strength (UTS),  þ11 MPa (þ  1.6 ksi) for yield strength, and that elongation is within þ3%. Table 3 demon strates that data taken from an integrally cast production casting are similar to that of a keel block casting. Larger castings have a great propensity to form discontinuities and be affected by section size and geometric and metallurgical effects on material properties. Foundry practices can be used to provide optimal properties. Analysis of CA6NM and 105/85 cast steel materials has shown that alloying reduces any impact of section thickness and geometry. This same analysis showed that shrinkage in a casting had little effect on yield strength. In theory, since shrinkage is along the centerline—a lowstress region—and fatigue failures typically occur from surface defects, the fatigue limit should likewise be minimally affected. Tests have shown that tensile properties are not greatly affected by different types of discontinuities. There is a mostly direct relationship between the severity level of a defect and the

Selection and Evaluation of Steel Castings / 985

Fig. 8

Keel block coupon according to ASTM A 703/A 703M and ASTM A 985/A 985M. Dimensions given in inches. See the ASTM standards for metric values.

Table 3 Number of tests required for various degrees of accuracy using double-legged keel block test specimens Acceptable variation

Tests required for indicated probability

Property tested

MPa

ksi

99%

95%

90%

80%

70%

Tensile strength

0.69 1.38 2.07 2.76 3.45 4.14 4.83 5.52 6.89 3.45 4.14 4.83 5.52 6.21 6.89 8.27 9.65 11.03 1% 2% 3% 1% 2% 3% 4% 5% 6%

0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 1 0.5 0.6 0.7 0.8 0.9 1 1.2 1.4 1.6

266 72 32 18 12 8 6 5 3 34 24 18 14 11 9 6 5 4 17 5 2 60 15 7 4 3 2

166 42 19 11 7 5 4 3 2 20 14 10 8 6 5 4 3 2 10 3 2 35 9 4 3 2 ...

177 30 13 8 5 4 2 ... ... 14 10 8 6 5 4 3 2 ... 7 3 2 24 6 2 ... ... ...

72 18 8 5 3 2 ... ... ... 9 6 5 4 3 3 2 ... ... 5 2 ... 15 4 2 ... ... ...

47 12 6 3 2 ... ... ... ... 6 4 3 3 2 ... ... ... ... 3 2 ... 10 3 2 ... ... ...

Yield strength

Elongation in 50 mm (2 in.)

Reduction in area

degradation in tensile properties. For instance, gas porosity will reduce the UTS by a little over 20 MPa (3 ksi) per severity level. Of the various types of porosity, linear shrinkage has the biggest impact on strength. However, even this type of defect only reduces the UTS by a little

over 55 MPa (8 ksi) per class. Keeping in mind location, use, and factor of safety, it is easy to see why steel castings with discontinuities can survive well past testing requirement levels. Testing on a hanger bracket casting has shown that defects do not necessarily lead to

part failure. Brackets with severity level 2 and five discontinuities had very little influence on the static and fatigue properties. Only in a few cases, where the defect was at or near the surface, did the location of the failure change. However, even in these cases, the properties of the casting were not impacted. Obtaining test specimens from an actual part is costly and renders the part useless. Thus, a cast coupon or separately cast keel block is used to determine that the material meets the requirements of the specification. However, in some instances, such as material thickness over 76 mm (3 in.), the specification may require that a test specimen be taken from an actual casting. Mechanical properties obtained from test bars represent the quality of the steel but do not necessarily represent the properties of the castings themselves. Solidification and heat treatment affect the properties of the casting, and these are dependent on section thickness, size, and shape. General Rules. Testing a section from a part is expensive but provides a data point on properties for the part. Testing a part to failure can likewise provide valuable information but comes at a cost. If a part with known level and type of discontinuity from NDE is tested, then setting the level one step higher for production parts should ensure desired performance characteristics. Working with a quality foundry will help ensure that discontinuities are minimized in the casting through optimized manufacturing techniques. Selecting a source based on best value versus lowest cost always pays for itself over the long run. As stated earlier, using analysis software is not only cost-effective but a very good means to identify discontinuities. Using zoning to inspect a casting will ensure cost-effective inspection. Areas of the part that see higher performance demands are inspected more frequently and to a higher level than other sections. Thus, sections with greater discontinuities that do not affect the part performance will not prevent the part from passing inspection. The impact of a discontinuity on performance depends on the casting size and use. A good starting point for NDE of steel castings is ASTM A 903 with acceptance level III indication length. Radiography is good to use in critical sections with level III as a starting point, and a rating per type of discontinuity should be applied. Ultrasonic testing can be used in castings with section sizes over 150 mm (6 in.). A tensile test should be carried out for each heat to ensure the material mechanical properties. Hardness testing should be done on critical sections. A hardness test is relatively low in cost, and hardness correlates with other mechanical properties. Since casting and welding are similar processes, the same NDE can be used. In the future, modeling casting production and solidification, in particular, will be key. The ability to link properties based on solidification patterns, design with these real-world properties, and craft NDE standards that assure performance promises a new generation of highperformance cast steel components.

986 / Steel Castings

Purchasing To obtain the highest-quality product, the part should be designed from inception to take advantage of the flexibility of the casting process. The foundry must have the part drawing, solid model or pattern equipment, and know the number of parts to be made. To take advantage of the casting process, the foundry should also know which surfaces are to be machined and where datum points are located. Reasonable dimensional tolerances must be indicated and should be appropriate for the casting process used. Tolerances are normally decided by agreement between the foundry and purchaser. Close cooperation between the purchaser’s design engineers and the foundry’s casting engineers is essential to optimize the casting design, in terms of cost and performance. Many specifications, such as ASTM International and ISO, have other requirements based on agreement between the purchaser and supplier; thus, communication between the designer, purchaser, and foundry is important in maximizing the benefits of steel castings. Purchasers should consider sourcing the foundry as the prime contractor. This invokes quality in the casting up front and can help assure that the delivery schedule is met. When making inquiries or ordering parts, all pertinent information must be stated on both the inquiry and order. This information should include all of the following components:  Casting shape—either by drawing, electronic

  



file, solid model, or pattern. Drawings or solid models should include dimensional tolerances, indications of surfaces to be machined, and datum points for locating. If only a pattern is provided, then the dimensions of the casting are as predicted by the pattern. Material specification and grade (e.g., ASTM A 27/A 27M—95 grade 60-30 class 1) Number of parts Supplementary requirements (e.g., ASTM A 781/A 781M—95 S2 radiographic examination) a. Test methods (e.g., ASTM E 94) b. Acceptance criteria (e.g., ASTM E 186 severity level 2, or MSS SP-54-1995) Any other information that may contribute to the production and use of the part (finishing and packaging requirements, documentation requirements)

To produce a part by any manufacturing process, it is necessary to know the design of the part, the material to be used, and the testing required. Machining. The foundry engineer is responsible for giving the designer a cast product that is capable of being transformed by machining to meet the specific requirements intended for

the function of the part. To accomplish this goal, a close relationship must be maintained between the customer’s engineering and purchasing staff and the casting producer. Jointly and with a cooperative approach, the following points must be considered:  The molding process—its advantages and

limitations

 Machining stock allowance to assure cleanup

on all machined surfaces  Design in relation to clamping and fixturing devices to be used during machining  Selection of material specification and heat treatment  Quality of parts to be produced Layout. It is imperative that every casting design, when first produced, be checked to determine whether all machining requirements called for on the drawings may be attained. This may be best accomplished by having a complete layout of the sample casting to make sure that adequate stock allowance for machining exists on all surfaces requiring machining. For many designs of simple configuration that can be measured with a simple rule, a complete layout of the casting may not be necessary. In other cases, where the machining dimensions are more complicated, it may be advisable that the casting be checked more completely, calling for target points and the scribing of lines to indicate all machined surfaces.

2. H. Davis, G. Troxell, and G. Hauck, The Testing of Engineering Materials, 4th ed., McGraw-Hill, 1982, p 314 3. M. Blair and T.L. Stevens, Ed., Steel Castings Handbook, 6th ed., SFSA and ASM International, 1995 SELECTED REFERENCES  C. Beckermann et al., “Analysis of ASTM



    



Research and Development Current and future research will always strive to improve the performance and manufacturing of steel castings. On-going research today (2008) is looking at modeling discontinuities and integrating the analysis with finite-element analysis software to facilitate the prediction of part performance. Modeling capabilities are also expanding to fill and flow analysis. Materials development centers around better performance characteristics, such as corrosion resistance, or better mechanical properties, such as strength to weight. Improvements to manufacturing are working on faster tooling technology, pressurizing risers, new pattern materials, and automated grinding capability. REFERENCES 1. C.W. Briggs and R.A. Gezelius, The Effect of Mass upon the Mechanical Properties of Steel, ASM, Vol 26, 1938, p 367



  

 

X-Ray Shrinkage Rating for Steel Castings,” Paper presented at 54th SFSA T&O Conference (Chicago, IL), Nov 2000 C. Briggs, “The Evaluation of Discontinuities in Commercial Steel Castings by Dynamic Loading to Failure in Fatigue,” Steel Foundry Research Foundation, Rocky River, OH, 1967 J. Carpenter and B. Hanquist, Specifying Steel Castings—Keeping Alloy Composition in Mind, Mod. Cast. Corrosion: Environments and Industries, Vol 13C, ASM Handbook, ASM International, 2006 Corrosion: Fundamentals, Testing, and Protection, Vol 13A, ASM Handbook, ASM International, 2003 Corrosion: Materials, Vol 13B, ASM Handbook, ASM International, 2005 F. Peters et al., “Assessing Process and Product Variability,” Paper presented at 57th SFSA T&O Conference (Chicago, IL), Nov 2003 F. Peters et al., “Variability: Causes, Concerns, and Corrections,” Paper presented at 58th SFSA T&O Conference (Chicago, IL), Nov 2004 A. See and B. Hollandsworth, “Mechanical Properties of Test Bars Compared to Castings,” Paper presented at 55th SFSA T&O Conference (Chicago, IL), Nov 2001 Steel Castings Handbook, 6th ed., SFSA and ASM International, 1995 Steel Castings Handbook Supplement 3— Dimensional Tolerances, SFSA, 2003 Steel Castings Handbook Supplement 6— Repair Welding and Fabrication Welding of Carbon and Low Alloy Steel Castings, SFSA, 1985 Steel Castings Handbook Supplement 7— Welding of High Alloy Castings, SFSA, 2004 P. Wieser, Ed., Steel Castings Handbook, 5th ed., Edward Brother, 1980

ASM Handbook, Volume 15: Casting ASM Handbook Committee, p 989-991 DOI: 10.1361/asmhba0005306

Copyright © 2008 ASM International® All rights reserved. www.asminternational.org

Nonferrous Casting—An Introduction CASTING OF NONFERROUS ALLOYS on a tonnage basis is dominated by aluminum, which is cast by ingot and continuous processes for primary mills and by all foundry (shape casting) processes. In terms of shape casting, the percentage of nonferrous shape casting tonnage is roughly as follows (Ref 1):
Aluminum high-pressure die casting, 54%
Aluminum sand and permanent mold





casting, 12% Zinc high-pressure die casting, 11% Aluminum lost foam casting, 10% Sand mold casting of copper alloys, 10% Magnesium high-pressure die casting, 3%

The articles in this Section, “Casting of Nonferrous Alloys,” describe the shape casting of aluminum, copper, and zinc alloys along with articles on the continuous casting of aluminum and copper. Casting of magnesium alloys is detailed in the article “Magnesium and Magnesium Alloy Castings” in this Volume. This article briefly reviews the melt processing and casting of other nonferrous alloys other than aluminum, copper, magnesium, and zinc. More details are also given in the articles in the next Section, “Nonferrous Castings.”

Nickel Alloys Foundry practice for nickel-base alloys is, for the most part, similar to that used for cast stainless steels. Requirements for alloy cleanliness involve increased use of the ladle as a refining vessel, similar to steel (see the article “Steel Melt Processing” in this Volume). Melting involves vacuum induction melting (VIM), which is followed by consumable electrode remelting methods of electroslag remelting (ESR) and/or vacuum arc remelting (VAR). Electroslag remelting is not widely used, but several producers of critical rotating components in the gas turbine industry have adopted the use of the triple-melt sequence (Ref 2). First, VIM produces an initial electrode with low oxygen and precise chemistry, followed by ESR. The ESR electrode will be clean and sound but may contain freckles. The final segregation-free structure is obtained through the application of a third melting process (VAR) to the ESR ingot. Investment Casting of Superalloys. A number of casting processes can provide near-net

shape superalloy cast parts, but essentially all components are produced by investment casting. The characteristic physical and mechanical properties and complex, hollow shape-making capabilities of investment casting have made it ideal in applying the high-temperature properties of superalloys. Dimensional tolerances for superalloy investment castings are generally 0.075 mm (0.003 in.), with section thicknesses as low as 1.25 mm (0.05 in.) or less, and excellent surface finishes can be obtained. Nickel-base superalloy castings are produced by investment casting under vacuum, while most cobalt-base superalloys are produced by investment casting in air. Improvements in properties have been made not only through control of composition but also through more precise control of microstructure. The absence of grain boundaries in single-crystal alloys permits elements such as carbon, zirconium, and boron to be deleted from the composition. The resulting increase in melting point in turn provides improved flexibility in alloy composition and heat treatment (Ref 2).

Titanium Alloys (Ref 3) Titanium castings have been cast in machined graphite molds, rammed-graphite molds, and proprietary investments used for precision investment casting. The principal technology that allowed the proliferation of titanium castings in the aerospace market was the investment casting method coupled with the adaptation of hot isostatic pressing (HIP) to the most critical castings (see the article “Titanium and Titanium Alloy Castings” in this Volume). Significant design complexity, tolerance, and surface-finish control have been achieved, and large parts approaching 1.5 m (5 ft) in diameter can be cast. Castings generally are produced by vacuum arc remelting titanium in a copper, water-cooled crucible and pouring into molds. If the initial investment can be justified, hearth-melting processes may find application in the future for the remelting of alloy prior to casting. Of necessity, the casting mold systems must be relatively inert to molten titanium. Proprietary lost wax ceramic shell systems were developed by the various foundries engaged in

titanium casting. Usually, the face coats of the ceramic shells are made with the proprietary coatings, and conventional refractory systems are used to add shell strength. Regardless of mold type (e.g., investment or rammed graphite), foundry practice focuses on methods to control both the extent of the reaction of the molten alloy and mold and the subsequent diffusion of reaction products inward from the cast surface. Depth of surface contamination can vary from nil on very thin sections to more than 1.5 mm (0.06 in.) on heavy sections. On critical aerospace structures, the brittle alpha case is removed by chemical milling. The depth of surface contamination must be taken into consideration in the initial wax pattern tool design for investment casting. The wax pattern and resultant casting are made slightly oversized, and final dimensions are achieved by careful chemical milling. Metal pouring temperature, mold temperature, casting force (if centrifugally cast), and thermal conductivity of the metal and the mold, as well as the cooling environment, are all factors in the production of a good investment-cast titanium part. Casting Processes. The molten metal resulting after remelting is poured into either an investment mold or a rammed-graphite mold as just described. Pattern making and investment-cast mold techniques are similar to those used with superalloy technology. Although rammed-graphite molds are different from investment molds, they are similar to conventional sand molds. Cores are used to cast hollow parts. Porosity continues to be a potential problem but one that is addressed for premium aerospace castings by the use of HIP, as described subsequently. Once the mold-metal reaction problem is addressed, the most significant problem—with respect to titanium and titanium alloy castings—is achieving sufficient levels of superheat in the molten metal to maximize flow and mold-fill characteristics. In many cases, either a centrifugal table or mold preheating (or both) is used to ensure proper mold filling. A significant difficulty related to large titanium castings is the problem of porosity. However, by use of HIP, internal soundness of titanium castings can be improved to the point that no porosity or small voids can be detected. Weld repair is used to close gross defects after HIP. Because strength increases and ductility (toughness) decreases as oxygen level increases,

990 / Casting of Nonferrous Alloys oxygen content is a matter of concern regarding titanium alloy castings. Control of oxygen levels in cast components is achieved mostly by selection of melt stock, but hearth melting can make an additional improvement possible. An ingot with a low oxygen content generally results in a casting with the lowest oxygen content. The exact role of revert in oxygen-content control and in alloy element segregation is not clear; both revert and virgin ingots are used. Oxygen also can be introduced to the casting from the surface mold-metal interaction. The rammed-graphite method is the oldest mold technique used to produce titanium castings. The method uses a mixture of graphite powder and associated binders and water additions that are rammed against a pattern to form a portion of the mold. Individual mold segments are then fired and assembled for casting. Most high-performance titanium alloy casting applications rely on the technique of investment casting.

Lead Melting of Lead Alloys. Lead and its alloys constitute a small but commercially important segment of the nonferrous metal casting industry. Principal lead alloys and their areas of importance include Pb-60Sn (solders), antimony-lead and calcium-lead (battery grids), and Pb-7Sn (terne plating). The melting and casting of pure lead and foundry alloys pose fewer problems than with other nonferrous systems. The major lead alloys have relatively low melting points (315 to 425  C, or 600 to 800  F) and are not susceptible to hydrogen absorption and the attendant gas porosity to any great extent. Dross formation can be a problem, especially with alloy constituents, so simple cover fluxes such as graphite or vermiculite are sometimes used. However, oxidation is relatively minor unless temperatures are excessive. In the terne plating industry, where lead-tin coatings are applied by hot dipping steel articles, principally for use in fuel systems, a cover flux is often used. This is similar to the flux used in galvanizing—a ZnCl2-NH4Cl mixture, often with other chloride salts such as KCl. The flux used in terne plating serves to preclean the steel, to absorb pretreatment pickling residue, to prevent spitting as the articles enter the plating bath, to alleviate bath dross formation, and to prewet the article being coated. In the die casting of lead alloy acid battery grids, no treatment is usually necessary at the die cast machine melting pot itself, although a melt cover can be used for heat insulation purposes and to prevent dross buildup. In remelt pots, where trim scrap and other scrap is melted, chloride-base drossing fluxes can be used. Pitch, sawdust, or wood chips are used to agglomerate the residues that accumulate on the surface.

The lead is usually first softened (refined) in the production of the newer Ca-Pb-Sn alloys used for maintenance-free battery grids from either primary metal or recycled scrap. Calcium is then added as an alloy, often under a protective cover such as sawdust, wood chips, or resin to prevent oxidation at the usual higher temperature at which alloying is facilitated. Refining of Lead. A number of fluxing procedures and refining processes are used in the smelting and refining of primary and secondary lead production. Several of these involve chemical flux-refining techniques somewhat analogous to those used with other nonferrous metals. Figure 1 is a flow chart showing a number of important sequential process steps in the refining of lead. These steps are briefly discussed here and are covered in greater detail in Ref 4. Copper Drossing. In copper drossing, the first step, the elements copper, iron, nickel, cobalt, and zinc can be removed, if present, by cooling the lead almost to its freezing point. The liquid solubility of these elements at that temperature is very low; thus, they can be removed as an intermetallic dross as they separate from the heavier lead melt. Copper will precipitate with any sulfur, arsenic, antimony, or tin, in that order. Lead sulfide may also form. These drosses are usually intimately intermingled with the lead metal phase. A molten phase (copper-lead sulfide) may form, and when the newer calcium lead battery scrap is remelted, calcium may be present in this phase, lowering the lead content. If oxides are present, there is also a slag phase, and silica can be added to increase its fluidity, resulting in the creation of iron, zinc, and lead silicates. It is also possible to remove copper from lead down to a level of 0.05% by stirring sulfur into the lead melt at temperatures near the freezing point. When silver or tin is present, copper can be reduced to even lower levels, but a reversion (copper reentering the metallic phase) can occur if stirring is excessive (over 10 min) or if temperature is increased (Fig. 2). Softening. In softening, the second process step, elements more readily oxidized than lead are removed by oxidation. These include tin, arsenic, antimony, and indium, whose free energies of oxide formation are greater than that of lead. The term softening relates to the effect these elements have on lead. When present, they contribute to solid-solution strengthening; thus, their removal results in softening. The Harris process of softening consists of oxidizing the impurities, converting them to sodium salts, and collecting them in a caustic soda melt slag phase that rejects molten lead, providing for relatively easy physical separation. Temperatures of best reactivity are fairly high (700 to 800  C, or 1290 to 1470  F), and thorough mixing of the metal with the salt flux is required. Tellurium, arsenic, and tin can be reacted directly with caustic soda to form lower-sodium salts. A further oxidation reaction, employing

Fig. 1

Flow chart showing steps in the lead-refining process. Impurities remaining are shown in brackets after each step. Source: Ref 4

sodium nitrate (NaNO3) as the oxidizing agent, forms the higher-sodium salts (arsenates, stannates, and antimonates). The intermixed metallic oxide products produced by these reactions are then removed and separated by various wet processing and precipitation methods. A simple softening process involves the refining of lead by oxygen at a somewhat reduced temperature (650  C, or 1200  F) directly in a refining kettle. A 22.5 Mg (25 ton) lead melt containing 0.6% Sb was refined to a level of just 0.02% by blowing 10 m3 (350 ft3) of oxygen for 1.5 h (Ref 5). Moreover, approximately 100 kg/m2 (20 lb/ft2) of antimony was oxidized with oxygen injection, versus a rate of only 10 kg/m2 (2 lb/ft2) using air. Because this is in excess of the expected stoichiometric ratio, the oxygen must also act catalytically. Desilvering is the term applied in lead refining to the removal of elements more noble than lead—silver, gold, and bismuth. These elements can be coprecipitated with lead and zinc by the Parkes process, which is the addition and mixing of zinc with the lead at a temperature just above the freezing point of lead (320  C, or 610  F). Because there is substantial metal value contained by silver- or gold-bearing precipitate residues, technologies (pressing,

Nonferrous Casting—An Introduction / 991

Fig. 3

Final lead refining using the modified Harris process. Source: Ref 4

fluxing techniques. Fractional oxidation or crystallization can yield liquid slag and solid crystals that are purer than the starting metal. Bismuth can also be precipitated from lead by adding alkali or alkaline earth metal reagents, such as calcium and magnesium (the Kroll-Betterton process). Magnesium metal and Pb-5Ca master alloy are stirred into the bismuth-containing molten lead at 420  C (790  F). As the lead is cooled back to its freezing point, CaMg2Bi2 crystals precipitate, leaving the lead refined of bismuth. The precipitate is intermixed with lead, and the latter can be removed by liquation by adding the mixture to the next batch and raising the temperature once again. Further reduction of bismuth is possible by stirring in antimony (or antimonial lead) in the presence of residual calcium and magnesium in the melt after the initial CaMg2Bi2 precipitate has been removed. The antimony addition forms a similar precipitate, which facilitates easier formation and substitution of bismuth for antimony in the precipitate at more dilute concentrations of bismuth. Final Refining. Before casting, a final refining step is carried out on the molten lead, using a modified Harris process at 450 to 500  C (840 to 930  F). Caustic soda or NaNO3 at the rate of 0.1% is stirred into the lead to decrease the zinc and antimony that may remain after previous refining steps to less than 1 ppm each, as shown in Fig. 3 (Ref 4). Arsenic, tin, calcium, magnesium, and iron will also be removed to very low levels with or before the antimony. The softened and highly purified lead thus produced is required in the production of battery alloys, whose electrochemical performance is highly dependent on the removal of impurities.

Fig. 2

REFERENCES

liquation, and zinc distillation, for example) have evolved for separating the noble metals from the precipitate. Zinc Removal. Zinc can be removed preferentially from the base lead metal by preferential oxidation or chlorination or by the Harris process without the use of an additional oxidizing agent:

1. Trends and Survey, AFS, 2002 2. M.J. Donachie and S.J. Donachie, Superalloys: A Technical Guide, ASM International, 2002 3. M.J. Donachie, Titanium: A Technical Guide, ASM International, 2000 4. T.R.A. Davey, The Physical Chemistry of Lead Refining, Proceedings of the Symposium on Lead-Zinc-Tin, AIME, 1980, p 477 5. J. Blanderer, The Refining of Lead by Oxygen, J. Met., Dec 1984, p 53

Decoppering of lead. (a) Effect of sulfur. Curve 1 represents the copper content of silver-free lead as a function of time while stirring in 0.1% S at 330  C (625  F), and curve 4 represents similar conditions for a lead containing 0.04% Ag. Curve 2 resulted when only 0.05% S was added, and curve 3 when a total of 0.1% S was added as ten equal additions of 0.01% S each, at 1 min intervals, approximating plant practice in which the sulfur is added gradually over 10 min. (b) Effect of silver. (c) Effect of tin. Source: Ref 4

Zn þ 2NaOH ! Na2 ZnO2 þ H2

Vacuum dezincing, however, provides much higher zinc recoveries than any of these techniques and is therefore preferred. Bismuth Removal. Bismuth is more noble than lead and therefore cannot be removed by chemical

ASM Handbook, Volume 15: Casting ASM Handbook Committee, p 992-1000 DOI: 10.1361/asmhba0005285

Copyright © 2008 ASM International® All rights reserved. www.asminternational.org

Dross, Melt Loss, and Fluxing of Light Alloy Melts Daniel E. Groteke, Q.C. Designs, Inc. David V. Neff, Metaullics Systems Division, Pyrotek Inc.

DROSS, which is the oxide-rich surface that forms on melts due to exposure to air, is a term that is usually applied to nonferrous melts, specifically the lighter alloys such as aluminum or magnesium. Dross comprises a physical mixture of oxides and entrapped molten or semimolten metal. The dross that forms on aluminum melts is normally considered to be an undesirable material, which is supported by the definition found in most dictionaries: “Scum on molten metal.” While it is certainly the source of much grief in operations that melt aluminum or aluminum alloys, the solid form of dross is far from being undesirable and, in truth, is a valuable co-product that can have a major effect on the profitability of the producer’s operation. This comes from the simple fact that, in addition to the oxides present in the dross, it often contains large quantities of entrapped useful aluminum alloy. The alloy values can potentially be lost from the process stream during the normal removal and cleaning process unless great care is exercised at the generating source or the alloys are recovered in subsequent recycling efforts. The composition of dross is dependent on the base analysis of the alloys being melted, but the major constituent in almost every case is aluminum oxide, with other contaminants added from the other alloy constituents and process contributors. Any new aluminum surface, whether solid or molten, will quickly form a protective oxide when the surrounding medium contains even a very small trace amount of oxygen from any source (free oxygen, moisture, or reducible oxides). Molten surfaces are subject to the same oxidants but are also exposed to other potential contributors in the layer of oxide-rich materials present on the melt. Industries conducting melting operations on aluminum and aluminum alloys operate a wide variety of types of primary melters and holders to provide molten material to their process lines. All are active dross producers, with their own characteristics influencing both the rate and quantity of oxide-bearing material generated. Further, the operation of transferring the molten material between the melters, holders,

and the final solidification point is an additional source of dross generation that can, in many cases, exceed the volumes generated in the actual initial melting.

Dross Formation Many different crystallographic forms of aluminum oxide can exist that are associated with molten aluminum processing. Up to 16 different forms have been identified, although in principal only the major ones need to be considered. The first oxide film to form on molten pure aluminum is generally agreed to be gamma-alumina (g), which is a thin film that inhibits further oxidation. After a brief incubation period, the g film transforms to an alpha-alumina (a) form, which oxidizes and grows at a much greater rate. When alloying elements are added to the pure aluminum, both the rate of formation and the nature of the resultant oxide are changed dramatically. While the common alloying elements (copper, iron, manganese, chromium, zinc, silicon, etc.) in foundry cast alloys have little effect on the rate of oxidation, other elements can play a major role in changing both the rate and volume of dross formation. The most important of these is magnesium, which is present in almost all commercial alloys as either a trace or major alloying element. A second contributing element is lithium, which is never used as an alloying element in conventional casting alloys but may be present in trace amounts in the scrap streams that are used to generate the secondary foundry alloys. Since these scrap streams constitute the source for the bulk of materials poured in the foundry casting industry, lithium, when present, can be one of the more significant elements encouraging dross formation in commercial operations. Even small quantities of magnesium in the cast alloy can convert the dross to a mixed oxide of MgO Al2O3 and Al2O3, with the percentage of the mixed oxide (spinel) increasing rapidly until a level of 2% Mg is reached in the alloy. At that approximate level, the




composition of the oxide reverts to pure magnesia, and the rate of formation increases. The oxide and mixed oxide films are continuous over the surface of the melt but can grow at rates that are multiples of the growth on solid metals. This growth is supported by the entrained alloy trapped in the filmy oxides and additionally by capillary action transporting fresh aluminum through the relatively porous layer of material. In addition to the growth of oxide films and dross on the melt surfaces that are induced by exposure to general atmospheric conditions, there are a number of factors encountered in industrial melters that can supplement and even alter the growth factors. These include turbulence induced by the mode of melting energy, the generation of atmospheres that are chemically more reactive, and even the addition of oxidants or contaminants with the melt stocks. These are illustrated in Fig. 1 and 2. Turbulence caused by direct impingement of burner flames on the melt can be an especially damaging condition in accelerating the rate of dross formation. The movement on the metal surface ruptures the oxide film, exposing a new surface to oxidizing gases at elevated temperature. These gases contain potentially high levels of moisture from entrained humidity or products of combustion, further adding to the oxidation potential. Similarly, the melt stocks themselves can be a source of dross and contaminants as they carry in their own layer of oxide and surface moisture that reacts to generate additional oxide, while also adding hydrogen to the melt. Further, many items of die cast home scrap and returns will carry in moisture and organic compounds from die lubes, machine oil and grease, paint, and so on. These add carbides and nitrides to the suspended inclusion matter in the metal and ultimately wind up in the dross. Finally, excessive melting temperatures, long holding periods, improper burner adjustment or electrical heating controls, worn elements, and deteriorating furnace doors that permit air

Dross, Melt Loss, and Fluxing of Light Alloy Melts / 993 aspiration into the furnace will contribute to increased oxidation at the melt surface and growth of the dross layer.

Influence of Melter Type on Dross Generation

Fig. 1

Complex interactions that occur between the atmosphere above the melt, the entrapment of useful aluminum alloy by the crumpled oxide films, and the resultant buoyancy of oxide inclusions in the melt due to the adsorption of hydrogen onto the complex surfaces of the inclusion

Fig. 2

Addition of supplementary volumes of contained surface oxides that are on the metal stocks entering the furnace molten metal. The presence of organic compounds will similarly react with the molten bath to add carbides and nitrides to the inclusion level. These suspended inclusions and oxides by themselves have a higher density than the molten aluminum but are carried to the dross layer by hydrogen gas adsorbed on the inclusion surface or gas bubbles generated by charge contaminants.

The various requirements for alloy composition, metal volumes, and pouring temperatures have created a very wide range in the size and types of melters and holders used in the aluminum shape-casting (foundry) industry. The very large aluminum casters make use of direct molten-metal deliveries from the secondary or primary refiners to minimize the need for re-melting cold stocks. It is not uncommon for these deliveries to cover distances of more than 160 km (100 miles) from the smelter plant, and they are delivered on a very tight schedule. The shipments are made in large refractory-lined ladles with sealed tops to retain the heat and, in some cases, are held at temperature in the receiving plant with auxiliary burners to supply molten alloy on an as-needed basis for the melting department. Over-the-road shipments of quantities as small as 4500 kg (10,000 lb) have been made, but a more common delivery quantity is approximately 7000 or 14,000 kg (15,000 or 30,000 lb). Typically, the size of the shipment is limited more by carrier and highway load limits than anything else. Surprisingly enough, tight specifications can be held on delivery temperature, time, and volume. Specification arrival temperatures of a total range of 40  C (75  F) are not uncommon, with arrivals scheduled within 30 min of a target delivery window. The use of molten deliveries will normally add a requirement for very large holding furnaces (with a minimum capacity of twice the delivery volume) at the receiving point to balance usage against deliveries and to minimize temperature surges and cold deliveries that can wreak havoc with process control. The elimination of the remelting operation does produce major benefits in terms of minimizing both dross generation and melting energy requirements. Further, since the holding furnaces are designed with smaller burners (merely holding metal at temperature requires less energy than melting), they also minimize the generation of dross on the hearth during the holding period. These savings are partially offset by higher transportation costs because of the unique carriers involved and their high maintenance costs, but the resultant savings in energy, labor, and melting loss more than compensate. With energy expenses constituting a larger portion of total melt-room costs, the savings associated with molten deliveries are becoming ever more important. The other major savings with molten deliveries are in the area of reduced melting losses. Reduced holding costs have already been noted but equally important is the minimization of melting loss associated with the lower volume

994 / Casting of Nonferrous Alloys of cold material being melted. With moltenmetal deliveries, the primary source of dross is that created by turbulence during the transfer. Most operations using hot metal deliveries experience less than 1% melting loss for the volume of material received. Those operations too remote from suppliers to use molten deliveries rely on primary melters in their own operations to convert their feedstocks to molten form. A variety of melter types are used in these operations, with the choice dictated by available capital, energy costs, melt demand, labor rates, and the types of charge materials that will be used or generated internally. The melter types in use include induction melters (both channel and coreless), resistance-heated furnaces (usually all wet-bath charge types), reverberatory furnaces (of wetbath, dry hearth, or combination charging modes), and stack melters (where the charge is preheated and dried before entering the melting zone) (see the article “Reverberatory and Stack Furnaces” in this Volume). All furnace types have their proponents, depending on selection factors mentioned before, and in general will sustain melting losses (dross generation) ranging from a low of 0.5% up to 4.0% or greater for similar charge stocks. There are also significant differences in melt quality generated by the various furnace types, which are addressed later in this section. For molten deliveries, the aluminum alloy caster does not sustain the melting loss on the original feedstocks, because that is absorbed at the refining plant (the caster’s only loss is usually that associated with the metal transfers and the minimal normal hearth losses). Among the other furnace types, the choice is usually made with primary consideration given to the size and surface area of the materials being melted. Dry hearth furnace types are usually chosen when the materials being charged have minimal surface area (large sows or full stacks of ingot). Wet-bath furnaces have advantages when the charge materials are mixed in section thickness, with the thinner materials being able to be quickly submerged to minimize additional surface oxidation. However, this benefit can be lost if there is any moisture or lubricant associated with the charge materials, because these contaminants are a major source of dross and dissolved hydrogen when added to molten aluminum. Stack melters have the advantage of energy efficiencies gained from preheat and drying of the charge materials before entering the melting zone but sustain somewhat higher melting losses when melting thin-gage stocks and flashings because of increased oxidation before reaching the melting zone. The other notable disadvantage of stack melters is that their best performance is developed with a mixed charge of heavy and light melt stocks. Gates and runners from large foundry or die castings cover a significant amount of surface area for their weight and will require size reduction at

the trim press or cutoff station and subsequent “bundling” to increase the charge density. Without this activity, the number of low-density charges will lead to increase labor expense, potentially higher melt loss, and a slowdown in the melting operation. An additional area of advantage for stack melters is that the holding hearth is usually sized at approximately 2 times the melting rate of the unit. Other volume-melting furnace types will frequently be sized at a multiple of 5 and even 10 times the melting rate. The larger volume in the hearth is associated with the need to minimize temperature fluctuations associated with the introduction of cold melt stocks to the molten volume. Side-by-side comparisons of oxidation/ melt losses have shown an advantage for stack melters when compared to dry hearth furnaces, but the types of charge materials used could have introduced some bias. Adherence to proper operating conditions is absolutely necessary to achieve low melt loss in stack melters. Resistance furnaces generally have lower melting losses because of the elimination of combustion atmosphere-related melt losses and the absence of any flame-induced turbulence on the surface of the melt. Again, there are trade-offs for these benefits, and they are normally associated with higher energy costs for power. Induction furnaces have historically been the preferred choice for scrap forms with very high surface areas (dried machine turnings or flashings, etc.) because of their rapid submergence of the materials. Normally, very little surface dross can be seen on some aluminum melts in induction furnaces. While it may not be visible, the intense stirring is normally viewed as a factor in breaking these oxides into smaller and smaller inclusions, which may then be suspended in the melts. The need to remove these with downstream cleaning will somewhat negate the low initial dross losses. While surface dross/metal loss may be low, the resultant residual oxides associated with fine-scale charges may reach 6 to 10% metal loss in the bulk metal when ultimate metal weight cast is compared with metal weight melted. Because of the high surface area associated with chips, borings, turnings, flash, and so on and the subsequent entrainment of the oxide surfaces with the melt, other forms of direct melting of these materials are coming into greater favor, especially those with mechanical or molten-metal pump-assisted devices designed for rapid submergence.

Influence of Charge Materials on Melt Loss As indicated earlier, the primary factors influencing melting loss, other than equipment or processing-related variables, are the surface-area-to-mass ratios of the charge materials and the condition of the surface itself. Large heavy-section melt stocks (i.e., sows or even

ingot shapes) have the largest ratio of mass to surface area and sustain proportionally minimal melt loss. In-house remelting of clean scrap, shot buttons, gates and runners, trim scrap, and so on will increase the melt loss because of the increased surface area. If the same materials carry moisture; sand or molding materials; metal working, finishing, or die-casting die; and release lubricants as they enter the melt, the loss could be doubled. A good rule of thumb is that for every 1% contaminant contained— organics and moisture—2% melt loss will result. Charging turnings, borings, and machining chips will usually result in as much as 10 to 15% dross formation or melt loss because of the high surface-area-to-volume ratio of such items. Other scrap forms, such as painted scrap or oily and greasy clean-out scrap from machines, are not currently being introduced into the melting furnaces of most foundries because of the downstream need for bag houses and other air-pollution-control equipment. These materials are processed in secondary refining plants that are equipped with such equipment and are often predried to remove the bulk of contaminants before introduction into molten metal. While adding cost to the recycling operation, the early removal of moisture, oil, and organic materials is recognized as essential to minimize associated melt losses and inefficiencies.

Influence of Operating Practices on Melt Loss The practices used in melting aluminum can have a major impact on the rate of dross generation and the resultant melting losses. With most melting being done in fuel-fired reverberatory furnaces, relatively small variations in a number of process parameters can have a large impact on the melting losses sustained. One of the first would be the adjustment of the fuel/ air ratios used in the burners, with strongly oxidizing conditions producing the obvious result of an increased rate of dross formation if there is any movement of the melt surface to expose fresh aluminum surfaces. A second and equally important factor is the holding temperature of the melt after melting has been achieved. The impact of temperature increases of only 25 to 55  C (50 or 100  F) can be quite significant, as may be seen in Fig. 3. It should be kept in mind that thermocouples in a molten aluminum bath actually measure the temperature of the bath itself. It is impossible to measure absolutely the temperature of just the top surface, which in reality may be considerably hotter than the bulk molten volume. Consequently, the temperatures shown in the graph of Fig. 3 can indeed be reached at the melt surface, with attendant rampant oxidation and resultant increase in dross thickness. Similarly, the thickness of the layer of dross over the melt can have an impact because of its insulating effect and the required increase in firing rate of the burner system to maintain

Dross, Melt Loss, and Fluxing of Light Alloy Melts / 995 process temperatures. The temperature in the dross layer rises, increasing the rate of oxidation of the metal being aspirated from the hearth volume, and can produce a logarithmic increase in melting losses. The impact of increases in the dross thickness on melt rate is shown in Fig. 4 and is recognized by most melting operators. For this reason, the thickness of dross on melter surfaces is most commonly maintained at less than 40 mm (1.5 in.). Typically, the temperature of the dross layer is approximately 50 to 85  C (100 to 150  F) higher than the melt temperature. Thus, high melt temperatures incur the risk of ignition of the drosses, with spontaneous burning of the aluminum metal (called thermiting), causing potentially catastrophic results. If allowed to continue, temperatures attained during these reactions will exceed 1650  C (3000  F) and can result in melting of furnace refractories and the total loss of the furnace. The industry preference for wet-bath charging of light-gage scraps can be the source of an additional incremental increase in melting losses if the drosses formed on the surface of the melt are subsequently exposed to direct burner flames. The surface drosses are a “soupy” mixture of oxides and entrained metal and can contain up to 80 or even 90% of useful alloy. Without prompt removal from the surface, or treatment to separate the metallics from

Fig. 3

Logarithmic increase in aluminum oxidation as the holding temperature is increased

the oxides and dirt, they will oxidize, consuming valuable metallics. Keys to reducing dross formation at the melt surface include not only careful control of charging practices and temperature control but also increasing opportunity for better heattransfer capability. Reducing the melt surface temperature allows a greater temperature difference between the heat source and the melt, resulting in greater heat transfer and hence faster melting practice. This is accomplished by convectional burners, by melt stirring, and, in larger furnaces, by application of molten-metal pumping. This allows heat to be more effectively transmitted to and throughout the bath from the radiation heat transfer provided by the heat source into the roof and upper sidewalls, which “radiate” heat into the melt. Dry hearth furnaces are immune to some of these problems because the surface oxides from the charge materials remain on the sloping hearth since the melted metal drains into the furnace. If wet or oily gates, runners, and shot buttons are placed on the dry hearth, they will not develop the same problems as if they were charged into a molten bath. They will, however, contribute a higher oxide load to the hearth, which can prevent complete drainback of the molten alloy and expose it to oxidation. It is for these reasons that the more efficient aluminum melting operations will emphasize skimming of the furnace hearths on a regular basis. Skimming should be conducted on an as-needed basis, usually once per shift, unless wet drosses are created. Recommended practices with the wetter drosses are to skim on a minimum of a 4 h schedule.

argon or nitrogen, and the latter is more common because of economics. Gas mixtures may also be used, usually adding an active halogen component—chlorine or fluorine compounds— in smaller amounts, normally only 5 to 10%, to the base inert gas. The purpose of this addition is to provide a better separation of solid particulate (i.e., aluminum oxide inclusions) from the liquid metal. Gas fluxing does not reduce dross; in fact, it increases dross! The principal function of gas fluxing is hydrogen removal, not dross treatment or recovery (see the article “Aluminum Fluxes and Fluxing Practice” in this Volume). In-Furnace Treatment with Solid Fluxes. Once formed on the surface of the melt, it is imperative that the dross be removed to maintain normal melt rates and to avoid further loss of the contained metallics. Methods used in the industry usually combine treatment of the dross on the hearth or charge well with routine maintenance on the sidewalls and bottom of the furnace. This is normally done by employing several different types and application of solid fluxes, which are chemical compounds and mixtures of chemical compounds designed to perform specific functions. Mixtures may include reactive ingredients as well as filler materials, which provide extension of volume and/or reduce cost. Table 1 lists several important chemical constituents that are included in many fluxes. The specific kinds of chemical fluxes include:

Furnace Fluxing

The three are commonly employed in the industry and vary in application, depending on the type and size of the melting unit. Cover fluxes are used, as the name implies, to cover the melt. These fluxes are comprised of salt compounds, which form eutectic mixtures that are liquid at normal aluminum melting temperatures. A typical base composition is 47.5% sodium chloride, 47.5% potassium chloride, and 5% sodium aluminum fluoride,

Fluxing is a word applied in a broad sense to a number of methods of treating a melt. It is important to understand what it means and where it is used. There are two types of fluxing that the metal caster commonly refers to in typical melting and handling operations. Gas fluxing is the application of a gas treatment to the melt. The gas may be inert, such as

Table 1 Typical fluxing compounds employed Compound

Fig. 4

Impact of a layer of hearth dross on the melt rate of a reverberatory melter

 Cover fluxes  Wall-cleaning fluxes  Drossing fluxes

Fluidizer (F) thickener (T)

Oxide surfactant

Chemically active

Exothermic

Gas released

Element added

AlF3 CaCl2 MgCl2 MnCl2 KF NaF NaCl KCl

F F F F F F F F

... ... ... ... ... ... ... ...

X ... ... X ... X ... ...

... ... ... ... ... ... ... ...

... ... ... ... ... ... ... ...

... ... ... ... ... Na ... ...

CaF2 Na3AlF6 Na2SiF6

T T T

X X X

... ... ...

... ... ...

... ... ...

... ... ...

KNO3

...

...

X

X

...

...

C2Cl6 K2CO3 Na2CO3

... ... ...

... ... ...

X X X

... ... ...

Cl2, AlCl3 CO2 CO2

... ... ...

K2TiF6 KBF4

... ...

... ...

X X

... ...

... ...

Ti B

996 / Casting of Nonferrous Alloys or cryolite. Covering fluxes become liquid on contact with the melt, help to agglomerate dirt and oxides, and provide some degree of protection of the melt against further oxidation and absorption of hydrogen. Wall-cleaning fluxes are used to help in softening the buildup on furnace walls to permit easier removal by mechanical cleaning. They are usually applied with a lance or flux gun or may simply be broadcast manually onto the walls. Slightly exothermic, these fluxes, under closed-door and high-fire conditions, increase the localized temperature of the oxide buildup and thereby soften or loosen the attached oxides, permitting easier mechanical removal and subsequent skimming. If the oxide buildup has been converted over time to the insidious form of aluminum oxide called corundum, it cannot be removed chemically. However, in its early stages sufficient loosening of the mildly adherent product can be accomplished to allow removal from furnace sidewalls with the combination of fluxing salts and hard work. Drossing Fluxes. Recognizing that a high percentage of useful metallics can be lost when wet drosses are removed from the furnace, many operations will attempt to reduce the metallic content by in-furnace treatment of the drosses with exothermic fluxes. These contain oxidants and fluorides to generate temperature and help promote a separation of the oxides

from the metallics. The exothermic reaction does consume aluminum values while generating heat and can be the source of a loss of up to 20% of the contained metallics in the dross, if the flux is applied too liberally. Drossing fluxes are designed to assist in separating the entrapped aluminum from the adherent oxide skin envelope, which surrounds the aluminum in the dross layer. The structure of dross can be likened to M&M candies (Mars, Inc.), as in Fig. 5, or a heavy mix of “corn flakes” of oxide films floating in a bath of milk. As discussed earlier, the nature of the dross is dependent on the materials being charged and the style of melting furnace. The fluxes used generally are reactive, which rely on good contact with the physical dross to achieve the separation, and the flux may contain an exothermic compound designed to add heat to accelerate the separation process. The application of heat and physical mixing increases the localized surface temperature and the fluidity and helps to loosen the loosely adherent oxide skin or shell, opening up this skin to allow the metallic aluminum to coalesce back into the bath. Usage practices for both types of flux vary widely in the industry, but all require agitation and intimate mixing of the flux and dross to achieve a good separation. The initial step is to spread the flux over the layer of dross. Application rates vary, but some flux manufacturers

recommend starting at approximately 1 kg/m2 (2 lbyd2) of hearth area. Following that uniform application, most operators rabble the dross and flux mixture to begin the separation. The furnace doors are then closed and the burners turned on high fire for a short period of time. The increase in temperature promotes flux reactions with the dross, improving the separation. Normally, the door is then reopened, and the operator then carefully rabbles, rakes, and chops through the dross layer with an appropriate tool to further mix the flux with the dross. The combined material is then removed from the furnace as quickly as possible to minimize thermal losses that can be quite excessive during the fluxing and cleaning operation. Care must be taken to allow the now-freed molten aluminum to drain from the dross/flux mixture to achieve best recovery of the entrapped metal. A typical operation is shown in Fig. 6. One should fully understand that the drossing flux process does not chemically reduce aluminum oxide; this can only be done electrolytically. Rather, correct application of the proper flux composition and application practice results in the physical separation or breaking apart of a weakly-bound oxide layer entrapping fresh aluminum, as depicted in Fig. 5, allowing the metallic liquid aluminum alloy droplets to coalesce and return to the main bath below the dross surface. The objective of any fluxing practice is to transform white dross, or metal-rich dross, into a black powdery dross that is low in metal content. Figures 7 and 8 exhibit this physical appearance comparison portraying the visual difference between untreated and treated dross, respectively. Metal-rich dross appears bright and shiny, certainly laden with metallic aluminum, whereas treated dross will appear duller and powdery. Typically, the aluminum content of untreated dross averages 85 to 90%. Initial fluxing treatment can recover approximately half of this amount, if done aggressively, rendering the remainder still economically and technically viable for secondary metal recovery in-house, with auxiliary equipment, or with an outside

Fig. 6 Fig. 5

Schematic view of dross shell and entrapped aluminum

dross

Furnace operator cleaning the sidewalls of the furnace and rabbling the flux into the hearth

Dross, Melt Loss, and Fluxing of Light Alloy Melts / 997

Fig. 9

Fig. 7

Typical range of dross fluxing aluminum recovery

Unfluxed white dross, which is aluminum-rich

Fig. 10

Pile of white dross collected after the turbulent fill of a transfer ladle of A380 alloy in a die casting plant. The aluminum content of the material is +95%.

of the same composition. Additionally, this returned metal that has already been paid for replaces the need for purchasing new metal units and minimizes the inventory cost of metal units that are in the recycling loop.

Drosses Generated in Metal Transfer

Fig. 8

Fluxed black powdery dross low in aluminum

dross processor. With either auxiliary in-house or outside dross processing, an additional 30 to 60% absolute metal content recovery is still possible. Figure 9 depicts the general recoveries and reduced metallic content in treated dross that may be achieved with aggressive furnace fluxing as a first step, followed by secondary fluxing treatment off-line in another dross recovery process. It should be noted that these furnace fluxing results are optimum percentages, achieved only with very careful practice. Recovery values may be decreased by 50% or more with “drag-

out” losses if good alloy is entrained in the granular dross being removed. Realization of the metal recovery in a drossfluxing or secondary process has a significant economic impact. Secondary processing may be done either with in-plant or external operations at recycling plants. External operations are far more costly than in-plant treatment because larger volumes of material require transportation, energy requirements are higher, and additional melting losses are incurred. Metal recovered is of the same alloy content and can be immediately returned to a furnace

An additional source of dross that can contribute significantly to the total metal losses in any operation that melts and transfers aluminum alloys are those drosses generated by turbulent metal transfer. When molten aluminum alloys are poured or pumped with a free fall of the metal stream, large quantities of air can be ingested into the falling stream as it enters the liquid bath. A comparable analogy is the generation of the “head” of foam that occurs when a glass of beer is poured. The white drosses generated by any turbulent transfer of metal can be quite significant and moreover are among the richest in contained aluminum alloy of all drosses generated in aluminum processing (Fig. 10). The entering stream of metal brings one layer of oxide to the pool but further ruptures the surface oxide on the waiting bath and generates new oxides at the turbulent point of entry. The problem is compounded by the fact that not all of the oxides generated in the turbulent transfer are collected in the dross layer on the surface of the melt. Often, the smaller oxides have not had enough time to coalesce into the dross layer and may be contained in the molten

998 / Casting of Nonferrous Alloys metal and carried downstream to the finishing operations that make cast or wrought product. There, they can be the source of lower mechanical properties, porosity in the final product, or even grow into hard spots with detrimental and costly losses in machining operations.

Economic Implications of Dross Treatment of the drosses found on any aluminum melt surface varies widely from operation to operation. Some foundries perform a very careful fluxing and cleansing process on the material, and others merely skim the material into a convenient container provided by the local scrap dealer. These drosses constitute a large portion of the melting loss sustained by all operating aluminum melters and will typically be in the range of 1 to 2% of the weight of clean and dry material, but may be as high as 6 to 10% for some of the more contaminated charge materials. Thus, the drosses generated will vary widely from operation to operation. The most efficient aluminum melters ship out drosses for reprocessing with metal contents of 40 to 50% Al, while those shops doing a poorer job of fluxing and treatment of the drosses in-house (by far the majority) may be shipping out materials with metal content approaching 75 or even 85%. Most aluminum melters generate many thousands of pounds of drosses during the course of a year’s operation, and these metalrich materials leave their plants through a variety of channels. Regardless of the departure channel, the drosses and skimmings will typically wind up at a remote processor who will incur container and transportation charges, the costs to remelt and process the material, the need to landfill the reclamation salts used to improve the recovery processes, the unavoidable metal losses associated with the processing and contained in the salt wastes, and the transportation costs back to the generating source, which will reuse the recovered material in its operation. It is for these reasons and costs that the melt shop generating the original dross will receive compensation for the materials shipped out at rates that vary from 10 to 40% of the true value of the metallics leaving the plant.

In-Plant Enhancement or Recovery of Dross Metallics Historically, most in-house aluminum dross processing was rejected by the foundry as not being part of their core business. Indeed, foundry ingot suppliers maintained buy-back contracts to acquire drosses generated. Several methodologies have been employed by secondary smelters and dross processors whose primary business is to process dross and recover metal values. References 1 to 3 provide further information about out-of-foundry dross processing. Within the foundry, there have been

simple technologies employed to recover dross, especially the little used “dross buggy,” described subsequently. Today (2008), however, there are newer systems available so that the foundry can efficiently and economically process its own drosses and still provide desirable residues for the dedicated dross processor or the secondary smelter. Since aluminum dross can contain upward of 95% Al metal, interspersed in a “lacy” network of aluminum oxide skim, there is a strong incentive to be able to return a good portion of this contained aluminum metal, which is indeed recoverable. With the exception of some wet skimmings, most drosses will contain a percentage of white dross that has the potential to thermite (a reaction where the contained aluminum values burn in the presence of air), consuming any aluminum metal values that are left. Once the thermite reaction has been allowed to start, it can only be stopped by removing all sources of oxygen, cooling the reacting mass to a point where the thermite reaction is quenched, or separating the remaining aluminum, which is the fuel for the reaction. This is the basic approach adopted by several methods of in-plant enhancement/recovery. Dross coolers are not commonly used in the foundry or die casting industry but have seen wide application in other aluminum melting operations. The dross from large aluminum melters can be treated in place with the proper fluxing action and then raked off periodically from the bath surface for further processing. However, dross must not be allowed to cool merely by removing it from the furnace and placing it on the floor or in a dross pan. If this is done, the residual heat contained in the dross pile will actually increase in temperature and thermite. When this occurs, any metallic aluminum that has been entrapped in the dross layer will burn to an oxide and therefore be lost from possible secondary metal recovery. For this reason one of the ways such a reaction can be quenched is to use a chamber dross cooler. The chamber dross coolers are usually large, tightly shielded steel clamshell devices that can be plumbed to provide an argon atmosphere. The shielding is required to maintain a tight seal against a vacuum induced by the thermiting reaction, which will attempt to pull in additional oxygen to supply the reaction. The argon atmosphere keeps oxygen from the cooling dross and stops further burning. Dross is pulled from the generating furnace into the transfer pan directly from the furnace and is then placed in the dross cooler, and the chamber is sealed. A rapid transfer is required to minimize losses that will occur if the thermiting reaction is allowed to proceed unchecked. Temperature controls and argon flow can be set and monitored to determine the appropriate cycle time. When the dross has cooled, some metallic aluminum can be recovered in the bottom of the pan or may be drained through an open hole into a lower ingot mold.

Depending on the extent of the normal thermiting reaction, the recoveries at the secondary plant that recycles the uncooled dross may produce recovery gains of as much as 10 to 20% from the cooled mass. Dross Presses. Another method of external dross processing outside the furnace is to use a dross press. The dross is raked off the furnace while still hot and then transported (quickly) in a specially designed dross press mold. The mold is a heavy-walled steel bucket that is placed in the press, as shown in Fig. 11. A press head contoured to the shape of the bucket is then lowered and pressed with force into the dross mass. The action is much like a juice press but without the twist/rotation. The pressing action squeezes the recoverable metallic molten aluminum from the dross, and the metal drains through a small hole in the bottom of the dross transporter into an ingot or sow mold placed below the press. This sow metal can then be sent directly back to the generating furnace, because it has the same composition as when it was taken from the melting furnace. Typical total recoveries in this process are approximately 50%. That is, the metallic content of the dross as removed from the furnace is approximately 80 to 85% as-raked, and after processing in the dross press will produce the squeezed product and metal-dross skulls that have been quenched in the heavy steel molds. Recoveries with the equipment are definitely a function of temperature losses in the transfer, and the process is subject to freezeups with cold transfers of material to the press, although the pressing operation takes only a few minutes.

Fig. 11

Working area of a dross press, with segmented pressing head, transfer vessel, and recovery pan placed below the unit

Dross, Melt Loss, and Fluxing of Light Alloy Melts / 999 One of the secondary benefits of the process is that the steel molds provide a real quench function to extinguish any thermiting reactions that may be occurring in the dross mass. Some operators view this as the primary benefit from the pressing operation and can raise recoveries from the processed dross by as much as 5 to 10%. Dross Buggy. It was mentioned previously that removing the aluminum fuel from a mass of white dross was an effective way to stop the thermite reaction that frequently follows a cleaning operation on an aluminum melter. Several manufacturers have made equipment to accomplish this on a volume basis.

Fig. 12

Schematic of the operating positions of the unit pictured in Fig. 13. In this application, all rotary movement of the reaction vessel is done manually, with only the mixing functions automated and timed. Courtesy of Q.C. Designs, Inc.

One type of equipment required to promote the separation includes an insulated ladle that must be preheated and used at the melting furnace to collect the dross as it is raked from the furnace. The ladle is then transported to the mixing station, where an additional flux addition is made, and the hot mass is stirred by a large paddle. As the metal content separates, it either drains into a collector pan below the ladle or is allowed to pool in the bottom of the ladle and then may be tapped into a sow mold. The recovered metal can be recycled back into the melt furnace, since it will normally be of the same composition as the original melt. This technique is quite simple; however, many metal casters are moving toward more advanced techniques. Dross Boss (Q.C. Designs, Inc.). In recent years, some modifications and adaptations of the concepts from the dross buggy have resulted in a new group of in-plant dross processors designed specifically to serve aluminum melting furnaces generating small to midsized volumes of dross. Process equipment is available to recover a high percentage of the contained alloy from quantities as small as 4.5 kg (10 lb) of hot dross and as large as 225 kg (500 lb). All sizes are available with varying levels of automation, with some smaller units capable of being manually stirred. As with the dross buggy, the new equipment family can only be used with hot dross transferred directly from the melting furnace. The specially designed reaction vessels do not require preheating and will accept a direct transfer of dross from the melter or processing station. The valuable metallics are recovered

after a small addition of exothermic flux is made to the dross mass, which is then stirred to develop a gravity separation of the materials. When the reaction and separation are complete, the recovered metal units are drained into either a collection ingot pan or transferred directly back to the melter in molten form. Either transfer method provides a major benefit in energy savings to minimize the cost of remelting the recovered material and associated melting losses, with 40 to 45% savings if transferred back at 425  C (800  F) and a total savings with molten transfers. While recoveries of up to 80% of the contained aluminum values have been achieved, a more normal recovery will be in the 60 to 70% range of the weight of charged dross and flux. The residual demetallized material will typically have between 25 and 35% contained aluminum and still have enough value to make it a desirable material for recycling and use in the chemical or construction industry. The midsized unit shown schematically in Fig. 12 and pictured in Fig. 13 will handle 70 kg (160 lb) of dross from a reverberatory well and is cycled hourly with approximately 45 kg (100 lb) of dross from a chip-melter installation. Smaller forms of systems are also in service to process the crucible skimmings generated by flux injection/degassing of transfer ladles in sand, permanent mold, and die casting operations. The automatic unit pictured schematically in Fig. 14 will accommodate a maximum charge of 25 kg (60 lb) of crucible skimmings, while recovering +50% of the weight of material transferred as good alloy. A partially automated unit of smaller size recovers more than 3.2 kg (7 lb) of metal per cycle from an average of more than 100 ladles processed per day. The largest number of applications for this equipment are found in service of wet-well

Fig. 14

Fig. 13

Dross Boss unit with 70 kg (160 lb) of hot dross capacity serving a well on a reverberatory melter. The average recovery of metallics is +50% on the weight of dross and flux charged, and the recovered metal is returned in molten form to the well.

Schematic view of an automatic skimming station process unit. The operator is required only to seal the drain hole and transfer the hot dross to the reaction vessel. All other operations of mixing, draining the recovered metal, dumping the demetallized dross, and cleaning the reaction vessel are controlled by a programmed logic controller unit. Courtesy of Q.C. Designs, Inc.

1000 / Casting of Nonferrous Alloys furnaces, where both the hearth and well drosses can be combined to yield quantities of dross that can be conveniently transferred to the process units.

Ways to Reduce Dross Formation The best way to initially deal with dross formation is to minimize it. This can be accomplished by a number of common-sense and conscientious maintenance procedures. The following list enumerates many, but probably not all, of the useful means to minimize dross formation that aluminum casting foundries can employ:  Use clean, dry charge materials.  Use charge materials with high mass-

to-surface-area ratios to minimize the amount of oxide introduced into the furnace.  Purchase molten alloy when feasible.

 Use fluxes to cover the melt and reduce        

oxidation. Minimize melt cycles by stirring and pumping to increase melting rates. Keep furnaces covered as much as possible, especially when holding metal, to reduce oxidation. Keep burner flames from direct impingement on the molten metal. Ensure proper fuel/air ratios in burners to minimize oxidizing conditions. Hold molten alloy at as low a temperature as possible. Transfer molten alloy as little as possible, and minimize turbulence or cascading during tap-out or pumped metal transfers. Employ the use of a dross reclamation system. Coat tools with an appropriate nonwetting wash.

 Keep furnace doors in good condition.  Maintain positive pressure in reverberatory

melting furnaces.

 Ensure that thermocouples are in good work-

ing order and properly calibrated to correctly monitor/control desired temperatures.

REFERENCES 1. V. Kevorkijan, Evaluating the Aluminum Content of Pressed Dross, JOM Feb 2002, p 34 2. D. Groteke and D. Neff, “A Guide to Reducing and Treating Aluminum Dross,” NADCA Publication 526, 2006 3. R. Peterson and L. Newton, Review of Aluminum Dross Processing, Light Met., 2002, p 1029

ASM Handbook, Volume 15: Casting ASM Handbook Committee, p 1001-1008 DOI: 10.1361/asmhba0005286

Copyright © 2008 ASM International® All rights reserved. www.asminternational.org

Aluminum Alloy Ingot Casting and Continuous Processes Elwin Rooy, Elwin Rooy & Associates

INGOT CASTING is the vital conduit between molten metal provided by primary production and recycling and the manufacture of aluminum and aluminum alloy products. Extrusions, forgings, sheet, plate, and foil begin as billet and fabricating ingot. Sand, permanent mold, investment, and pressure die castings typically originate in alloyed remelt ingot. A number of ingot casting processes have been developed to assure the soundness, integrity, and homogeneity required by these downstream manufacturing processes. The importance of ingot casting technology to subsequent wrought and shape casting processes is measured by the wide range of material properties and characteristics that can be consistently and reliably developed in these products.

process but must be scalped for indirect extrusion. Ingots for Rolling. Terminologies differ but ingots cast to provide starting stock for sheet, plate, and foil are called sheet ingots or rolling ingots and are of essentially rectangular cross section (Fig. 3). Sizes vary widely and are controlled by coil size considerations and hot mill widths and break-down mill openings. Maximum ingot dimensions are approximately 66 cm (26 in.) in thickness, 200 cm (80 in.) in width, and up to 915 cm (360 in.) in length.

Ingot Forms Remelt Ingot. The metal castings industry requires ingot (Fig. 1) that can be remelted to provide specification-compliant, on-composition molten metal and alloys. Sizes vary according to the volume of operations and the furnace sizes and types that are used for remelting. Typical ingot sizes are 13.5, 22.5, 315, 675, and 990 kg (30, 50, 700, 1500, and 2200 lb). Shape is usually determined by the ingot supplier, but there is general industry conformity. Dimensions, especially thickness relative to width and length, and casting practices are designed to minimize or avoid surface cracks and open shrinkage voids. These and other defects that permit water contamination and retention represent significant safety risks in furnace charging and melting. Ingots up to 22 kg (50 lb) are often notched so that pieces can be broken to facilitate charging and more accurately top off crucible furnace charges. Billets (Fig. 2) are round, cross-sectional lengths used extensively for extrusion. Sizes conform to extrusion press container dimensions, which typically range from 15 to 28 cm (6 to 11 in.) in diameter. Billets with as-cast surfaces are acceptable in the direct extrusion

Fabricating Ingot. Most often consumed in forging, fabricating ingot may be of rectangular or circular cross section. In either case, initial fabricating steps usually involve preworking of the ingot by a series of upsets and draws to assure thoroughly wrought metallurgical structures before performing final hand (open die) or die forging. The use of extruded stock from billet and plate from rolling ingot is also common. Round cross-sectional ingots for forging range in diameters up to 97 cm (38 in.). Fabricating ingot may also be the starting form for rolled aluminum bar and rod and for rolled and drawn wire. Particle Ingot and Powder. Granulated, pebbled, and flake forms are used in chemical processes as reactants and catalysts. Aluminum shot is used for deoxidation in steel production. Atomized aluminum powder is used in applications that include paint, roofing materials, explosives, and powder metallurgy parts.

Molten-Metal Processing Molten metal must be composition controlled and processed before ingots are cast. Alloying, degassing to reduce dissolved hydrogen, treatment to separate and remove nonmetallic contamination, grain refinement, and, in some cases, the addition of elements influencing

Fig. 1

Ingot for foundry remelting. (a) Automatic ingot stacker at casthouse. (b) Stack of aluminum ingot at die casting foundry. Courtesy of Light Metal Age

Fig. 2

Vertical direct chill billet casting machine. Courtesy of Light Metal Age, Dec 1998

1002 / Casting of Nonferrous Alloys

Fig. 3

Rolling ingot from direct chill casters with mold frame in background. Courtesy of Loma Machine

metallurgical structure are all within the province of ingot cast shop operations. The principles of degassing, flux treatment, filtration, and grain refinement have been covered in the Section “Molten-Metal Processing and Handling” in this Volume. Because of the scale and generally common design of most ingot casting facilities, a degree of commonality in process types for addressing the treatment of molten metal has evolved. Large reverberatory furnaces up to 135 Mg (300,000 lb) capacity are used for melting and/or holding. Alloying is accomplished through the addition and solution of metallurgical metals and master alloys. Compositions are confirmed by chemical analysis. In-furnace treatments for adjusting magnesium concentration, separating entrained oxides and other nonmetallics, and reducing dissolved hydrogen levels include the use of solid and gaseous fluxes. In-line processes for the efficient removal of dissolved hydrogen and included matter are almost universally in use. Rotary degassers capable of reducing sparging gas bubble sizes for the most efficient hydrogen removal are preferred over diffusers or other systems. Even micrometer-sized nonmetallics may be removed by the most efficient molten-metal filters, which include deep-bed, rigid media, and porous ceramic designs. Additions of grain refiners are also typically made in-line using rod feeders that introduce highly refined 9.5 mm (3=8 in.) diameter master alloys of aluminum, titanium, titanium-boron, or carbides.

Ingot Casting Processes Nearly all ingots in the form of billet, sheet, and fabricating ingot are produced by the direct chill process or one of its many variations. These process variations are also used to produce a percentage of remelt ingot, including T-cross-sectional lengths as an alternative to open-mold sow, but open-mold casting of these products is more common. Firms specializing in sophisticated complex mold tooling have contributed to standardization across the industry, but specialization based on proprietary designs, developments, and practices continues to complicate discussions of ingot casting process technology. Process principles are less controversial and serve the purpose of this discussion.

Open-Mold Casting Alloyed and unalloyed remelt ingots are predominantly cast in open steel or cast iron molds. Larger forms (>315 kg, or 700 lb), termed sows, are the international standard for unalloyed primary aluminum. Sow may be ladle poured directly from the reduction plant. The pouring practice can be automated by in-line and rotary conveyor systems. Smaller remelt ingot forms (14 to 23 kg, or 30 to 50 lb), termed pig, are more typically cast on conveyor systems. In either case, automation can extend from metering pouring devices to identification, stacking, and bundling after solidification.

Fig. 4

Vertical direct chill casting. (a) General arrangement of process. (b) Mold assembly. Source: Ref 1, 2

Direct Chill Process As originally developed, the direct chill (DC) process is the vertical discontinuous casting of ingot in water-cooled aluminum molds (Fig. 4). The lateral dimensions and shape of the ingot are formed by the mold, while the bottom or butt of the ingot solidifies against a closely fitting contoured bottom block positioned within the mold before molten metal is introduced. Molten metal flows into the cavity formed by the mold and bottom block through a pin or float-controlled spout. As solidification begins, the bottom block is withdrawn from the mold and continues to lower at controlled rates. In this way, cooling water flowing over the surface of the mold or through passages in the mold directly impinges on the ingot surface. Water-cooling patterns may be complex. In some cases, the mold is internally cooled through elaborate passageways and flow-control devices. In others, the mold is externally

Aluminum Alloy Ingot Casting and Continuous Processes / 1003 cooled through water impingement. The pattern of water contact with the emerging ingot can also be manipulated to increase or decrease heat flux as a function of the ingot geometry. Sheet ingot molds are contoured to generate parallel rolling faces in the desired ingot thickness under dynamic steady-state conditions for a given casting speed. However, the solidification that occurs at the start of casting is analogous to that of the permanent mold process for engineered castings, so that the bottom of the ingot is thicker (butt swell) than desired. At the same time, stresses generated as casting progresses deform the ingot butt ends upward, a condition referred to as butt curl. Both conditions can be minimized by moderating the severity of water cooling at the start of the cast and by providing insulation over the center of the bottom block, which results in the retention of higher postsolidification metal temperature. Control of the water-cooling rate has been achieved through controlled variations in dissolved carbon dioxide. Additional developments affecting degrees of butt swell and butt curl are contoured bottom blocks and molds whose convexity can be dynamically altered during the cast. Mold lubrication is required to prevent attack or adherence by the metal being cast. Manually applied greases have been largely replaced by injecting selected oils through a network of small-diameter holes in the upper area of the mold or through grooved rings. A DC casting complex comprises a mold table on which the mold is mounted and through which water is directed to the mold, a platen-mounted bottom block, and a lined casting pit. Maximum length is determined by the depth of the casting pit and the mechanism by which the bottom block movement is controlled. In most cases, the bottom block, attached to a platen, is raised and lowered hydraulically, although cable systems continue to be in limited use. When the desired ingot length is reached, the mold table is tilted or slid away so that the solidified ingot can be removed from the casting pit (Fig. 2, 5b). Removal is usually by overhead cranes with sling or tong attachments. When the ingot has been removed from the casting pit, the mold table is repositioned, and the bottom block is raised into the mold for the next casting cycle. The use of multiple molds and bottom blocks is normal for all production DC units. Large sheet ingot may be cast seven or more at a time, and the largest billet stations are equipped with tooling for multiples exceeding 100.

DC Process Variations Level-Pour DC (FDC) Process with Header. Mold designs, following the successful development of the DC process, include mounting an insulating header above the water-cooled aluminum mold (Fig. 6). Its purpose is to increase aluminostatic pressure and

to promote thermal differentiation during solidification. A corollary benefit is the ability to introduce molten metal to the mold under level-pour conditions through the insulating body rather than through downspouts. The insulating header and mold are separated by metal rings that provide passage for the mold lubricant. The oil ring arrangement also minimizes heat flux between the insulating header and water-cooled mold. The vertical FDC process is most extensively used in the production of billet and large-diameter fabricating ingot. Hollow Round FDC. Hollow round ingot, commonly used for the production of extruded tube, is cast using a sequence of cores lowered with the ingot or by a water-cooled mold piece that forms the inside diameter. Air-Injected FDC. An increase in the overhang of the insulating header and the circumferential injection of air at the header-mold interface effectively reduce mold contact. The result is improved surface quality and a reduction in surface and subsurface disturbances in metallurgical structure. Originally developed in Japan, the process and its variations have had profound influences on billet production. Horizontal Direct Chill (HDC) Casting. The discontinuous nature of DC and FDC casting and the cost of deep casting pits inevitably led to development efforts for the horizontal casting of ingot using DC process principles (Ref 1). In concept, a headered mold turned 90 from vertical is fed with processed alloyed molten metal through a refractory or ceramic insulating panel. A conventional bottom block is used to start casting, but thereafter, ingot is drawn from the mold at controlled rates by a conveyor system. Clamps and a flying saw cut ingot to desired lengths on the conveyor (Fig. 7). The first successful use of the HDC process was the production of multiple-extrusion billet strands (Fig. 8a). Some time later, the process was adapted to produce foundry remelt ingot of extraordinary density and quality relative to open-mold ingot (Fig. 8b). T-bar in weights corresponding to remelt sow is also cast by the HDC process (Fig. 9), although T-bar ingot is more typically produced by the DC process. The ultimate achievement in HDC was its adaptation for full cross-sectional sheet ingot. A significant concern in this development was the gravity-induced variation in metallurgical structures across the thickness dimension. While these variations would result in unacceptable product characteristics in certain alloys and products, extensive industrial experience and testing have proven the process highly advantageous for many common alloy sheet products, with no appreciable effects on product performance. Electromagnetic casting provides ingot surfaces that permit the as-cast application of sheet ingot in hot rolling or reduce scalping requirements. An induction coil shaped to provide the desired ingot cross section provides an

Fig. 5

Direct chill caster with (a) mold frame in position and (b) mold frame partially tilted up. Courtesy of Light Metal Age, Feb 1995

Fig. 6

Insulating header above the water-cooled mold. Source: Ref 2

electromagnetic field through which molten metal passes. Eddy currents induced in the molten column are strongly repulsed by the electromagnetic field. As a result, molten metal is suspended and lowered into the cooling water curtain without mold contact. While a nonmagnetic mold or shroud may still be employed for safety reasons, there is normally no mold contact, and solidification conditions that affect conventional DC process surface and subsurface structures are avoided. Ingot surfaces are typically smooth, and alternating segregation bands (Fig. 10) and oxide folds are avoided.

Particle Processes Aluminum and its alloys can be valuable as reactants and catalysts in chemical processes. Depending on the application, it may be desirable to maximize or minimize surface-tovolume ratio. For this reason, processes have been developed to produce a variety of particle

1004 / Casting of Nonferrous Alloys

Fig. 8

Horizontal direct chill caster. (a) Four-strand billet caster. (b) Twenty-strand foundry alloy ingot caster. Courtesy of Light Metal Age, Feb 1995. Source: Ref 3

Fig. 7

Mold package for horizontal casting. Source: Ref 2

Fig. 10

Fig. 9

T-bar ingot-handling system for a continuous horizontal universal caster. Courtesy of Hertwich Engineering-Austria

ingot sizes and shapes. Most particle products are unalloyed, but purity can be crucially important. Many particle products are cast from refined metal of >99.99% purity. Most particle products are sorted by size and dried following production. Highly automated systems separate and package powder fractions. Pebbled ingot is formed by introducing droplets of molten aluminum into flowing water. The droplets can be formed in a vibrating mold wash-coated pan in which holes have been drilled. Droplet size is controlled by the

opening dimension in the distribution pan and its position relative to the water stream. Ovality is controlled by the cooling water stream velocity. Granulated Ingot. Granules are produced by the deposition of molten aluminum droplets on a moving water-saturated porous fiber belt. The granules are characterized by flat and rounded surfaces. Flake Ingot. Flakes of aluminum particles are produced by droplets of molten aluminum falling onto cooling water flowing over an

Characteristic alternating bands of direct chill ingot

inclined surface. An alternate process drops a stream of molten aluminum onto a rapidly revolving water-covered cylinder. Deoxidation Products. Particles used in steelmaking for deoxidation are called deox shot and are formed by pouring molten aluminum onto a rotating disk above a cooling water tank. The particles leaving the disk under centrifugal pressure may be directed into the water tank or pass through a screen that reduces particle size before quenching. Aluminum powder is produced by air-inspirated atomization or by the mechanical or pneumatic disruption of a molten aluminum stream.

Continuous Processes While the DC process and its variations dominate in the production of ingot for wrought

Aluminum Alloy Ingot Casting and Continuous Processes / 1005 product manufacture, the development of continuous processes (Ref 4) has resulted in significant production volume with equally significant advantages. Continuous processes provide commercial alternatives to conventional ingot casting. Among the more important incentives for continuous operation are:     

Reduce capital investment Improve recovery Eliminate process steps Conserve energy Promote product uniformity

Alternatives to conventional ingot for rolled products, sheet, and foil include strip and slab casting. Each offers production rates at significantly reduced capital cost when compared with the equipment, facilities, and mills required for conventional hot rolling. A number of intermediate processing steps are eliminated. Neither process requires scalping, which reduces throughput recovery and contributes to increased melt losses. No stress relief, homogenization, preheating for hot rolling, and reheating during hot rolling are required. These processing steps, not always mandatory for all products, involve repetitive energy-intense heating and cooling. There are no end-crop losses that are experienced in conventional hot rolling. Typically, edge trim is eliminated or greatly reduced. Because strip and slab casting are continuous, ingot structural differences associated with starting and stopping in vertical DC casting that represent nonsteady-state solidification conditions are avoided, and product uniformity is improved by sustained solidification under steady-state conditions. Twin-roll strip casting (Fig. 11) is a continuous process in which molten metal is introduced through refractory or ceramic spout tips into the gap between counterrotating watercooled rolls, producing strip at conventional reroll gages (3.18 to 11.4 mm, or 0.125 to 0.450 in.) that typically serve as stock for subsequent cold rolling or, in some cases, may be rolled in-line to finished sheet gages. Depending on size, design, and operational practices, the average annual capacity of a strip caster is approximately 18,000 Mg (40 million lb). Casting speeds range widely but average approximately 2.0 to 2.5 m/min (7 to 8 ft/ min). Strip cast stock is the principal source of aluminum foil. Rolls must be effectively coated to prevent adherence of the solidified strip. This is accomplished by automated spray applications. Solidified sheet is coiled directly and sheared to coil size without interruption. Widths are established by edge dams or other methods. Thickness is established by the roll gap. Molten metal contacts the rolls behind the roll nip so that solidification precedes the exit point and an element of extrusion is included in the solidification process. The amount of setback of the feed tip through which

Fig. 11

Typical arrangement of a twin-roll horizontal caster used in the aluminum industry. Source: Ref 5

Fig. 12

Block caster

molten metal flows into the roll gap is a significant and controllable process variable. Because of extremely high heat-extraction rates, strip casting is alloy-sensitive, and highly alloyed compositions or those that require specific microstructures may not be suitable for production by these processes. This is especially true for dispersion-hardened compositions in which severe centerline segregation is encountered. Slab Casting. Molten metal solidified between moving belts or blocks produces a slab of 10 to 45 mm (0.4 to 1.8 in.) thickness that is reduced in-line by high-torque, slow-speed rolling mills to produce either hot-band or finishgage sheet. Molten metal is introduced through refractory or ceramic feed tips. Thickness results from the positioning of upper and lower belts/blocks. Width is determined by the position of edge dams. As in the case of twin-roll casters, the process is continuous. Slab casters (Fig. 12) demonstrate production rates an order of magnitude greater

(180,000 Mg/yr, or 400 million lb/yr) than strip casters and, as opposed to strip casting, are relatively alloy-tolerant. Casting speeds approximate 4 m/min (13 ft/min). The metallurgical structure of the slab and the reroll gage product is similar to that of metal produced from the hot rolling of conventionally cast ingot. In the twin-belt system (Fig. 13), tension and elaborate guide and support rolls are used to maintain flatness and minimize distortion. Cooling is accomplished by arrays of cooling water applicators and removal devices located on the reverse sides of each belt. The twin-block caster employs interlocking precision-machined blocks that form the rolling surfaces. The blocks are water cooled and dried during the cycle time that blocks are not in contact with the slab. Despite the precision of the blocks, molten aluminum nevertheless penetrates the seams formed by adjacent blocks, and while penetration is limited, sheet cosmetics after rolling are affected.

1006 / Casting of Nonferrous Alloys cross section and belt from either the 12:00 or 3:00 positions, using level transfer in the former and spout control in the latter. The belt is held over the cavity through tension and guides for up to 235 of wheel travel. The solidified bar is stripped from the wheel cavity and guided into an in-line multistand mill to produce coilable rod at typical diameters of 9.5 to 12.7 mm (0.375 to 0.500 in.). For electrical conductance alloys, conductivity is established by composition and through control of mill entry temperature and the reduction/cooling sequence in the mill. Production rates can exceed 5700 kg/h (12,500 lb/h). Continuous Extrusion Processes. Concepts for the direct production of extrusions from either molten metal or compacted scrap have been explored and developed.

Fig. 13

Fig. 14

Twin-belt slab caster. Courtesy of Light Metal Age. Source: Ref 6

Wheel-belt caster

Comparisons with Conventional Product. In all cases, metallurgical structure results from alloy composition, solidification conditions, and the rate of postsolidification cooling. Process differences and resulting differences in metallurgical structure may be considered advantageous or disadvantageous, depending on the product and its application. Solidification conditions characteristic of strip casting result in structures that are more susceptible to centerline segregation. The degree of segregation can be moderated by adjustments in casting rate, cast thickness, tip setback, and cooling as well as through downstream thermal treatments, but it remains a factor in any comparison with DC cast product. By comparison, solidification rates in slab casting approximate those of conventional DC casting, resulting in similar properties. Crystallographic texture, which dominates anisotropy and formability, results from the relationship of relative amounts of hot and cold reductions in the forming of wrought products.

For this reason, there are significant differences between continuous strip and slab processes, in which the extent of hot reduction is necessarily limited, and sheet and foil originating in thick ingot that undergoes extensive hot working to the reroll gage. Wheel-Belt Processes. Rod for electrical applications is almost universally produced in continuous wheel-belt processes (Fig. 14), which generate a trapezoidal bar that is rolled in-line to rod dimensions in triaxial multistand mills. Alloys for mechanical applications may also be produced by specialized wheel-belt casting machines. There are a large number of caster designs and a wide range of production capabilities. Casting rates are controlled by wheel material and diameter, belt material, and cooling. Belt material and its length are associated with longer run-times. Not all casters are capable of producing high-strength, heat treatable alloys. Molten aluminum is typically introduced into the cavity formed by the water-cooled wheel

Solidification in the DC Process The conditions under which solidification occurs in the conventional DC process are complex and involve mold length, cooling water flow rates, method of mold cooling, water distribution, mold length, casting rate, ingot size, metal temperature, and alloy. Molten metal initially flowing into the space created by the mold and bottom block solidifies on contact. The thin shell of solidified aluminum on the mold surface contracts, creating an air gap between the shell and mold wall. The heat mass of molten metal is sufficient to remelt the solidified shell, causing molten metal to again flow against the mold surface. This liquation sequence is repetitive until the shell thickness increases sufficiently that the heat mass at the solidification front no longer prevents its continuous growth. The first metal to solidify is relatively pure compared to alloy composition, according to Hoopes law. Metal flowing into the air gap after remelting of the ingot shell is alloy-enriched, resulting in a region of reverse segregation. The result is, in addition to normally occurring macrosegregation across the ingot cross section, an ingot surface and subsurface affected by alternating alloy-depleted and alloy-enriched zones (Fig. 10) also containing oxide films. A stable, solidified shell is formed before exiting the mold, at which point thermal extraction is greatly accelerated by direct contact with the flow of cooling water. The DC process is capable of producing dense metallurgical structures. The depth of the liquid-solid zone at the solidification interface is limited by steep thermal gradients, and there is an unlimited source of thermally differentiated liquid metal for shrinkage compensation. The ratio of thermal gradient and rate of solid metal formation also favors fine dendrite arm spacing and enhances grain-refiner effectiveness. Inclusions and hydrogen porosity are avoided by in-line metal treatment and the control of turbulence during mold filling and subsequent molten-metal flow.

Aluminum Alloy Ingot Casting and Continuous Processes / 1007

Postsolidification Processes Rolling and fabricating ingot or billet may be thermally treated to reduce residual stresses, homogenize the microstructure, and remove the surface disturbed zone and dimensional irregularities from scalping. Stress Relief. Residual stress levels can be sufficient to cause failure by cracking or splitting during casting. Failure results when residual stresses created by the differences in postsolidification cooling rates across the ingot cross section exceed the strength of the alloy at the temperature where failure occurs. The tendency for such failure is increased by the elevated-temperature strength of the alloy, ingot thickness, and severity of cooling. One of the techniques developed to reduce the risk of failure during casting is the removal of cooling water from the ingot surface at a point below the depth of the molten crater, which allows reheating and promotes the homogeneity of temperatures across the ingot thickness. For many thick or large-diameter highstrength alloys, postsolidification stress relief through furnace treatments approximating annealing is necessary. Homogenization. Depending on composition, ingot may also be homogenized using long, higher-temperature heating cycles and controlled cooling from homogenization temperature. The intent is the uniform distribution of soluble phases across grains and the control of the morphology of certain insoluble phases and other effects based on the limited diffusion of insoluble phases. Principles are those applicable to the elimination or minimization of microsegregation by diffusion. Homogenization may be incorporated in preheating procedures for fabricating operations. Scalping. As-cast ingot surfaces are typically characterized by surface and subsurface variations in structure and chemistry, by dimensional variations, and, in some cases, by surface defects such as oxides, oxide/salt patches, and shallow cracks. Scalping is performed to remove the disturbed zone—defined as the surface layer comprised of alternate segregation and oxide folds—the oxide surface formed during solidification, surface defects, and to provide dimensional regularity. The only significant exception is the use of billet with as-cast surfaces in direct extrusion. In direct extrusion, the surface of the billet is retained in the press at the end of the extrusion cycle. Electromagnetically cast ingot is still typically scalped, but scalping depth is sharply reduced. Removal of the oxide that forms at the surface before and during solidification improves sheet/plate finish and reduces hot mill roll coating.

charging, this discussion of safety is limited to ingot casting. It is not intended that the discussion be comprehensive nor form the exclusive basis for safe practices. Every aluminum ingot casting facility involves molten aluminum and large quantities of water in its operations. Conditions may result in the risk of explosions, ranging from minor to violent reactions, affecting equipment, facilities, and the safety of personnel. The most serious safety concerns are related to the DC process and its variations. There are a number of ways molten aluminum may become exposed to the cooling water. Bleedouts occur when the ingot shell melts below the bottom of the mold. This could result from improper mold cooling, excessive metal temperature, or excessive casting rate. Ingot may adhere to the mold wall, resulting in a hangup. Lubrication failure and mold condition are probable factors. When the solidified body of metal falls or when it is forced from the mold, molten aluminum may flow directly into the casting pit. Equipment malfunctions can contribute to these and other unsafe conditions. Rules and practices have been developed to minimize these risks. Most are derived from extensive testing in which quantities of molten aluminum were intentionally introduced into water-filled containers. Testing permitted the assessment of the influence of metal quantities, addition rates, composition and temperature, water temperature and cleanliness, drop height, stream dimensions, and container materials and coatings. The analysis of large numbers of these tests resulted in conclusions regarding variables that either promoted or diminished the probability of explosion occurrence. A generalized summary of conclusions is instructive:  Explosions were not experienced if metal

stream diameter was less than 70 mm (23/4 in.).

 Violent explosions are most likely to occur 

     

Safety



While there are substantial concerns over the safety of molten-metal transport and furnace



when molten aluminum flows into water at depths between 50 and 760 mm (2 and 30 in.). Explosions were more likely to occur when drop height ranged from 38 to 1220 mm (1½ to 48 in.). No explosions occurred with drop heights exceeding 3 m (10 ft). Explosion tendencies are inversely related to water temperature. Explosion tendencies increase with metal temperature. Explosion tendencies are reduced when water is oil-contaminated. Oxidized iron or steel surfaces promote explosions. The presence of hydroxides on container surfaces promotes the occurrence of explosions. The presence of lithium, lead, and bismuth in the alloy increases the risk of explosions. Container walls with inorganic coatings reduce the risk of explosions. Any form of impact concurrent with the introduction of molten aluminum into the

water-filled container increases the risk of explosions.  Explosions do not occur (with the exception of lithium-containing alloys) unless liquid metal contacts the container surface. Ingot casting operations sensitive to these conclusions have adopted practices that reduce the probability of explosions. Casting pit walls, platens, and bottom block assemblies are coated with selected inorganic coatings and/or compounds. Many of these that are based on petroleum or coal tars are no longer used for environmental reasons. Current recommendations can be obtained from The Aluminum Association, which has sponsored programs for improving ingot casting safety and also summarizes explosion-related incidents for the aluminum industry. Coatings are routinely inspected and repaired. Pumps removing water from the casting pit are operated to maintain a minimum water depth of 914 mm (36 in.) (over debris that may collect there). Debris is routinely removed. Casting crews are trained to recognize the development of unsafe conditions during the cast and to use appropriate abortive procedures. Fail-safe engineering solutions involving cooling water supply and cylinder operation have been designed and implemented. REFERENCES 1. C.M. Adam, Overview of D.C. Casting, Proceedings of the 1980 Conference on Aluminum-Lithium Alloys, The Metallurgical Society, 1981, p 39–48 2. “Alcoa Casting Systems,” Report 22, Alcoa Technology Marketing Division, Oct 1974 3. H. Zeillinger and A. Beevis, Universal Continuous Horizontal Caster for Ingot, Billet, or Bar, Light Met. Age, June 1997, p 20 4. R.E. Spear and K.J. Brondyke, Continuous Casting of Aluminum, J. Met., April 1971 5. B.Q. Li, Producing Thin Strip by Twin-Roll Casting, Part I: Process Aspects and Quality Issues, J. Met., May 1995, p 29 6. P. Regan, Recent Advances in Aluminum Strip Casting and Continuous Rolling Technology—Implications for Aluminum Can Body Sheet Production, Light Met. Age, Feb 1992, p 58 SELECTED REFERENCES  M.G. Chu and J.E. Jacoby, Macrosegregation

Characteristics of Commercial Size Aluminum Alloy Ingot Cast by the Direct Chill Method, Light Metals 1990, TMS, 1990  D.P. Cook, B.Q. Li, and J.W. Evans, Electromagnetic Casting: Mathematical and Physical Models, Light Metals 1989, TMS, 1989  J.M. Drezet and M. Plata, Thermomechanical Effects During Direct Chill and Electromagnetic Casting of Aluminum Alloys, Part I, Light Metals 1995, TMS, 1995

1008 / Casting of Nonferrous Alloys  S.G. Epstein, Causes and Prevention of Molten

 B. Frischnecht and K.P. Maiwald, Roll

 R. Sautebin and W. Haller, Industrial Appli-

Aluminum-Water Explosions, Proceedings of the Fourth International Aluminum Extrusion Technology Seminar, 1988  S.G. Epstein, Causes and Prevention of Molten Aluminum-Water Explosions, Light Metals 1991, TMS, 1991  J.P. Faunce, F.E. Wagstaff, and H. Shaw, New Casting Method for Improving Billet Quality (Air-Slip), Light Metals 1984, TMS, 1984  H.G. Fjaer and E.K. Jensen, Mathematical Modeling of Butt Curl Deformation of Sheet Ingots, Light Metals 1995, TMS, 1995

Caster Applications and Developments, Light Metals 1988, TMS, 1988 Guidelines for Handling Molten Aluminum, The Aluminum Association Y. Ishii, HDC Process for Small Diameter Ingot, Light Metals 1989, TMS, 1989 I. Jin, L.R. Morris, and H.D. Hunt, Centre Line Segregation in Twin Roll Cast Aluminum Alloy Slab, Light Metals 1982, TMS, 1982 A. Mo et al., Modelling the Surface Segregation Development in DC Casting, Light Metals 1994, TMS, 1994

cation of Electromagnetic Casting, Light Metals 1985, TMS, 1985  R.E. Spear and K.J. Brondyke, Continuous Casting of Aluminum, J. Met., April 1971  W. Szczpiorski and R. Szczpiorski, The Mechanical and Metallurgical Characteristics of Twin-Belt Cast Aluminum Strip, Light Metals 1991, TMS, 1991  R.P. Taleyarkhan, S.H. Kim, and C.L. Knaff, Fundamental Studies on Aluminum-Water Explosion Prevention in Direct Chill Casting Pits, Light Metals 2001, TMS, 2001

  



ASM Handbook, Volume 15: Casting ASM Handbook Committee, p 1009-1018 DOI: 10.1361/asmhba0005287

Copyright © 2008 ASM International® All rights reserved. www.asminternational.org

Aluminum Shape Casting ALUMINUM CASTING ALLOYS are the most versatile of all common foundry alloys and generally have the highest castability ratings. As casting materials, aluminum alloys have the following favorable characteristics:
Good fluidity for filling thin sections
Low melting point relative to those required

for many other metals


Rapid heat transfer from the molten alumi







num to the mold, providing shorter casting cycles Hydrogen is the only gas with appreciable solubility in aluminum and its alloys, and hydrogen solubility in aluminum can be readily controlled by processing methods. Many aluminum alloys are relatively free from hot-short cracking and tearing tendencies. Chemical stability Good as-cast surface finish with lustrous surfaces and little or no blemishes

Aluminum is one of the few metals that can be shape cast by essentially all existing processes, including pressure die casting, permanent mold, clay/water-bonded sand, chemically bonded sand, plaster mold, and investment casting. Important variations include molding and pattern distinctions such as lost foam (evaporative pattern), shell and V-mold, and process derivatives such as squeeze casting, lowpressure permanent mold, vacuum riserless casting, and semisolid forming based on rheocasting/ thixocasting principles. Of these casting methods, high-pressure die casting of aluminum is the dominant nonferrous casting process (Fig. 1). This article provides an overview on the common methods of aluminum shape casting. Details on melting and melt treatment are covered in the article “Dross, Melt Loss, and Fluxing of Light Alloy Melts” in this Volume. Structure control and grain refinement are detailed in the following articles in this Volume:

Melting and Melt Handling Aluminum and aluminum alloys can be melted in a variety of ways. Coreless and channel induction furnaces, crucible and openhearth reverberatory furnaces fired by natural gas or fuel oil, and electric resistance and electric radiation furnaces are all in routine use. The nature of the furnace charge is as varied and important as the choice of furnace type for metal casting operations. The furnace charge may range from prealloyed ingot of high quality to charges made up exclusively from low-grade scrap. Even under optimum melting and melt-holding conditions, molten aluminum is susceptible to three types of degradation:
With time at temperature, adsorption of

hydrogen results in increased dissolved hydrogen content up to an equilibrium value for the specific composition and temperature.
With time at temperature, oxidation of the melt occurs; in alloys containing magnesium, oxidation losses and the formation of complex oxides may not be self-limiting.
Transient elements characterized by low vapor pressure and high reactivity are reduced as a function of time at temperature; magnesium, sodium, calcium, and strontium, upon which mechanical properties directly or indirectly rely, are examples of elements that display transient characteristics.

Turbulence or agitation of the melt and increased holding temperature significantly increase the rate of hydrogen solution, oxidation, and transient element loss. The mechanical properties of aluminum alloys depend on casting soundness, which is strongly influenced by hydrogen porosity and entrained nonmetallic inclusions. Reductions in dissolved hydrogen content and in suspended included matter are normally accomplished through treatment of the melt before pouring, as discussed in the following articles in this Volume:
“Degassing”
“Molten-Metal Filtration”
“Aluminum Fluxes and Fluxing Practice”

Gravity Casting. Regardless of the types of melting and holding furnaces and the particular gravity casting process used, there is great concern for reducing or eliminating dissolved hydrogen and entrained oxides. These procedures are less frequently employed for pressure die casting, in which concerns are focused on the dominant process-related causes of casting unsoundness, namely, entrapped gas and pouring injection-associated inclusions. Sensitivity to melt quality varies with the casting process and part design and necessitates special consideration of relevant criteria for each application. In general, the melt is processed to achieve hydrogen reductions and


“Aluminum and Aluminum Alloy Castings”
“Modification of Aluminum-Silicon Alloys”
“Grain Refinement of Aluminum Casting

Alloys”


“Refinement of the Primary Silicon Phase in

Hypereutectic Aluminum-Silicon Alloys”

Fig. 1

Nonferrous casting processes by tons poured. Source: From data of AFS 2002 survey.

1010 / Casting of Nonferrous Alloys the removal of oxides to meet specific casting requirements. Modification and grain-refiner additions are made as appropriate to the given alloy and end product. Die Casting. Different melt preparation practices are employed in die casting operations because process-related conditions are more dominant in the control of product quality than those controlled by melt treatment. For this reason, degassing for the removal of hydrogen, grain refinement, and modification or silicon refinement in the case of hypereutectic silicon alloys are often intentionally neglected. The movement toward higher-integrity die castings has brought into focus the importance of the same melt-quality parameters established and used in the gravity casting of aluminum alloys. In high-production die casting operations, the consumption of internal and external scrap is of primary importance in reducing base metal costs for the predominantly secondary alloy compositions that are consumed. Scrap crushing, shredding, and pretreatment of various types precede melting, often in efficient induction systems. Oxides entrained in the melt as a result of this sequence of operations are dealt with through the use of salts and/or reactive gas fluxing. Melt treatment is typically confined to this and to rudimentary fluxing in holding furnaces to remove gross oxide and to facilitate the maintenance of minimum furnace cleanliness. A concern in die casting is the formation of complex intermetallics that are insoluble at melt-holding temperatures and/or precipitate under holding conditions or during transfer to and injection from the hot or cold chamber. These intermetallics (sludge) affect furnaces, transfer systems, and, by inclusion, the quality of the castings produced. Die casters are familiar with composition limits that prevent sludge formation. A common rule is that iron content plus two times manganese content plus three times chromium content should not exceed the sum of 1.7%. This limit is arbitrary and inexact, it is often assigned values from 1.5 through 1.9%, and it is subject to the specific composition and actual minimum process temperature. Pouring. It is critically important that the metal be drawn and poured according to the best manual or automatic procedures. These procedures avoid excessive turbulence, minimize oxide generation and entrainment, and limit regassing of hydrogen. Frequent skimming of the melt surface from which metal is drawn may be necessary to minimize oxide contamination in the ladle. Siphon ladles that fill from below the melt surface are used for these purposes, but most often, coated and preheated ladles of simple design are employed. The process of repetitive drawing and skimming inevitably degrades melt quality, and this necessitates reprocessing if required melt-quality limits are exceeded. Pouring should take place at the lowest position possible relative to the pouring basin or sprue opening. Once pouring is initiated, the sprue must be continuously filled to

minimize aspiration and to maintain the integrity of flow in gates and runners. Countergravity mold-filling methods inherently overcome most of the disadvantages of manual pouring. Proprietary casting processes based on low pressure, displacement, or pumping mechanisms may be considered optimum for preserving the processed quality of the melt through mold filling, but some important and relevant considerations apply. Melt processing by fluxing is more difficult in some cases because the crucible or metal source may be confined. If, as in the case of low-pressure casting, the passage for introducing metal into the mold is used repetitively, its inner surface becomes oxide contaminated and a source of casting inclusions. In other countergravity casting, mold intrusion into the melt and devices employed to displace or pump metal to the cavity may be the source of turbulence, moisture reactions, and the possibility of hydrogen regassing. Automated pouring systems are common in the die casting industry (see the article “Automation in High-Pressure Die Casting” in this Volume). Robotized ladle transfer, as well as metered pumping, may nevertheless incorporate features and reflect provisions to protect molten-metal quality through sound drawing, transfer, and pouring techniques. The same techniques have application in die casting operations in which these operations are performed manually. Hot chamber operation offers apparent metal transfer advantages over cold chamber operation. Recent developments in the use of siphons or vacuum legs to the cold chamber in pressure die casting offer new and interesting opportunities for upgrading the quality of the metal deliverable to the die cavity. Gating and Risering Principles. The methods for introducing metal into the casting cavity, for minimizing degradation in metal quality, and for minimizing the occurrence of shrinkage porosity in the solidifying casting differ among the various casting processes, primarily as a function of process limitations. However, the objectives and principles of gating and risering are universally applicable: Somewhat different techniques in gating and risering are used for different alloys. In general, riser size and the need for stronger thermal gradients increase with more difficult-to-cast alloys. Crack sensitivity or hot shortness forces compromises in the steps normally taken to achieve directional solidification. Extensive localized chilling may aggravate crack formation. In these cases, more uniform casting section thickness, larger fillets, more gradual section thickness changes, larger risers and, in some cases, riser insulation, and more graduated chilling offer the best prospects for success. In alloys that are more difficult to feed but are relatively insensitive to cracking at elevated temperature, establishing thermal gradients by selective chilling (and heating as in permanent mold casting) usually provides good casting results. Examples are alloys high in eutectic

content, as well as purer compositions, in which the solidification range is limited. In these cases, localized areas of shrinkage are inevitable in the absence of adequate thermal gradients accompanied by effective risering. Gating of Die Castings. Conventional pressure die casting uses gating principles that are different from those for gravity casting. Although the fundamental principles employed in gravity casting remain desirable, injection under high pressure at significant metal velocity precludes application of many of the rules that govern gravity casting processes. In pressure die casting, gates and runners are the means by which molten metal is transferred from the shot chamber or injection system to the die impression. The gating system must be designed to permit the attainment of cavity pressurization without reducing cycle time by its own mass and solidification time. It must also be of minimum size to maximize the gross-to-net-weight ratio and to minimize trimming and finishing costs. The objective of gating in die casting is filling of the die cavity by establishing uninterrupted frontal flow. This requires the prevention of excessive turbulence and mixing within the die cavity, and it minimizes the entrainment of air and volatiles derived from the injection system and from lubricants contained in the die cavity. Runners are usually semiellipsoidal and decrease in cross-sectional area in the direction of metal flow. Their respective cross-sectional areas should exceed corresponding gate dimensions. At no time should gate thickness exceed that of the casting. Gates are normally severely tapered at the entry point(s) to facilitate removal during trimming with minimal risk of damage to the casting. The greatest progress has been made in research dedicated to die design and the design of metal entry conditions, including sprue, runner, gating, and parameters of injection. Based on carefully analyzed applied research as well as modeling of the die-filling sequence, mathematical and instrumented programs are now available to die casters for the development, modification, and control of gating design and operation. Much less sophisticated mathematical relationships and nomographs have been employed for many years to determine the relative dimensions of plunger diameters, runners, gates, and process parameters, including fill rate and velocity, plunger velocity, gate velocity, and system pressures.

Casting Process Selection Selection of a casting process may depend on several different factors, such as:
Casting

process considerations: requirements for fluidity, resistance to hot tearing, minimization of shrinkage tendencies
Casting design considerations: draft, wall thickness, internal passages. (For example,

Aluminum Shape Casting / 1011












parts with undercuts and complex internal passageways can usually be made by sand, plaster, or investment casting but may be impractical or impossible to produce in permanent mold or pressure die casting.) Mechanical property requirements: strength and ductility, hardness, fatigue strength, toughness, impact strength, specification limits Physical property requirements: electrical and thermal conductivity, specific gravity, expansion characteristics Process requirements: machinability, brazeability, weldability, impregnation, and chemical finishing Service requirements: pressure tightness, corrosion resistance, wear resistance, elevated-temperature strength, dimensional and thermal stability Economics: volume, productivity, process yield, material costs, tooling costs, cost of machining, welding, and heat treatment

In most cases, dimensions, design features, and material property requirements limit the range of candidate processes. Typical design details for aluminum castings are shown in Fig. 2.

Fig. 2

Dimensional tolerances are compared in Tables 1 to 3, for different shape casting processes. Table 4 compares general characteristics of sand casting, permanent mold casting, and die casting of aluminum. Although aluminum alloys can be produced by any of the available casting methods, highpressure die casting is the dominant method (Fig. 1). This process is very common to nonferrous metals with lower-temperature melting points. It is ideally suited for high production rate and volume production of dimensionally accurate parts with excellent surface finish. One of the important reasons for the success of die castings has been the development of high-speed precision equipment. Another is the extension of die casting technologies to larger castings with heavier wall thicknesses. When two or more casting methods are feasible for a given part, the method used very often is dictated by costs. Quality factors are also important in the selection of a casting process. When applied to castings, the term quality refers to both degree of soundness (freedom from porosity, cracking, and surface imperfections) and levels of mechanical properties (strength and ductility). The following

tabulation presents characteristic ranges of cooling rate for the various casting processes: Casting processes

Cooling rate 

C/s



F/s

Dendrite arm spacing mm

in

Plaster, 0.05–0.2 0.09–0.36 0.1–1 0.004–0.04 dry sand Green sand, 0.1–0.5 0.18–0.9 0.05–0.5 0.002–0.02 shell Permanent 0.3–1 0.54–1.8 0.03–0.07 0.001–0.003 mold Die 50–500 90–900 0.005–0.015 0.002–0.0006 Continuous 0.5–2 0.9–3.6 0.03–0.07 0.001–0.003

However, it should be kept in mind that in die casting, although cooling rates are very high, air tends to be trapped in the casting, which gives rise to appreciable amounts of porosity at the center. Extensive research has been conducted to find ways of reducing such porosity; however, it is difficult if not impossible to eliminate completely, and die castings often are lower in strength than low-pressure or gravity-fed permanent mold castings, which are more sound in spite of slower cooling. Expendable Mold Methods. Sand casting is the major type of expendable mold method. The advantages of typical sand casting are

Suggested design details for aluminum alloys. Recommended dimensions are averages, and the use of either larger or smaller numerical factors may result in more difficult casting or defects. Source Van Hoen, Vol. III.

1012 / Casting of Nonferrous Alloys Table 1

Dimensional limits and tolerances for aluminum castings produced by different processes

Sand castings

Diameter of cored holes: Tolerance  þ0.75 mm (0.030 in.). If cored hole is to be used for clearance and a tolerance of þ0.75 mm (0.030 in.) is necessary, hole should be ordered  undersized and machined to tolerance. Location of cored holes: þ1.25 mm (0.050 in.) approximate minimum for large-volume  acceptance (þ 0.38 mm, or 0.015 in.) for precision sand castings); can be obtained either by direct tolerance or by reducing hole size correspondingly. If more than one diameter is located from a common centerline, a concentricity tolerance for all diameters can be maintained within  þ0.75 mm (þ þ0.015 in., for precision sand 0.030 in.) (þ 0.38 mm, or  method), except as noted under “diameter of cored holes.” A tolerance of þ0.25 mm  (0.010 in.) (commonly specified) cannot be met consistently. Straightness: þ0.75 mm (0.030 in.) per 25 mm (1 in.) of length. Min of 0.13 mm (0.005 in.)  for precision sand castings. Allowance for machining: 1.8 mm (0.070 in.): on long, thin castings allowance should be increased to 3.18 mm (0.125 in.) Parallelism: Same as under “Straightness” Permanent mold castings

Maximum length of core supported at one end: 10  core diam Draft on outer surfaces: 1 min, 3 preferred Draft in recesses: 2 min, 5 preferred Draft on cores: ½ min, 2 preferred Diameter of core: 6.4 mm (¼ in.) min Allowance for machining: 0.8 mm (1/32 in.) min for castings up to 255 mm (10 in.) long 1.2 mm (3/64 in.) desirable; 1.2 mm (3/64 in.) min for castings over 255 mm (10 in.) long, 1.6 mm (1/16 in.) desirable; 1.6 mm (1/16 in.) for surfaces formed by sand cores Minimum radius of fillet: Average thickness of joining walls Straightness: 0.25 to 150 mm (0.010 to 6 in) plus 0.030 mm (0.0012 in.) per 25 mm (1 in.) of length up to 915 mm (36 in.) Die castings

Draft on outside surface: ½ min Draft in cored holes: 0.50 mm (0.020 in.) per 25 mm (1 in.) of depth for 2.5 to 3.2 mm (1/10 to 1/8 in.) diam; 0.41 mm (0.016 in.) per 25 mm (1 in.) of depth for 3.2 to 6.3 mm

versatility in a wide variety of alloys, shapes, and sizes. Alloys considered hot short because of cracking tendencies during solidification are more easily cast in green sand since molds offer reduced resistance to dimensional contraction during solidification. Lower mold strength is also advantageous for parts with widely varying section thicknesses and intricate designs. These advantages are diminished in chemically bonded molds, which display greater rigidity than green sand molds. There are only practical limitations in the size of the parts that can be cast. Minimum wall thickness is normally 4 mm (0.15 in.), but thicknesses as little as 2 mm (0.090 in.) can be achieved. Suggested tolerances for sand castings are in Table 5. Sand castings are relatively lower in dimensional accuracy with poor surface finish; basic linear tolerances of  þ30 mm/m (þ 0.03 in./in.), with a minimum tolerance of 20 mm/m (0.020 in./in.), and surface finishes of 6 to 13 mm (250 to 500 min.) root mean square (rms) are typical. Chemically bonded sands offer improved surface finish and dimensional accuracy and relatively unlimited shelf life, depending on the type of binder used. Achievable dimensional tolerances can be substantially improved using precision methods in the forming and assembly of dry sand mold components. Surface quality can be improved by using a finer grade of facing sand in the molding process. Strength is typically lower as a result of slow solidification rates.

(1/8 to ¼ in.) diam; 0.30 mm (0.012 in.) per 25 mm (1 in.) of depth for 6 to 25 mm (¼ to 1 in.) diam; 0.30 mm (0.012 in.) per 25 mm (1 in.) of depth. plus 0.05 mm (0.002 in.) per 25 mm (1 in.) of diam for diameters over 25 mm (1 in.) Plaster mold castings

Maximum length of core supported at one end: 5  diam of core Draft on cores: Zero draft often permissible, ½ min, otherwise, with 2 preferred Diameter of core: 6.3 mm (¼ in.) min Minimum radius of fillet: Sharp corners can be cast Flatness: 0.038 mm (0.0015 in.) per 25 mm (1 in.) of length Shell mold castings

Draft on outside surfaces: ½ min, 2 preferred Straightness: 0.05 mm (0.002 in.) per 25 mm (1 in.) of length Investment castings

Maximum length of core supported at one end: 5  diam of core (on some parts 10  diam) Location of cored hole: 0.10 mm (0.004 in.) per 25 mm (1 in.) of dimension Concentricity: 0.05 mm (0.002 in.) per 25 mm (1 in.) Straightness: 0.13 mm (0.005 in.) per 25 mm (1 in.) of length Centrifugal permanent mold castings

Diameter of core: 3.2 mm (1/8 in.) min Draft on core: ½ min for 1  diam of core, 1½ preferred; 1 min for 2  diam, 2 preferred; 1½ for > 2  diam, 3 preferred Contour surfaces: 0.13 mm (0.005 in.) min for surfaces up to 50 mm (2 in.) long, 0.25 mm (0.010 in.) preferred; 0.25 mm (0.010 in.) 25 mm (1 in.), min for surfaces over 50 mm (2 in.) long; 0.025 mm (0.001 in.) for each additional 25 mm (1 in.) 0.76 mm (0.030 in.) max Minimum radius of fillet: 0.13 mm (0.005 in.) fillet radius equal to web thickness preferred Angularity: 0 50 per 25 mm (1 in.) up to 25 mm (1 in.); 0 50 per 25 mm (1 in.) for each additional 25 mm (1 in.) to 300 max

Other expendable mold methods include:
Lost foam casting with unbonded sand mold

compacted around an expendable polystyrene pattern
Shell mold casting, which surpasses ordinary sand castings in surface finish and dimensional accuracy (table and cool at slightly higher rates; however, equipment and production costs are higher, and the size and complexity of castings that can be produced are limited.
Plaster casting with a permeable (aerated) or impermeable plaster mold. The high insulating value of the plaster allows castings with thin walls to be poured. Minimum wall thickness of aluminum plaster casting is typically 1.5 mm (0.060 in.), Plaster molds have high reproducibility, permitting castings to be made with fine details and close tolerances; basic linear tolerances of  þ5 mm/m (þ 0.005 in./in.) are typical. The surface finish of plaster castings is excellent; aluminum castings attain finishes of 1.3 to 3.2 mm (50 to 125 min.) rms. Suggested tolerances are in Table 6.
Investment casting of aluminum most commonly employs ceramic molds and expendable patterns of wax, plastic, or other low-temperature melting materials. Aluminum investment castings can have walls as thin as 0.40 to 0.75 mm (0.015 to 0.030 in.) basic linear tolerances of  þ5 mm/m

(þ 5 mils/in), and surface finishes of 1.5 to 2.3 mm (60 to 90 min.). Because of porosity and slow solidification, the mechanical properties of many aluminum investment castings are typically lower than those demonstrated by other casting processes. The interest of the aerospace and other industries in the combination of accurate dimensional control with controlled mechanical properties has resulted in the use of improved technologies to produce premium-quality castings by investment methods. Castings in the premium strength range can be achieved with molten-metal treatments, gating, and solidification conditions that are not typical for conventional investment castings. Investment casting applications include: instrument parts, impellers, compressor vanes, gears, ratchets, pawls, scrolls, speed brakes, wing tips, and aircraft pylons. Permanent Mold Casting. In principle, permanent mold casting is analogous to expendable mold casting processes. In this case, the molds are machined cast, wrought or nodular iron, cast steel, or wrought steel and can be reused repetitively until damage or wear necessitates repair or replacement. Intricate details and undercuts can often be cast using segmented steel cores. Permanent mold tooling is typically more expensive than that required for sand casting and other expendable mold processes and is justified by the volume of production.

Aluminum Shape Casting / 1013 Table 2

Dimensional tolerances of aluminum alloy castings Tolerance for dimensional limits, mm (in.)

Casting process

Under 25 mm (1 in.)

Over 25 mm (1 in.), add per 25 mm (1 in.)

Across parting line Sand þ0.38 (þ  0.015) Die . . .. Permanent mold  þ0.38 (þ 0.015) Plaster mold þ0.25 (þ  0.010) Shell mold þ0.25 (þ0.010)   Centrifugal(a) þ0.25 (þ  0.010)(b) Investment . . ..

. . .. . . .. +0.051 (+0.002) +0.038 (+0.0015) +0.030 (+0.0012) +0.025 (+0.001)(c) . . ..

Between points for one part of mold Sand þ0.38 (þ0.015)   Die . . .. Permanent mold  þ0.38 (þ 0.015) Plaster mold þ0.13 (þ0.005)   Shell mold þ0.13 (þ  0.005) Centrifugal þ0.25 (þ0.010)(b)   Investment . . ..

. . .. . . .. +0.025 (+0.001) +0.038 (+0.0015) +0.030 (+0.0012) +0.025 (+0.001)(c) . . ..

Sketch of dimensional requirement

Between points produced by core and mold Sand þ0.38 (þ0.015) . . ..   Die . . .. . . .. Permanent mold þ0.38 (þ0.015) +0.05 (+0.002)   Plaster mold þ0.25 (þ +0.038 (+0.0015)  0.010) Shell mold þ0.25 (þ0.010) +0.038 (+0.0015)   Centrifugal(a) þ0.38 (þ +0.025 (+0.001)(e)  0.015)(d) Investment . . .. . . ..

(a) Using permanent molds. (b) For dimensions under 50 mm (2 in.)  þ0.13 mm (þ 0.005 in.) min. (c) For dimensions over 50 mm (2 in.) +0.25 mm (+0.010 in.) per 25 mm (1 in.) min. (d) For dimensions under 50 mm (2 in.)  þ0.25 mm (þ 0.010 in.) min. (e) For dimensions over 50 mm (2 in.) þ0.38 mm (þ0.015 in.) per 25 mm (1 in.) min.  

While the principles and mechanics of gravity casting are similar, the metallurgical structure of permanent mold castings reflects the refinement of higher solidification rates. Typical and specified minimum mechanical properties, including ductility, are higher than those of expendable mold castings. The improved mechanical properties of permanent mold castings provide part of the justification for selecting this process over competing gravity casting options. There is a downsprue into which molten metal is introduced from a pouring basin, from which the metal flows into runners, risers, infeeds, and casting cavity. Directional solidification is promoted by selective chilling of mold sectors by air, mist, or water. An insulating coating is used to protect the mold from the molten aluminum and to facilitate removal of the casting from the mold after solidification is complete. Typical mold dressings or washes are suspensions of talc, various metal oxides such as zirconia, chromia, iron oxide and titania, colloidal graphite and calcium carbonate in water, and sodium silicate. The thickness and thermal characteristics of these coatings are used to locally increase or decrease heat absorption during solidification. As a result of wear, these coatings must be periodically repaired or replaced to ensure consistent process performance and casting results. Mold surfaces are periodically blasted with dry ice or

mild abrasives to remove coatings and scale, after which new mold coatings are applied. The permanent mold process is less alloy-tolerant than most expendable mold processes. The most popular permanent mold alloys display superior castability, such as those of the Al-Si, Al-Si-Mg, and Al-Si-Cu (Mg) families. Mold rigidity is a challenge in the casting of hot-short alloys in which liquidus-solidus range and elevated-temperature strength combine to increase the tendency for cracking during and after solidification. Determined efforts to cast even the most difficult foundry alloys, such as low-iron aluminum-copper alloys, have nevertheless been successful, and alloys with limited castability are routinely cast in permanent molds. Suggested tolerances are in Table 7. Permanent mold castings can be produced in sizes ranging from less than 0.5 kg (1 lb) to more than several hundred pounds. Surface finish typically varies 3.8 to 10 mm (150 to 400 min.). Basic linear tolerances are approximately  þ10 mm/m (þ 0.01 in./in.), and minimum wall thicknesses are approximately 2.5 mm (0.100 in.). Centrifugal Casting. Centrifugal force in aluminum casting involves rotating a mold or a number of molds filled with molten metal about an axis. Baked sand, plaster, or graphite molds have been used, but iron and steel dies are most common. Centrifugal castings are generally, but not always, denser than conventionally poured castings and offer the advantage of greater detail.

Wheels, wheel hubs, motor rotors, and papermaking and printing rolls are examples of aluminum parts produced by centrifugal casting. Aluminum alloys suitable for permanent mold, sand, or plaster casting can be cast centrifugally. Squeeze Casting. Although a number of process developments have been referred to as squeeze casting, the process by which molten metal solidifies under pressure within closed dies positioned between the plates of a hydraulic press is the only version of current commercial interest. Squeeze casting has been successfully used for a variety of ferrous and nonferrous alloys in traditionally cast and wrought compositions. Applications of squeeze-cast aluminum alloys include reciprocating engine pistons, brake rotors, automotive and truck wheels, and structural automotive frame components (Fig. 3). Squeeze casting is simple and economical, is efficient in its use of raw material, and has excellent potential for automated operation at high rates of production. Semisolid forming incorporates elements of casting, forging, and extrusion. It involves the near-net shape forming of metal parts from a semisolid raw material that incorporates a uniquely nondendritic microstructure. Semisolid forming is more costly than conventional casting but offers unique properties and consistently excellent quality. In addition, the viscous nature of semisolid alloys provides a natural environment for the incorporation of thirdphase particles in the preparation of reinforced metal-matrix composites.

Pressure Die Casting of Aluminum Pressure die casting came into existence in the early 1820s in response to the expanding need for large volumes of cast print type. Iron or steel dies had been used in casting print type in lead-base alloys in the 17th century. Iron molds were also used in colonial times to cast pewter. Intensive efforts to employ iron and steel molds in the casting of aluminum resulted in commercial “permanent mold” operations by the first decade of the 20th century. Progress in die casting aluminum was limited until the development of the cold chamber process in the 1920s. In pressure die casting, molten metal is injected under pressure into water-cooled dies. Pressure is maintained until the part has solidified. Molten metal usually enters the mold by the action of a hydraulic ram in a containment chamber (shot chamber), resulting in rapid filling of the mold cavity. While lubrication is required to facilitate the separation of the casting from the die surface, the dies are otherwise uninsulated. Dies are usually machined from high-quality tool steels. The die casting process has undergone significant changes through the evolution of machine design and instrumentation as well as process

1014 / Casting of Nonferrous Alloys Table 3 Typical tolerances of aluminum alloy castings Metal

Aluminum-silicon (400 series) Aluminum-copper (200 series) Aluminum-silicon (300 series) Aluminum-silicon (400 series) Aluminum-copper (200 series) Aluminum-silicon (300 series) Aluminum-magnesium (500 series) Aluminum-magnesium (500 series) Aluminum-copper (200 series) Aluminum-silicon (400 series) Aluminum-silicon (300 series) Aluminum-silicon (400 series) Aluminum-silicon (300 series) Aluminum-silicon (300 series) Aluminum-magnesium (500 series) Aluminum-silicon (400 series) Aluminum-silicon (400 series) Aluminum-copper (200 series) Aluminum-silicon (300 series) Aluminum-magnesium (500 series) Aluminum-silicon (300 series) Aluminum-silicon (400 series) Aluminum-copper (200 series) Aluminum-copper (200 series) Aluminum-magnesium (500 series) Aluminum-silicon (300 series) Aluminum-silicon (400 series) Aluminum-copper (200 series) Aluminum-magnesium (500 series) Aluminum-silicon (300 series) Aluminum-silicon (300 series) Aluminum-magnesium (500 series) Aluminum-copper (200 series) Aluminum-silicon (300 series) Aluminum-silicon (300 series) Aluminum-magnesium (500 series) Aluminum-silicon (400 series) Aluminum-copper (200 series) Aluminum-silicon (300 series) Aluminum-magnesium (500 series) Aluminum-silicon (400 series) Aluminum-copper (200 series) Aluminum

From dimension

To dimension

Lower tolerance

Upper tolerance

Premium tolerance

Parting line increment

Process(a)

mm

in.

mm

in.

mm

mm

mm

mm

Centrifugal Centrifugal Centrifugal Ceramic Ceramic Ceramic Ceramic EPC/full mold EPC/full mold EPC/full mold EPC/full mold Gravity die Gravity die Green sand Green sand Green sand Investment Investment Investment Permanent mold Permanent mold Permanent mold Permanent mold Plaster Plaster Plaster Plaster Plaster Plaster Plaster Pressure die casting Pressure die casting Pressure die casting Pressure die casting Shell Shell Shell Shell No-bake/air set/CO15 No-bake/air set/CO17 No-bake/air set/CO18 No-bake/air set/CO28 Sand

25 25 25 2.5 2.5 2.5 2.5 2.5 2.5 2.5 2.5 25 25 25 25 25 25 25 25 25 25 25 25 25 25 25 25 25 25 25 25 25 25 25 25 25 25 25 25 25 25 25 0 152

1 1 1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 0 6

305 305 305 610 610 610 610 254 254 254 254 152 152 152 152 152 254 254 254 152 152 152 152 2540 2540 2540 2540 2540 2540 2540 152 152 152 152 152 152 152 152 152 152 152 152 152 1016

12 12 12 24 24 24 24 10 10 10 10 6 6 6 6 6 10 10 10 6 6 6 6 100 100 100 100 100 100 100 6 6 6 6 6 6 6 6 6 6 6 6 6 40

2.5 2.5 2.5 0.25 0.25 0.25 0.25 0.38 0.38 0.38 0.38 0.64 0.64 0.38 0.38 0.38 0.25 0.25 0.25 0.23 0.23 0.23 0.23 0.25 0.25 0.25 0.25 0.25 0.25 0.25 0.13 0.13 0.13 0.13 0.25 0.25 0.25 0.25 0.25 0.25 0.25 0.25 0.81 0.81

in.

0.1 0.1 0.1 0.01 0.01 0.01 0.01 0.015 0.015 0.015 0.015 0.025 0.025 0.015 0.015 0.015 0.01 0.01 0.01 0.009 0.009 0.009 0.009 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.005 0.005 0.005 0.005 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.032 0.032

3.8 3.8 3.8 1.5 1.5 1.5 1.5 0.89 0.89 0.89 0.89 1.3 1.3 0.5 0.5 0.5 1.3 1.3 1.3 0.30 0.30 0.30 0.30 1.3 1.3 1.3 1.3 1.3 1.3 1.3 2.0 2.0 2.0 2.0 0.36 0.36 0.36 0.36 0.36 0.36 0.36 0.36 0.81 0.81

in.

0.15 0.15 0.15 0.06 0.06 0.06 0.06 0.035 0.035 0.035 0.035 0.05 0.05 0.02 0.02 0.02 0.05 0.05 0.05 0.012 0.012 0.012 0.012 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.08 0.08 0.08 0.08 0.014 0.014 0.014 0.014 0.014 0.014 0.014 0.014 0.032 0.032

1.3 1.3 1.3 0.13 0.13 0.13 0.13 0.05 0.05 0.05 0.05 0.25 0.25 0.51 0.51 0.51 0.13 0.13 0.13 0.13 0.13 0.13 0.13 0.38 0.038 0.038 0.038 0.038 0.038 0.038 0.025 0.025 0.025 0.025 0.23 0.23 0.23 0.23 0.23 0.23 0.23 0.23 ... ...

in.

0.05 0.05 0.05 0.005 0.005 0.005 0.005 0.002 0.002 0.002 0.002 0.01 0.01 0.02 0.02 0.02 0.005 0.005 0.005 0.005 0.005 0.005 0.005 0.0015 0.0015 0.0015 0.0015 0.0015 0.0015 0.0015 0.001 0.001 0.001 0.001 0.009 0.009 0.009 0.009 0.009 0.009 0.009 0.009 ... ...

0 0 0 0.25 0.25 0.25 0.25 0 0 0 0 0.38 0.38 0.792 0.76 0.76 0 0 0 0.38 0.38 0.38 0.38 0.25 0.25 0.25 0.25 0.25 0.25 0.25 0.25 0.25 0.25 0.25 0.25 0.25 0.25 0.25 0.51 0.51 0.51 0.51 ... ...

in.

0 0 0 0.01 0.01 0.01 0.01 0 0 0 0 0.015 0.015 0.0312 0.03 0.03 0 0 0 0.015 0.015 0.015 0.015 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.02 0.02 0.02 0.02 ... ...

First/base tolerance mm/m

10 10 10 6.0 6.0 6.0 6.0 3.0 3.0 3.0 3.0 15.6 15.6 25 25 25 1.0 1.0 1.0 15.6 15.6 15.6 15.6 2.0 2.0 2.0 2.0 2.0 2.0 2.0 6.0 6.0 6.0 6.0 5.0 5.0 5.0 5.0 20 20 20 20 ... ...

in./in.

0.01 0.01 0.01 0.006 0.006 0.006 0.006 0.003 0.003 0.003 0.003 0.015625 0.015625 0.025 0.025 0.025 0.001 0.001 0.001 0.015625 0.015625 0.015625 0.015625 0.002 0.002 0.002 0.002 0.002 0.002 0.002 0.006 0.006 0.006 0.006 0.005 0.005 0.005 0.005 0.02 0.02 0.02 0.02 ... ...

(a) EPC, evaporative pattern casting. Source: American Foundry Society, 2008

Table 4 Comparison of sand casting, die casting, and permanent mold casting processes for aluminum alloys Casting process Factor

Cost of equipment Casting rate Size of casting External and internal shape Minimum wall thickness Type of cores Tolerance obtainable Surface finish Gas porosity Cooling rate Grain size Strength Fatigue properties Wear resistance Overall quality Remarks

Sand casting

Permanent mold casting

Die casting

Lowest cost if only a few items required Lowest rate Largest of any casting method Best suited for complex shapes where coring required 3.0–5.0 mm (0.125–0.200 in.) required; 4.0 mm (0.150 in.) normal Complex baked sand cores can be used

Less than die casting 11 kg/h (25 lb/h) common; higher rates possible Limited by size of machine Simple sand cores can be used, but more difficult to insert than in sand castings 3.0–5.0 mm (0.125–0.200 in.) required; 3.5 mm (0.140 in.) normal Reusable cores can be made of steel, or nonreusable baked cores can be used Best linear tolerance is 10 mm/m (10 mils/in.)

Highest 4.5 kg/h (10 lb/h) common; 45 kg/h (100 lb/h) possible Limited by size of machine Cores must be able to be pulled because they are metal; undercuts can be formed only by collapsing cores or loose pieces 1.0–2.5 mm (0.040–0.100 in.); depends on casting size

Poorest; best linear tolerance is 300 mm/m (300 mils/in.) 6.5–12.5 mm (250–500) min.) Lowest porosity possible with good technique 0.1–0.5  C/s (0.2–0.9  F/s) Coarse Lowest Good Good Depends on foundry technique Very versatile as to size, shape, internal configurations

4.0–10 mm (150–400 min.) Best pressure tightness; low porosity possible with good technique 0.3–1.0  C/s (0.5–1.8  F/s) Fine Excellent Good Good Highest quality ...

Steel cores; must be simple and straight so they can be pulled Best linear tolerance is 4 mm/m (4 mils/in.) 1.5 mm (50 min.); best finish of the three casting processes Porosity may be present 50–500  C/s (90–900  F/s) Very fine on surface Highest, usually used in the as-cast condition Excellent Excellent Tolerance and repeatability very good Excellent for fast production rates

Aluminum Shape Casting / 1015 Table 5

Suggested dimensional tolerances for sand castings Type A dimension: between two points in same part of mold, not affected by parting plane or core Tolerance, þ, mm (in.) 

Specified dimension mm (in.)

Critical

Up through 152 (6) Over 152 to 305 (6 to 12)

0.76 (0.030) 30 + 3.0 mm/m (0.030 + 0.003 in./in.) over 152 mm (6 in.) 48 + 2.0 mm/m (0.048 + 0.002 in./in.) over 305 (12 in.)

Over 305 (12)

Type B dimension: across parting plane. A-type dimension plus following:

Noncritical

1.0 (0.040) 40 + 4.0 mm/m (0.040 + 0.004 in./in.) over 152 mm (6 in.) 64 + 2.0 mm/m (0.064 + 0.002 in./in.) over 305 (12 in.) Type C dimension: affected by core. A-type dimension plus following:

Projected area of casting, A1  A3, cm2 (in.2)

Additional tolerance for parting plane mm (in.)

Projected area of casting affected by core, A3  G, cm2 (in.2)

Up through 65 (10) Over 65 to 325 (10 to 50) Over 325 to 645 (50 to 100) Over 645 to 1615 (100 to 250) Over 1615 to 3225 (250 to 500)

0.5 (0.020) 0.89 (0.035) 1.1 (0.045) 1.5 (0.060) 2.3 (0.090)

Up through 65 (10) Over 65 to 325 (10 to 50) Over 325 to 645 (50 to 100) Over 645 to 3225 (100 to 500) Over 3225 to 6450 (500 to 1000)

D dimension: draft

Additional tolerance for core, mm (in.)

0.5 (0.020) 0.89 (0.035) 1.1 (0.045) 1.5 (0.060) 2.3 (0.090)

F dimension: allowance for finish Draft, deg

Location

Outside wall Recesses Cores

Critical

2 3 2

Noncritical

Maximum dimension, mm (in.)

3 5 3

Up through 152 (6) Over 152 to 305 (6 to 12) Over 305 to 455 (12 to 18) Over 455 to 610 (18 to 24)

E dimension: minimum wall thickness, 3.8 mm (0.150 in.)

Nominal allowance, mm (in.)

1.5 2.3 3.0 4.6

(0.060) (0.090) (0.120) (0.180)

Minimum diameter of cored holes, 6.4 mm (0.250 in.)

development and controls. The demand for larger, more complex castings with improved quality and lower cost has led to the development and promotion of specialized die casting machines capable of higher rates of production and improved performance. There are two basic concepts: hot and cold chamber operation. In the hot chamber process, the shot chamber and piston are immersed in molten metal. Metals such as magnesium and zinc that do not aggressively attack the materials of construction can be efficiently cast by this method, with production rate advantages. Despite intensive efforts to develop hot chamber process designs and materials that could be used in aluminum casting, none have been commercially successful. Except in rare cases, all aluminum die casting is performed in cold chamber equipment in which the shot chamber is filled with each cycle, and the chamber and

piston assembly are not continuously in contact with molten aluminum. Die casting machine designs are also differentiated by parting plane orientation. In practice, the dies are mounted on platens that can operate in either vertical or horizontal directions. Early die casting was typically performed with vertical die movement. Today (2008), with exceptions, vertical die casting is restricted to rotor production. Locking pressure defines the capacity of the machine to contain the pressure generated during the injection cycle. The larger the plan area of the casting and the greater the hydraulic pressure applied, the greater the required locking pressure. Die casting machines are designed with locking pressures from as little as 25 tonnes to more than 4500 tonnes corresponding to injection pressures of up to 280 MPa (40 ksi).

The principles of directional solidification and gravity-based gating and risering are essentially inapplicable to die casting. Gates and runners are used to convey metal from the shot chamber to the die cavity. Geometrical considerations are observed to minimize turbulence, air entrapment, and fragmentation of the metal stream, but no effort other than the use of sustained pressure is used to promote internal soundness. Techniques are used to intensify pressure during the solidification phase to decrease the volume fraction of internal porosity. With metal velocities exceeding 30 m/s (100 ft/s) and solidification rates exceeding 1000  C/s (1800  F/s), the greatest quality concern is entrapped gases, including combustion or volatilization components of the lubricant and turbulence-related inclusions. Rapid filling of the mold (20 to 100 ms) and rapid solidification under pressure combine to produce a consistently dense, fine-grained, and highly refined surface structure with excellent properties, including fatigue strength. Internal unsoundness affects bulk properties, the acceptability of machined surfaces, and pressure tightness. Impregnation is routine for die castings that must contain gases or liquids under substantial pressure. Internal unsoundness generally prevents full heat treatment and welding because of the risk of blister formation when die castings are exposed to elevated temperatures. Lower-temperature thermal treatments for stabilization or hardening are routinely used. In special cases and in restricted casting areas, limited welding can also be performed. The die casting process is the least alloy-tolerant of important commercial casting processes. Solidification conditions require alloys of superior castability and that display good resistance to cracking at elevated temperatures. The highly castable alloys of the aluminum-silicon family are the most common. Of these, alloy 380.0 and its variations comprise approximately 85% of total die casting production. These compositions provide attractive combinations of cost, strength, hardness, and corrosion resistance in the as-cast state, with excellent fluidity and resistance to hot cracking. Aluminum-silicon alloys lower in copper, such as 360.0, 364.0, 413.0, and 443.0, offer improved corrosion resistance and excellent castability. Hypereutectic aluminum-silicon alloys, including 390.0, have become more important in wear-resistant applications. Magnesium content is usually controlled at low levels to minimize oxidation and the generation of oxides in the casting process. Most commonly used die casting alloys specify restrictive magnesium limits. Nevertheless, aluminum-magnesium alloys can be die cast. Alloy 518.0, for example, is specified when the highest corrosion resistance and the brightest, most reflective finish are required. Iron contents of 0.7% or greater are preferred to maximize elevated-temperature strength, to facilitate ejection, and to minimize soldering

1016 / Casting of Nonferrous Alloys Table 6 Suggested dimensional tolerances for plaster castings Type A dimension: between two points in same part of mold, not affected by parting plane or core Specified dimension mm (in.)

Critical

Up through 25 (1) Over 25 (1)

Tolerance,  þ, mm (in.)

0.13 (0.005) 5.0 + 1.0 mm/m (0.005 + 0.001 in./in.) over 25 mm (1 in.)

Type B dimension: across parting plane. A-type dimension plus following: Additional tolerance for parting plane, mm (in.)

Projected area of casting, A1  A3, cm2 (in.2)

Up through 65 (10) Over 65 to 325 (10 to 50) Over 325 to 645 (50 to 100) Over 645 (100)

0.13 0.25 0.51 0.76

(0.005) (0.010) (0.020) (0.030)

D dimension: draft

Noncritical

0.25 (0.010) 10 + 2.0 mm/m (0.010 + 0.002 in./in.) over 25 mm (1 in.)

Type C dimension: affected by core. A-type dimension plus following: Projected area of casting affected by core, A3  G, cm2 (in.2)

Additional tolerance for core, mm (in.)

Up through 65 (10) Over 65 to 325 (10 to 50) Over 325 to 645 (50 to 100) Over 645 (100)

0.13 (0.005) 0.51 (0.020) 0.76 (0.030) 1.1 (0.045)

F dimension: allowance for finish Draft, deg

Location

Outside surface Recesses Cores

Critical

0 0 0

Noncritical

Maximum dimension, mm (in.)

Nominal allowance, mm (in.)

2 2 2

Up through 125 (5) Over 125 to 305 (5 to 12) Over 305 to 455 (12 to 18)

0.51 (0.020) 0.76 (0.030) 1.0 (0.040)

E dimension: minimum wall thickness, 1.5 mm (0.060 in.)

to the die face. Iron content is usually 1  þ 0.3%, but greater concentrations are also used. Improved ductility through reduced iron content has been an incentive resulting in widespread efforts to develop a tolerance for iron as low as approximately 0.25%. These efforts focus on process refinements, design modifications, and improved die lubrication. At higher iron concentrations, there is a risk of exceeding solubility limits of coarse Al-Fe-Cr-Mn segregate at molten-metal temperatures. Sludging or precipitation of segregate is prevented by chemistry controls related to metal temperature. A common rule is: Fe + 2 (Mn) + 3(Cr) + 1.7, where element values are expressed in weight percent. Aluminum alloys can be die cast to tolerances of  þ4 mils/in. (þ 0.004 in./in.) and commonly have finishes as fine as 1.3 mm (50 min.). Parts are cast with walls as thin as 1.0 mm (0.040 in.). Cores, which are made of metal, are restricted to simple shapes that permit drawing or removal after solidification is complete. Suggested tolerances are in Table 8.

Minimum diameter of cored holes, 6.4 mm (0.250 in.)

Premium Engineered Castings A premium engineered casting is one that provides higher levels of quality and reliability than found in conventionally produced castings. Premium engineering includes intimately detailed design and control of each step in the manufacturing sequence. The results are minimally variable premium strength, ductility, soundness, dimensional control, and finish. Castings of this classification are notable primarily for mechanical property performance that reflects extreme soundness, fine dendrite arm spacing, and well-refined grain structure. Premium engineered casting objectives require the use of chemical compositions competent to display superior properties (see the article “Aluminum and Aluminum Alloy Castings” in this Volume). Most of the compositions designated as premium engineered alloys had their origins in the 1950s and early 1960s. Alloy A356.0, registered in 1955, and B356.0, in 1956, were derivatives of 356.0 alloy, which

was originally developed in 1930. Similarly, alloy C355.0 dates from 1955, while the parent 355.0 was first used in 1930. Alloys 359.0 and 354.0 were developed in 1961. Alloy A357.0 (1962) had its origins in Tens-50 alloy that was also first registered in 1961. Premium engineered aluminum castings represent the culmination of decades of research and development involving molten-metal treatment, methods for the removal of included matter, measurement or assessment of dissolved hydrogen, gating development, microstructural modification and refinement, casting processes, mold materials, and the development of highstrength, ductile alloys. Each of these developments was necessary to advance casting capabilities to meet an increasingly challenging range of application requirements. Melt processing and solidification of aluminum are discussed in more detail in other articles in this Volume. Fluxing and degassing are key aspects in the production of premium aluminum castings. More details on the melting, melt treatment, and melt handling of aluminum are discussed in other articles in this Volume. To a large extent, melt quality has been assessed by variations of the Straube-Pfeiffer test in which the relationship of hydrogen solubility and pressure was qualitatively measured (see the article “Approaches to Measurement of Melt Quality” in this Volume). The absence of entrained oxides was assessed by their influence on hydrogen precipitation under reduced pressure. A semiquantitative approach using sample densities was developed and used extensively as a process control tool. For greater sensitivity, controlled vibration during sample solidification at absolute pressures of 2 to 5 mm Hg (0.04 to 0.10 psi) was employed. Real-time measurement of dissolved hydrogen by partial pressure diffusion in molten metal supplemented vacuum test results. Validation of hydrogen assessments was provided by solid-state extraction techniques. In terms of solidification, simulation models and programs help put into practice the principles of directional solidification for more complex cast parts. The relationship of properties and dendrite arm spacing also has significantly influenced the premium engineered casting effort in terms of both strength and ductility. Mold Materials. Investment casting produced small parts in which dimensional accuracy and surface finish were important criteria. Plaster molds were used in the production of dimensionally accurate cast parts such as impellers and tire molds. There were corollary pattern, mold material, and mold processing developments. Differences in the molding and heat extraction characteristics of differing sands were measured, and at the same time, supplier developments in dry sand binders were studied and evaluated. In permanent mold, variations in mold wash chemistry and application were used with air, mist, and water cooling to control solidification. On occasion, copper and ceramic inserts were

Aluminum Shape Casting / 1017 Table 7 Suggested dimensional tolerances for permanent and semipermanent mold castings Type A dimension: between two points in same part of mold, not affected by parting plane or moving parts Specified dimension mm (in.)

Critical

Up through 25(1) Over 25(1)

Tolerance, þ, in. 

0.25 (0.010) 10 + 1.5 mm/m (0.010 + 0.0015 in./in.) over 25 mm (1 in.)

Type B dimension: across parting plane. A-type dimension plus following:

Noncritical

0.38 (0.015) 15 + 2.0 mm/m (0.015 + 0.002 in./in.) over 25 mm (1 in.)

Type C dimension: affected by moving parts. A-type dimension plus following: Additional tolerance, mm (in.)

Additional tolerance for parting plane mm (in.)

Projected area of casting, A1  A3, cm2 (in.2)

Up through 65 (10) Over 65 to 325 (10 to 50) Over 325 to 645 (50 to 100) Over 645 to 1615 (100 to 250) Over 1615 to 3225 (250 to 500)

0.25 0.38 0.51 0.64 0.76

Projected area of casting affected by moving part, A3  G, cm2 (in.2)

(0.010) (0.015) (0.020) (0.025) (0.030)

Up through 65 (10) Over 65 to 325 (10 to 50) Over 325 to 645 (50 to 100) Over 645 to 3225 (100 to 500) Over 3225 to 6450 (500 to 1000)

D dimension: draft

Metal core or mold

0.25 0.38 0.38 0.56 0.81

(0.010) (0.015) (0.015) (0.022) (0.032)

Sand core

0.38 (0.015) 0.64 (0.025) 0.76 (0.030) 1.0 (0.040) 1.5 (0.060)

E dimension: minimum wall thickness

Draft, deg Location

Outside surface Recesses Cores

Critical

2 2 1

Noncritical

3 5 2

Maximum dimension, mm (in.)

Minimum wall thickness, mm (in.)

Up through 76 (3) Over 76 to 152 (3 to 6) Over 152 (6)

3.5 (0.140) 4.1 (0.160) 4.8 (0.188)

F dimension: allowance for finish Nominal allowance, mm (in.) Maximum dimension, mm (in.)

Metal core or mold

Up through 152 (6) Over 152 to 305 (6 to 12) Over 305 to 455 (12 to 18) Over 455 to 610 (18 to 24)

0.76 (0.030) 1.1 (0.045) 1.5 (0.060) 2.3 (0.090)

Sand core

1.5 2.3 3.0 4.6

(0.060) (0.090) (0.120) (0.180)

Minimum diameter of cored holes, 9.5 mm (0.375 in.)

designed into the mold to promote solidification directionality. The use of steel chills, contoured sections, and mold plates was normal in sand casting. The use of all available mold components from most insulating (foamed plaster) to most rapid heat extraction (water-cooled copper) could be incorporated in mold designs to reproducibly alter solidification in order to promote the highest possible degree of internal soundness. Computer simulations of mold filling dynamics and solidification that incorporate differences in heat extraction through finite-element analysis are gradually supplanting art and instinct in process designs.

The result is a composite mold design of significant complexity using dry sand and at least several other mold materials for each configuration. Mold Filling. The inevitable degradation of melt quality that occurs in drawing and pouring through conventional methods was recognized as a significant barrier to achieving premium engineered quality levels and properties. Very significant efforts had been made to evaluate different pouring and gating approaches. These included siphon ladles, sprue/gate/runner ratios, runner overruns, dross traps, pouring cup designs, strainers and screens, and nonvortexing cross-sectional geometry of downsprues and

runners. The use of optimal gating designs and rigorously controlled practices are integral components of premium casting technology. Several developments have altered mold-filling options; these developments are described in the following paragraphs. First, the low-pressure casting process achieved commercial importance in the United States in the 1950s. A number of challenges that were unaddressed by commercial low-pressure systems were successfully met. A camcontrolled back-pressure method based on gross casting weight was used to retain residual metal levels at the top of the feed tube. This feature prevented the inclusion-spawning characteristic of normal low-pressure cycles. In-gate filtration and screening methods were also devised. The range of part designs and alloys that were cast would be considered unusual today (2008), when the low-pressure process has become principally known for automotive wheel production. Instead, the low-pressure method was considered a means of nonturbulent mold filling with a number of additional advantages that included reduced gross/net weight and lower pouring temperature. Some examples were diesel engine and compressor pistons, air conditioner compressor bodies, bearings, furniture parts, and missile fins. Geometric symmetry, which is normally a criterion for low-pressure production, was not considered a prerequisite, and many of the castings that were produced used conventional risering rather than exclusively relying on the in-feed for shrinkage compensation. These developments were adapted to premium engineered plaster and dry sand parts. High-speed rotors and impellers were excellent examples, but many other premium engineered casting configurations were made by low-pressure mold filling. Second, countergravity mold-filling methods were developed involving the use of mechanical or induction pumps. Third, various techniques have been developed and are in use for filling molds quiescently by displacement of molten metal. Fourth, solidification time was not a significant factor in expendable mold production when extensive chilling was used, but it was always a factor in permanent mold. For this reason, and to overcome other low-pressure process limitations, vacuum riserless casting (VRC) was developed in the early 1960s. Rather than pressurize a contained molten reservoir, the application of a vacuum on the mold cavity drew metal from the bath through a short fill tube. The metal source was exposed for periodic treatment, the distance from subsurface metal entry to the casting cavity was minimal, dies were extensively chilled, and the process could be highly automated. While only relatively small and simple shapes were produced by the VRC method, productivity and mechanical properties were exceptional. More than 20 million air conditioner pistons and millions of rocker arms were produced by this process.

1018 / Casting of Nonferrous Alloys

Fig. 3

Automotive parts produced by squeeze casting

Table 8 Suggested dimensional tolerances for die castings Type A dimension: between two points in same part of die, not affected by parting plane or moving parts Specified dimension, mm (in.)

Up through 25 (1) Over 25 to 305 (1 to 12) Over 305 (12)

Critical

0.10 (0.004) 4.0 + 1.5 mm/m (0.004 + 0.0015 in./in.) over 25 mm (1 in.) 20.5 + 1.0 mm/m (0.0205 + 0.001 in./in.) over 305 mm (12 in.)

Type B dimension: across parting plane. A-type dimension plus following: Projected area of casting, A1  A3, cm2 (in.2)

0.13 0.20 0.30 0.38

(0.005) (0.008) (0.012) (0.015)

D dimension: draft

Up through 2.5 (0.1) Over 2.5 to 13 (0.1 to 0.5) Over 13 to 25 (0.5 to 1.0) Over 25 to 76 (1.0 to 3.0) Over 76 to 230 (3.0 to 9.0)

6 3 2 1

/32

1

Projected area of casting affected by moving part, A3  G, cm2 (in.2)

Up through 65 (10) Over 65 to 130 (10 to 20) Over 130 to 325 (20 to 50) Over 325 to 645 (50 to 100)

Additional tolerance for moving part, mm (in.)

0.13 0.20 0.30 0.38

(0.005) (0.008) (0.012) (0.015)

E dimension: minimum wall thickness, 2.3 mm (0.090 in.) Draft, deg

Critical

Noncritical

0.25 (0.010) 10 + 2.0 mm/m (0.010 + 0.002 in./in.) over 25 mm (1 in.) 32 + 1.0 mm/m (0.032 + 0.001 in./in.) over 305 mm (12 in.)

Type C dimension: affected by moving parts. A-type dimension plus following:

Additional tolerance for parting plane, mm (in.)

Up through 325 (50) Over 325 to 645 (50 to 100) Over 645 to 1290 (100 to 200) Over 1290 to 1935 (200 to 300)

Depth of draw, mm (in.)

Tolerance, þ, mm (in.) 

Noncritical

18 6 3 1½ 1

F dimension: allowance for finish Maximum dimension, mm (in.)

Nominal allowance, mm (in.)

Up through 125 (5) Over 125 to 305 (5 to 12) Over 305 to 455 (12 to 18)

0.51 (0.020) 0.76 (0.030) 1.0 (0.040)

Minimum diameter of cored holes, 3.56 mm (0.140 in.)

Fifth, the level pour process for premium engineered castings was developed. This process had its origins in the direct-chill process that was developed for fabricating ingot. It was natural that the shared concerns for metal distribution and solidification principles would result in the synthesis of process concepts. Aluminum-tin alloy bearings, which were typically hollow or solid cylinders, could be produced by the direct-chill process, but segregation and safety concerns led to a variation in which metal was introduced to the bottom of a permanent mold through a moving pouring cup that traversed the length of the mold. Quiescent flow and the continuous layering of molten metal provided improved internal quality. Excellent soundness was obtained without the use of the extensive risering normally required for these long-solidification-range alloys, pouring temperature was reduced, and the solidified structure was more chemically homogeneous than in conventionally cast parts. With determined engineering, the same approach was successfully used for more complex configurations with more challenging metallurgical requirements. The cost of doing so in permanent mold was not typically justified. However, aircraft and aerospace parts in the limited quantities normally associated with sand casting offered exciting opportunities for level pour technology, and a large number of prototype and production parts were produced by this method. In its final form, the assembled mold was lowered on a hydraulic platform through a trough arrangement that provided nonturbulent flow of metal through entry points that paralleled the vertical traverse of the mold. Metal flow was controlled by the dimensions of the entries and the lowering rate, which could be modulated for cross-sectional variations as a function of mold travel. Characteristic of the direct-chill process on which it was based, the level pour process features quiescent moltenmetal flow, minimized feeding distances, and reduced pouring temperatures. ACKNOWLEDGMENTS Portions adapted from:
J.G. Kaufman and E. Rooy, Aluminum Cast-

ing Alloys: Properties, Processes, and Applications, American Foundry Society and ASM International, 2004
A. Kearry and E. Rooy, Properties and Selection: Nou Ferrous Alloys and SpecialPurpose Materials, Vol 2, ASM Handbook, ASM International, 1990
E. Rooy, Casting, Vol 15, ASM Handbook, ASM International, 1988, p 743
Van Hoen, Aluminum, Vol. III Fabrication and Finishing, American Society for Metals, 1967

ASM Handbook, Volume 15: Casting ASM Handbook Committee, p 1019-1025 DOI: 10.1361/asmhba0005288

Copyright © 2008 ASM International® All rights reserved. www.asminternational.org

Copper Continuous Casting* Derek E. Tyler, Olin Brass (Retired) Richard P. Vierod, Olin Brass

COPPER ALLOY PRODUCTS such as strip, billet, rod, or tube are continuous cast, defined as the continuous solidification and withdrawal of product from an open-ended shaping mold. Methods include both vertical and horizontal casting, depending on product size, shape, and volume. Casting vertically has certain inherent technical advantages. The symmetry of cooling promotes a uniform and predictable solidification growth pattern and uniform axial loading on the freshly solidified shell as it is withdrawn from the mold. In tube or hollow section casting, the vertical process has particular merit. The disadvantages of vertical casting are mostly logistic: difficulty in handling long lengths of section; cut-off can be more difficult to engineer and control; and it is generally a semicontinuous operation. Horizontal casting requires lower capital investment, is compatible with lower production rates, and is a continuous operation. This article briefly reviews the history and methods of copper alloy continuous casting; the information is drawn from the very detailed and extensive coverage of the subject in Ref 1 and the numerous publications of equipment supply companies such as Rautomead, SMS Meer, and so on. Although a relatively simple process, the metallurgical complexities of continuous casting involve the thermal and mechanical interactions between the mold and the moving solidifying shell of the casting. The variations of alloy chemistry, physical, thermal, and mechanical properties invoke detailed changes in the process protocols. The heat transfer within the mold is a key factor in the successful production of a quality product. The equipment suppliers have their own proprietary designs for the mold assembly, and some of the various design details of the dies and cooling assemblies are described in Ref 1. Typically, the molds (dies) are fabricated from copper alloy components that may or may not be sleeved with high-quality graphite inserts. When sleeves are used, precise machining and polishing of the surfaces is essential to provide intimate contact between graphite and

plate cooler to maximize heat transfer. Precise control of the withdrawal parameters then becomes of paramount importance.

History Many of the techniques developed for copper alloys have been adapted for aluminum and steel. In a general timeline of developments (Table 1), the breakthrough in continuous casting of nonferrous alloys can be credited to Eldred in 1930 (Ref 2). He developed a machine using graphite as the mold material

to cast copper rods and later a number of copper-base alloys. Prior to that, the continuous casting of metals had been practiced for almost a century earlier with patents by Sellers and Laign (Ref 3, 4). In 1938, Poland and Lindner were granted a U.S. patent (Ref 7) for a vertical casting machine very similar to Eldred’s. The graphite mold was cooled by a close-fitting metal water jacket (Fig. 1). The potential of graphite as a suitable mold material was quickly appreciated by various companies, such as American Smelting and Refining Company (Ref 8) and Flocast (Ref 17). The Asarco process (Ref 8), patented

Table 1 Historical timeline of continuous casting Year

Development related to continuous casting of copper

1840 1843

The first recorded patent in the nonferrous field was by Sellers for the manufacture of lead pipes (Ref 3). About the same time, Laign filed a patent in America for a method of continuous casting nonferrous metal tube (Ref 4). U.S. patent was granted to Swedish engineer Pehrson for the first horizontal closed-head system for continuous casting of cast iron bars (Ref 5). The Hazelett process was introduced as the first ingotless rolling plant for continuous casting of copper wire rod (Contirod process) and production of continuous cast and sheared copper anode plate for electrolytic refining (Contilanod) (Ref 6). Eldred developed a machine using graphite as the mold material to cast copper rods and later a number of copper-base alloys (Ref 2). Poland and Lindner were granted a U.S. patent for a vertical casting machine very similar to Eldred’s (Ref 7). The Asarco process was patented by the American Smelting and Refining Company for the continuous casting of phosphorus-deoxidized copper (Ref 8). The Properzi process was introduced in Italy to continuously cast and roll lead rod used for the manufacture of lead pellets for shotgun cartridges. The plant is used today (2008) for the large-scale production of aluminum rod and copper rod (Ref 9). An inexpensive machine for the continuous casting of bronzes was developed at the Tin Research Institute, England (Ref 10). United Wire, Edinburgh, patented their Unicast system for continuous casting brass and bronze rods (Ref 11). The Swiss company Alfred Wertli introduced the first industrial horizontal continuous caster for the production of cast iron rods and later expanded into continuous casting plants for a full range of copper-base alloys and shapes—rod, billet, and strip (Ref 12). Technica-Guss of Wurzburg, West Germany, introduced horizontal continuous casting systems tailored to individual customer requirements, producing strip, billets, round bars, tubes, and profiles in a range of copper-base alloys (Ref 13). The Southwire Company of Carrolton, GA, introduced continuous casting of copper wire by the Southwire continuous rod system, using a high-speed casting wheel mold (Ref 14). Outokumpu O.Y. of Finland introduced and patented the Outokumpu upward casting process for producing copper rod (Ref 15). Rautomead Dundee introduced horizontal and vertical continuous casting equipment based on the all-graphite system, with the Unicast principle of integrated melt, stabilization, and cast from a single crucible (Ref 16).

1914 Mid-1920s

1930 1938 1930s Late 1930s

1950s Early 1950s 1957

1960s

1964 1969 1978

*Portions adapted with permission from: Robert Wilson, A Practical Approach to Continuous Casting of Copper-Based Alloys and Precious Metals, Institute of Materials (IOM Communications Ltd.), 2000

1020 / Casting of Nonferrous Alloys (2008) is a subsidiary of SMS Meer. The company is a supplier of horizontal and vertical casting plants covering large installations for billet and strip casting to smaller plants for strip, wire, and tube casting of copper alloys and precious metals.

Vertical Continuous Casting (Ref 1, 18)

Fig. 1

Poland and Lindner vertical caster. Source: Ref 1

by American Smelting and Refining Company, was primarily designed for the continuous casting of phosphorus-deoxidized copper, but it is widely used today (2008) for a range of copper-base alloys. The Properzi process (Ref 9) was introduced in Italy in the late 1930s to continuous cast and roll lead rod used for the manufacture of lead pellets for shotgun cartridges. The plant is used today (2008) for the highvolume production of aluminum rod and copper rod. Among the first production vertical casting units to be introduced was the TRI equipment developed by the Tin Research Institute in England in 1950 (Ref 10). The equipment consisted of an induction melting unit that feeds a tundish, which was attached to a graphite mold in a tapered steel water jacket. Withdrawal of the product was by means of two grooved rolls situated below the mold. Following the advent of the TRI system, a number of vertical casting processes appeared, such as the Unicast process introduced by United Wire of Edinburgh. The TRI and Unicast equipment filled a need for equipment to produce tin bronze in the form of rod and tube and also a selection of brasses. Since the 1950s, United Wire plants have been installed worldwide, particularly in Britain, France, Italy, and the United States. Rautomead has become a major supplier of such equipment In 1957, the Swiss company Alfred Wertli (Ref 12) introduced the first industrial horizontal continuous caster for the production of cast iron rods and later expanded into continuous casting plants for a full range of copper-base alloys and shapes—rod, billet, and strip. In the 1960s, Technica-Guss (Ref 13) of Wurzburg, West Germany, introduced horizontal continuous casting systems tailored to individual customer requirements, producing strip, billets, round bars, tubes, and profiles in a range of copper-base alloys. In 1989, Technica became a member of the Mannesmann Demag AG group as Demag Technica GmbH and today

As noted, the Unicast system was developed in the 1950s for vertical continuous casting of bronze and brass rod approximately 16 to 19 mm (0.63 to 0.75 in.) in diameter as feedstock for the manufacture of fine wire mesh used for papermaking machines. The unit consisted of an integrated continuous casting plant in which the complete process of melting, alloying, holding, and casting takes place in one selfcontained furnace (Ref 1). The Unicast furnace differs from the modern version only in that refractory brick insulation is used throughout instead of low-thermal-mass insulation. The integrated melting, homogenizing, and casting in an all-graphite system produces a high-quality product with minimal residualelement impurities and low oxygen level, capable of being drawn to very fine wire. The initial casting machines manufactured and used internally by United Wire were vertical casters of approximately 1 tonne capacity with coiling equipment to give workable coils for subsequent rolling to 5 mm2 (0.008 in.2) and then drawn down to suitable wire sizes. United Wire operates these casting units today (2008) on a range of brasses, bronzes, and nickel-silver alloys, producing rod of high quality. A vertical continuous casting plant for copper alloy tubes and bars consists of a channeltype induction furnace, positioned on a tilting frame and feeding into a water-cooled graphite mold, with microprocessor-controlled withdrawal system and automatic tube cut-off. Mold or die change is made without emptying the holding furnace by incorporating a back-tilting mechanism, which is also a safety feature in vertical casting, because the melting unit can be tilted off the casting position in case of malfunction. The product range in tubes is 20 to 125 mm (0.8 to 5.0 in.) outside diameter and in bars 12 to 80 mm (0.5 to 3.1 in.) diameter. Strand lengths are generally in the range of 3 to 4 m (10 to 13 ft).

Vertical Semicontinuous Slab Casters Since the original invention of Unicast and TRI vertical continuous casting, the most commonly used plant for slab casting is still the proven design and low investment costs of vertical semicontinuous casters (Fig. 2). The layout consists of a casting pit with integrated casting cylinder, by which the casting table and products are moved up and down. The liquid metal can be supplied from the melting or holding

furnace via launder into the molds mounted on top of the oscillating casting machine. It is common for multiple molds to be mounted in the casting table; this allows for very high production rates. At the end of the casting cycle, the produced slabs are pulled out of the pit with an overhead crane and a hydraulic tool for clamping of the slab (Fig. 2b). This disadvantage can be avoided by use of a vertical semicontinuous caster with incorporated discharge device (Fig. 3). This layout shows a larger casting pit with integrated casting cylinder and casting table, to which the discharge device is attached by a swivel joint. The liquid metal solidifies in the mold on top of the oscillating casting machine. When the casting cycle is completed, the cast slabs are pushed from the vertical position slightly to the direction of the discharge device, and the discharging operation starts. The vertical semicontinuous slab casters can be used to cast a wide range of alloys and sizes. In addition to copper and brass, copper nickel and other copper alloys are produced. Modern casters incorporate components such as an automatic melt-level control system to provide automation and stable casting conditions for a consistently good product quality. Also, adjustable slab molds can be used, by which the casting width can be varied and thus the inventory of the molds reduced (Fig. 4). Most often, copper alloy molds without sleeves are used for slab casting.

Vertical Continuous Slab Casting When casting copper slabs, the highest production rates and consistent quality product can be achieved with a fully continuous vertical caster. The general layout is similar to the semicontinuous mode; a holding furnace supplies the liquid metal via an automatic melt-level control into the molds, which are mounted on top of the oscillating casting machine (Fig. 5). The solidified strands pass through the secondary water-cooling box and are transported by a withdrawal unit. After cutting, the products are tilted by a tilting basket into a horizontal position and further transported by a roller conveyor and lifting system. Appropriate sensors detect the melt level in the mold, and the flow of liquid metal is regulated automatically by an electronic control. The holding furnace is typically induction heated and has a specific shape for vertical casting, with an attached launder section or forehearth, thus supplying the metal by the shortest route into the mold without splashing. The solidified slabs are firmly clamped and transported by the withdrawal unit. The vertical movable saw clamps to the strands and travels together with the strands during the cutting operation. After cutting, the products are taken over by a tilting basket and tilted into a horizontal position for further transport on subsequent roller tables.

Copper Continuous Casting / 1021

Fig. 2

Vertical semicontinuous casting plant. (a) Layout without discharge device. (b) Slab discharging by overhead crane. Courtesy of Demag Technica GmbH of SMS

Fig. 4

Adjustable molds for vertical semicontinuous casting of slabs. Courtesy of Demag Technica GmbH of SMS

Upcasting Methods Fig. 3

Vertical semicontinuous casting plant with discharge device. Courtesy of Demag Technica GmbH of SMS

The Outokumpu Upcasting Method. The upward casting process was introduced and patented in 1969 by Outokumpu O.Y., Finland

(Ref 15), with the first production unit coming into operation in 1970 for casting oxygen-free, small-diameter copper rod. This system has all the technical advantages of casting in the vertical mode and, for small-diameter rod, none of the disadvantages. The method consists of a graphite die partially immersed in molten metal, with the upper part surrounded by a water-cooled jacket (Fig. 6) (Ref 1). The assembly is located just above the metal top surface, with the graphite die only just immersed into the liquid and maintained precisely in position by an electronic level-sensing control. The action of vertical pulsed withdrawal of the rod raises the metal beyond the lower extremity of the cooler, and solidification takes place. In the melting and transfer system, Outokumpu exposes the liquid metal to graphite or charcoal, resulting in deoxidation of the melt to a level of the order of 5 ppm oxygen. The machine operates on a multidie system, casting, for example, 12 mm (0.5 in.)

1022 / Casting of Nonferrous Alloys

Fig. 5

Vertical continuous casting. Courtesy of Demag Technica GmbH of SMS

diameter rods at speeds on the order of 3 m/min (10 ft/min). Rautomead Upwards Vertical Casting. Rautomead International, Dundee, introduced a modified upwards vertical casting process (Ref 16) based on graphite melt containment technology and using submersed dies with inert gas protection. The equipment is used primarily for the production of small-diameter, high-purity copper rod with oxygen levels on the order of 5 ppm at casting speeds on the order of 4.0 to 4.5 m/min (13 to 14.5 ft/min). The machine is also adaptable to alloy systems such as bronzes and brasses in rod form and also tube. By utilizing an all-graphite containment system and incorporating a specially designed graphite filter bed, deoxidation of copper to 5 ppm oxygen is ensured. Pressure Upcast System. A pressure Upcaster (Ref 19) was developed as a production unit at Dundee Institute of Technology (now University of Abertay, Dundee). During the casting operation, an inert gas applied to the sealed steel furnace casing exerts pressure on the molten metal in the graphite crucible, raising it into the graphite die where it solidifies and is withdrawn through a water-cooled jacket vertically upward in a conventional pulsed mode. On reverting to atmospheric pressure, metal drains to the crucible. The equipment is primarily intended for casting small-section high-purity copper rod in the range of 1.5 to 10 mm (0.06 to 0.40 in.) in diameter. A process using some of the same principles was explored in the 1970s by Chase Brass to produce copper and brass rod. A higher throughput was targeted, and the cast rod was reduced in-line to wire. The process remains in use to produce copper bus bar.

Horizontal Continuous Casting

Fig. 6

Principle of upward casting. Source: Ref 1

Casting in the horizontal mode facilitates product handling and generally occupies less space. There are inherent problems as opposed to vertical casting that mainly relate to gravity-induced directional cooling, which, in most cases, can be accommodated in the process protocols. The horizontal plants produce billets for subsequent extrusion in copper or brass in section sizes between 80 and 400 mm (3.1 and 16 in.) in diameter, operating as single- or multistrand machines. Smaller horizontal casters are used for the production of tube, bar, and sections in a full range of copper alloys. The Swiss company Alfred Wertli (Ref 12) introduced the world’s first industrial horizontal continuous caster for the production of cast iron rods and later expanded into continuous casting plants for a full range of copper-base alloys. State-of-the-art horizontal continuous casting lines for copper alloys are mostly designed to cast two narrow strips up to 450 mm (18 in.) wide or one strip up to 800 mm (32 in.) wide. Thin-strip withdrawal machines are designed to cast two strands simultaneously, or each strand can be independently withdrawn.

Copper Continuous Casting / 1023 The molds can also be configured for billet casting 100 to 400 mm (9 to 16 in.) in diameter, bar and tube casting in the size range of 25 to 350 mm (1.0 to 14 in.) in diameter, and for small-diameter rod and wire 12 to 25 mm (0.5 to 1.0 in.) in diameter. The original Wertli (Ref 12) concept consists of a channel-type induction furnace and holding furnace, together with graphite die and cooler assembly and runout track with withdrawal machine and cut-off device. Molten metal flows from the melting furnace to a holding or casting furnace that acts as a reservoir of molten metal, maintaining the required casting temperature. Water-cooled graphite dies are attached to the holding furnace or crucible. During the continuous casting operation, metal flows into the graphite casting die, where it solidifies. The solidified strands are intermittently withdrawn in a pull-pause sequence by means of withdrawal equipment. After leaving the graphite die, which is housed within the primary cooler, the cast strands pass through a secondary cooler in the form of a water sparge that removes the surplus heat contained in the solidified casting. The crucible can be manufactured in a refractory ceramic or from graphite. Integral ceramic crucibles are used extensively in induction melting and casting furnaces. These are the most energy-efficient furnaces and consist of melting units feeding a casting unit or a single induction-heated casting unit. The design varies depending on the application. The metal type and production rate will determine the crucible capacity and power rating. Frequency is chosen to suit these parameters and is selected from 150, 250, 500, 1000, 3000, and 10,000 Hz. The high frequencies apply to small crucible capacity, decreasing for the larger installations. Induction melting and casting furnaces use either integral or removable crucible assemblies, depending on the casting operation. Precast ceramic crucibles with graphite support carriers are used in either induction-heated or resistance-heated furnaces (Ref 1). The full range of Rautomead graphite resistance-heated furnaces use this basic design, ranging from small table-top units to installations with crucible capacity of 2500 kg (5500 lb) (copper). Graphite can only be used in a nonoxidizing atmosphere; therefore, the crucible and die assembly must be housed in a sealed furnace and protected with an inert gas, either nitrogen or argon. Most high-grade coppers, brasses, tin bronzes, phosphor bronzes, and aluminum bronzes can be successfully cast in an allgraphite crucible and die assembly (Ref 1). The volume of the crucible is dependent on the application and may vary in capacity from several tonnes to 1 kg (2 lb) or less. A crucible liner or protecting sleeve is frequently fitted, particularly with larger crucibles. This liner is manufactured in graphite and protects the main crucible against abrasion and oxidation. The seal between crucible and graphite die is often made by means of a grafoil gasket sheet or washer. Grafoil consists of graphite

in flexible lamellar form, which is compressible, forming a gastight seal and providing a liquidtight seal. Graphite baffles can be fitted within the crucible and held in position between the lower and upper graphite sleeves or liners (Ref 1). A baffle with suitable perforations provides an upper and lower chamber to facilitate melting and homogenization of the charge in the upper section prior to this metal entering the casting die. Another, most important function is to allow sufficient time for deoxidation of the melt and thus avoid attack on the graphite die. Unicast Horizontal Casting System. In the early 1970s, the first Unicast horizontal casting plant was installed by Timex Corporation, Dundee, for continuous casting brass rod for watch case manufacture, operating under conditions similar to the United Wire vertical casters. This operation was extremely successful, utilizing as feedstock 100% internally generated brass scrap. The recycled scrap from trim and machining operations on case manufacturing, together with high-quality press shop and screw machine residue, made the process economically viable. The chemistry of the product could be closely controlled, reducing trace element impurities to limits unobtainable on purchased stock. Since its inception in 1978, Rautomead, Dundee, has manufactured a wide range of continuous casting machines for the nonferrous metals industries, primarily for copper-base alloys and precious metals. The Rautomead resistanceheated all-graphite system is based on the United Wire Unicast technology. The design has been refined, particularly in the area of refractory insulation heating element configuration and unit modular construction. The majority of the machines operate in the horizontal mode, with a few special-purpose machines casting vertically downward. The construction of the Rautomead machines is an integrated all-graphite melt-and-cast system. The casters range from small table-top units with crucible capacities of 2 to 50 kg (4.5 to 110 lb) (copper) to large billet and strip casters with crucible capacities to 2500 kg (5500 lb) (copper).

Strip Casting The horizontal strip casting process was originally developed by Technica-Guss, which is now part of SMS Meer. It has been widely used in the industry for more than 30 years as a nearnet shape casting process for alloys that are considered difficult to hot roll, such as nickelsilver mint alloys and phosphor bronze for electronic applications. Horizontal continuous casters are manufactured for casting a range of strip widths. Typical strip widths are 450 mm (18 in.), 650 mm (26 in.), or larger, with a common thickness being between 14 and 20 mm (0.6 and 0.8 in.). The produced coil can be automatically discharged onto a subsequent discharge table.

A majority of horizontal strip casters are equipped with an in-line milling machine by which the top and bottom of the cast surface is cleaned of oxides and segregations so that a perfect coil is produced for direct passing into the cold rolling mill. Modern strip casters comprise high-performance coolers with temperature-sensoring systems to survey the solidification and apply nitrogen for reduction of surface oxides. Furthermore, a tight temperature control in the holding furnace is essential for stable casting conditions and product quality, which is realized with the stepless power control of the inductor by the inverter technique. A typical casting speed of a horizontal strip caster is up to 200 mm/min (8 in./min), resulting in a production of approximately 6,000 tons/year for strips 650 by 16 mm (26 by 0.6 in.). Due to the limited production rate, many companies who are using this casting process operate several machines; even eight machines are in operation in one company. Horizontal strip casters are used exclusively for the production of copper alloys and, in very few exceptions, worldwide for copper. To overcome the limitations of the thin-strip casting process regarding output and copper casting, SMS Meer developed a new concept for a vertical strip caster (Ref 18). Wertli Strip Casting. The Wertli strip casting plants include in-line operations such as milling equipment, traveling shear, coiler, and die and track friction. Hydraulically amplified electric drives are used to achieve forces in the range of 40 to 80 kN (9000 to 18,000 lbf) while maintaining motion accuracy. The Wertli drive concept is designed to handle such forces and accelerations by using backlash-free lowratio gears together with a high-precision servo motor with hydraulic amplification. To achieve a mechanically backlash-free drive, backlashfree gears are used between the driving motor and the rollers that drive the strands. Slippage between the cast strand and the drive roller is to be avoided if a precise strand motion must be maintained over long periods of casting. The tight gripping of the strands is achieved by hydraulic press-down cylinders. A cooling system with an enhanced water-cooled surface increases heat transfer (approximately 10%) compared to a conventional copper plate cooler (Ref 1). Hazelett Casting Process. The Hazelett steel belt casting invented in 1920 has been developed for various product forms and today (2008) is used in the production of copper wire, rod, strip, and anode. The metal is usually melted by induction and is delivered via a tundish to a straight-through mold formed by tensioned steel belts and edge dam blocks. Fast-film heat extraction from the mold is achieved by a proprietary design for the application and removal of high-flow-rate water cooling. The use of special coatings on the belt is also important. Strip up to 1.25 m (4.10 ft) wide and Contirod, a rectangular cast bar at

1024 / Casting of Nonferrous Alloys 6 to 60 tonnes/h depending on plant capacity, can be achieved. The Contilanode process is used for producing high-quality copper anode. The cast anode plate can be maintained geometrically to within close limits. Hanger lugs are cast in shaped recesses and thus become an integral part of the anode body.

Wheel Casting Properzi Wheel Casting Technology. Properzi first used wheel casting technology on copper in the 1950s and commercially introduced the continuous casting and rolling process for copper rod in 1963. Molten copper is poured into a revolving casting wheel from a gas-fired melting and refining furnace. The copper rim of the wheel is grooved to receive the molten metal, which is then retained in the groove by a steel belt. The solidified metal leaves the wheel and passes through a rolling mill without interruption. The casting wheel has a “U” profile, a shape that evolved to control solidification and heat transfer as the metal traversed the cooling segments of the wheel. A Cu-Cr-Zr alloy mold is used for the casting of all electrolytic tough pitch (ETP) copper. The position, alignment, and adjustment of the individual cooling spray nozzles located around the wheel are of the utmost importance in controlling the solidification and uniformity of the grain structure of the cast bar. A layer of acetylene soot, applied to both cavity and band, serves as a release agent and insulator, which provides uniformity of heat transfer. During each rotation of the wheel, the soot is stripped by high-pressure water sprays, then reapplied. The cast “D” section passes through two twohigh break-down roll stands, followed by six to eight three-high roll stands to yield product for rod, narrow strip, trolley wire, and other applications. Southwire Continuous Casting Rod Process. Following some research and development with Properzi in approximately 1960, the Southwire Company of Georgia, United States, introduced a continuous casting process for the high-speed production of ETP copper rod. The process has been described in detail in several published papers (Ref 20–23). The SCR process, as it is known, incorporates a continuous melting, holding, casting, rolling, pickling, and coiling system. The casting wheel provides a trapezoidalshaped casting groove in the periphery of a copper alloy ring. This ring is closed by an endless steel belt through an arc of approximately 180 to 210 , the belt being held in place by idler wheels and tensioners. The casting groove and the contact side of the steel band are coated with a controlled layer of soot that serves as a release agent and provides uniformity of heat extraction. The cast bar passes to the rolling mill through a trimming and descaling

Fig. 7

Schematic of Ohno continuous casting. Source: Ref 1

operation. The mill itself is composed of a number of roughing, intermediate, and finishing two-roll stands. The alternating vertical and horizontal shaft stands produce a repetitive series of oval-to-round reductions. In the continuous casting of ETP copper with oxygen content in the melt of approximately 400 ppm, the dissolved oxygen reacts with the impurities present during solidification, precipitating these out of the solid solution and resulting in improved annealability and electrical conductivity of the product. The claim for success of the SRC process is the ability to control the amount of superheat to a very close range immediately prior to casting, generally approximately 25  C (45  F) above the liquidus. The use of a high-purity cathode and the close control of temperature results in solidification in a columnar grain pattern with good bar quality. In subsequent rolling in the SCR process, the high temperature and severe initial reductions in the first pass cause dynamic recrystallization. Chia (Ref 24) described the mode of solidification on SCR tough pitch copper rod.

Ohno Continuous Casting Process The Ohno continuous casting concept is based on the application of the Ohno separation theory of solidification (Ref 25), with the continuous cast ingot consisting of unidirectional solidified structure with no equiaxed crystals. The process is described by Ohno and McLean (Ref 26). The patented process (Ref 27) differs from conventional techniques in that molten metal is poured into a heated mold rather than into a cooled mold or die. The mold is heated externally and its temperature maintained above the solidification point of the metal being cast. As a result, no metal nucleates on the mold surface. The Ohno process has been adopted by Furukawa, Japan (Ref 28), for the production of

oxygen-free high-purity copper rod. The rod has a structure characterized by longitudinal crystals or may even develop into a single crystal in some growth conditions. This special structural material is used in high-resolution audio signal transmission, having low impurities, no grain boundaries transverse to the direction of signal transmission, smooth surface finish, and excellent physical properties. The production casting equipment used is essentially as shown in Fig. 7, with some refinements, including a melting furnace and a casting furnace with precise metal-level control, ensuring constant metastatic pressure on the solidification front. The high-purity copper charge is deoxidized using carbonaceous material in the melting furnace before transfer to the casting furnace. REFERENCES 1. R. Wilson, A Practical Approach to Continuous Casting of Copper-Based Alloys and Precious Metals, Institute of Materials (IOM Communications Ltd.), 2000 2. B.E. Eldred, U.S. Patent 1,868,099, 1932 3. G.E. Sellers, U.S. Patent 1908, 1840 4. L. Laign, U.S. Patent 3023, 1843 5. A.H. Pehrson, U.S. Patent 1,088,171, 1914 6. Hazelett Process, Iron Steel Eng., Vol 43 (No. 6), 1966, p 105 7. Poland and Lindner, U.S. Patent 2,136,394, 1938 8. A. Kreil et al., Asarco Process, Met. Rev., Vol 5, 1960, p 413–446 9. Properzi Process, Met. Rev., Vol 6 (No. 22), 1961 10. E.C. Ellwood, J. Inst. Met., Vol 84, 1955– 1956, p 319–326 11. I.E. Ewen, United Wire Unicast Process, U.K. Patents 894,783, 894,784, and 934,484 12. T.P. Wertu, Alfred Wertli AG, Winterthur, Switzerland 13. Technica Guss, Wurzburg, West Germany

Copper Continuous Casting / 1025 14. Southwire Revolutionizes Non-Ferrous Rod Production with SCR System, Vol. 33, Southwire Mag., June 1975 15. M. Rantaneno, Upward Continuous Cast of Copper Wire, Wire Ind., July 1976 16. Rautomead International, Dundee, Scotland 17. A. Krell et al., Flocast Process, Met. Rev., Vol 5, 1960, p 413–446 18. “Copper and Copper Alloy Rolled Products—Trends in Equipment/Auxiliaries and Applications,” Seminor at Hotel Le Royal Meridiea (Mumbai), Jan 18–19, 2006 19. R. Wilson, Pressure Upcast, U.K. Patent GB 2,236,498B, 1992, and U.S. Patent 5,090,471, 1992

20. U. Sinha and R. Adams, Southwire Continuous Rod Process: Innovations for Quality Improvements, Wire J. Int., June 1993 21. U. Sinha and R. Adams, “Southwire Continuous Rod: A Method to Produce HighQuality Rods for Fine Wire Drawing and Special Applications,” Conference Indian Copper Development Centre and Winding Wires Manufacturers Association of India, Oct 1988 22. G.T. Hudson, “The Production of Copper Rod by SRC Process,” internal paper, Southwire Company, Carrolton, GA. 23. L.C. Richards et al., “Continuous Casting— Its History, Impact and Future,” Metals Week Copper Conference, Dec 10, 1989

24. H. Chia, International Conf., Inst. Wire and Mach. Assoc. (Torremolinos, Spain), April 1979 25. A. Ohno, Solidification, The Separation Theory and Its Practical Application, Springer Verlag, New York 26. A. Ohno and A. McLean, Ohno Continuous Casting, Adv. Mater. Process., Vol 4, 1995, p 43–45 27. A. Ohno, Japan Patent 1,049,148; U.S. Patent 4,515,204; and Germany Patent 3,246,470 28. K. Nakano, “Continuous Casting of Copper and Copper Alloys,” R&D Division of Furukawa Electric Company, Japan

ASM Handbook, Volume 15: Casting ASM Handbook Committee, p 1026-1048 DOI: 10.1361/asmhba0005303

Copyright © 2008 ASM International® All rights reserved. www.asminternational.org

Casting of Copper and Copper Alloys ALL COPPER ALLOYS can be successfully cast in sand. Sand casting allows the greatest flexibility in casting size and shape and is the most economical casting method if only a few castings are made (die casting is more economical above
50,000 units). Virtually all copper alloys can be cast successfully by the centrifugal casting process. Castings of almost every size from less than 100 g to more than 22,000 kg (50,000 lb) have been made. Permanent mold casting is best suited for tin, silicon, aluminum and manganese bronzes, and yellow brasses. Dies casting is well suited for yellow brasses, but increasing amounts of permanent mold alloys are also being die cast. Size is a definite limitation for both methods, although large slabs weighing as much as 4500 kg (10,000 lb) have been cast in permanent molds. Brass die castings generally weigh less than 0.2 kg (0.5 lb) and seldom exceed 0.9 kg (2 lb). The limitation of size is due to the reduced die life with larger castings. Because of their low lead contents, aluminum bronzes, yellow brasses, manganese bronzes, low-nickel bronzes, and silicon brasses and bronzes are best adapted to plaster mold casting. For most of these alloys, lead should be held to a minimum because it reacts with the calcium sulfate in the plaster, resulting in discoloration of the surface of the casting and increased cleaning and machining costs. Size is a limitation on plaster mold casting, although aluminum bronze castings that weigh as little as 100 g (0.25 lb) have been made by the lost wax process, and castings that weigh more than 150 kg (330 lb) have been made by conventional plaster molding. This article describes the casting characteristics and practices of copper alloys. Successful production of copper and copper alloy castings depends on three important factors:  An understanding of casting and solidifica-

tion characteristics of copper and its various alloys  Adherence to proper foundry practices, including melting practices (e.g., selection of melting furnace and molten-metal treatments), pouring practices, and gating and risering techniques  Proper selection of the casting process, which, in turn, depends on the size, shape, and technical requirements of the product

This article addresses each of these factors. For more detailed information, see Ref 1 and the other references listed at the end of this article.

Castability The castability of alloys is generally influenced by their shrinkage characteristics and their freezing range (which is not necessarily related directly to shrinkage). Castability should not be confused with fluidity, which is only a measure of the distance to which a metal will flow before solidifying. Fluidity is thus one factor determining the ability of a molten alloy to completely fill a mold cavity in every detail. Castability, on the other hand, is a general term relating to the ability to reproduce fine detail on a surface. Colloquially, good castability refers to the ease with which an alloy responds to ordinary foundry practice without requiring special techniques for gating, risering, melting, sand conditioning, or any of the other factors involved in making good castings. High fluidity often ensures good castability, but it is not solely responsible for that quality in a casting alloy. Table 1 presents foundry characteristics of selected standard alloys, including a comparative ranking of both fluidity and overall

castability for sand casting; number 1 represents the highest castability or fluidity ranking. The effect of the freezing range on castability is discussed in the next section, “Control of Solidification.” Copper alloys are classified as high-shrinkage or low-shrinkage alloys. The former class includes the manganese bronzes, aluminum bronzes, silicon bronzes, silicon brasses, and some nickel-silvers. They are more fluid than the low-shrinkage red brasses, more easily poured, and give high-grade castings in the sand, permanent mold, plaster, die, and centrifugal casting processes. With high-shrinkage alloys, careful design is necessary to promote directional solidification, avoid abrupt changes in cross section, avoid notches (by using generous fillets), and properly place gates and risers; all of these design precautions help avoid internal shrinks and cracks. Turbulent pouring must be avoided to prevent the formation of dross becoming entrapped in the casting. Liberal use of risers or exothermic compounds ensures adequate molten metal to feed all sections of the casting. Pure copper is extremely difficult to cast, as well as being prone to surface cracking and porosity problems. Casting characteristics of copper can be improved by small additions of elements such as beryllium, silicon, nickel, tin, zinc, chromium, and silver. The copper-base

Table 1 Foundry properties of the principal copper alloys for sand casting UNS No.

Common name

Shrinkage allowance, %

C83600 C84400 C84800 C85400 C85800 C86300 C86500 C87200 C87500 C90300 C92200 C93700 C94300 C95300 C95800 C97600 C97800

Leaded red brass Leaded semired brass Leaded semired brass Leaded yellow brass Yellow brass Manganese bronze Manganese bronze Silicon bronze Silicon brass Tin bronze Leaded tin bronze High-lead tin bronze High-lead tin bronze Aluminum bronze Aluminum bronze Nickel-silver Nickel-silver

5.7 2.0 1.4 1.5–1.8 2.0 2.3 1.9 1.8–2.0 1.9 1.5–1.8 1.5 2.0 1.5 1.6 1.6 2.0 1.6

Approximate liquidus temperature 

C

1010 980 955 940 925 920 880 ... 915 980 990 930 925 1045 1060 1145 1180



F

Castability rating(a)

Fluidity rating(a)

1850 1795 1750 1725 1700 1690 1615 ... 1680 1795 1810 1705 1700 1910 1940 2090 2160

2 2 2 4 4 5 4 5 4 3 3 2 6 8 8 8 8

6 6 6 3 3 2 2 3 1 6 6 6 7 3 3 7 7

(a) Relative rating for casting in sand molds. The alloys are ranked from 1 to 8 in both overall castability and fluidity; 1 is the highest or best possible rating.

Casting of Copper and Copper Alloys / 1027 casting alloy family can be subdivided into three groups according to solidification (freezing) range. The three groups are as follows. Group I alloys are alloys that have a narrow freezing range, that is, a range of 50  C (90  F) between the liquidus and solidus. These are the yellow brasses, manganese and aluminum bronzes, nickel bronzes, (nickel-silvers), manganese (white) brass alloys, chromium-copper, and copper. Liquidus/solidus temperatures for these alloys are shown in Table 2. Group II alloys are those that have an intermediate freezing range, that is, a freezing range of 50 to 110  C (90 to 200  F) between the liquidus and solidus. These are the berylliumcoppers, silicon bronzes, silicon brass, and copper-nickel alloys. Liquidus/solidus temperatures for these alloys are shown in Table 3.

Table 2 Solidification ranges for group I copper alloys Liquidus temperature Alloy type



UNS No.

Copper Chrome copper Yellow brass

C81100 C81500 C85200 C85400 C85700 C85800 C87900 Manganese bronze C86200 C86300 C86400 C86500 C86700 C86800 Aluminum bronze C95200 C95300 C95400 C95410 C95500 C95600 C95700 C95800 Nickel bronze C97300 C97600 C97800 White brass C99700 C99750

C

1083 1085 941 941 941 899 926 941 923 880 880 880 900 1045 1045 1038 1038 1054 1004 990 1060 1040 1143 1180 902 843



F

1981 1985 1725 1725 1725 1650 1700 1725 1693 1616 1616 1616 1652 1913 1913 1900 1900 1930 1840 1814 1940 1904 2089 2156 1655 1550

Solidus temperature 

C

1064 1075 927 927 913 871 900 899 885 862 862 862 880 1042 1040 1027 1027 1038 982 950 1043 1010 1108 1140 879 818



F

1948 1967 1700 1700 1675 1600 1650 1650 1625 1583 1583 1583 1616 1907 1904 1880 1880 1900 1800 1742 1910 1850 2027 2084 1615 1505

Group III alloys have a wide freezing range, well over 110  C (200  F), even up to 170  C (300  F). These are the leaded red and semired brasses, tin and leaded tin bronzes, and highleaded tin bronzes. Liquidus/solidus temperatures for these alloys are shown in Table 4. Control of Solidification. Production of consistently sound castings requires an understanding of the solidification characteristics of the alloys as well as knowledge of relative magnitudes of shrinkage. The actual amount of contraction during solidification does not differ greatly from alloy to alloy. Its distribution, however, is a function of the freezing range and the temperature gradient in critical sections. Manganese and aluminum bronzes are similar to steel in that their freezing ranges are quite narrow— approximately 40 and 14  C (70 and 25  F), respectively. Large castings can be made by the same conventional methods used for steel, as long as proper attention is given to placement of gates and risers—both those for controlling directional solidification and those for feeding the primary central shrinkage cavity. Tin bronzes have wider freezing ranges (
165  C, or 300  F, for C83600). Alloys with such wide freezing ranges form a mushy zone during solidification, resulting in interdendritic shrinkage or microshrinkage. Because feeding cannot take place properly under these conditions, porosity results in the affected sections. In overcoming this effect, design and riser placement, plus the use of chills, are important. Another means of overcoming interdendritic shrinkage is to maintain close temperature control of the metal during pouring and to provide for rapid solidification. These requirements limit section thickness and pouring temperatures, and this practice requires a gating system that will ensure directional solidification. Sections up to 25 mm (1 in.) in thickness are routinely cast. Sections up to 50 mm (2 in.) thick Table 4 Solidification ranges for group III copper alloys Liquidus temperature Alloy type

Table 3 Solidification ranges for group II copper alloys

Alloy type

UNS No.

Beryllium-copper C81400 C82000 C82200 C82400 C82500 C82600 C82800 Silicon brass C87500 Silicon bronze C87300 C87600 C87610 C87800 Copper-nickel C96200 C96400

Liquidus temperature

Solidus temperature





C

1093 1088 1116 996 982 954 932 916 916 971 971 916 1149 1238



F

2000 1990 2040 1825 1800 1750 1710 1680 1680 1780 1780 1680 2100 2260

C

1066 971 1038 899 857 857 885 821 821 860 860 821 1099 1171



F

1950 1780 1900 1650 1575 1575 1625 1510 1510 1580 1580 1510 2010 2140

Leaded red brass

UNS No.

C83450 C83600 C83800 Leaded semired brass C84400 C84800 Tin bronze C90300 C90500 C90700 C91100 C91300 Leaded tin bronze C92200 C92300 C92600 C92700 High-leaded tin bronze C92900 C93200 C93400 C93500 C93700 C93800 C94300



C

1015 1010 1004 1004 954 1000 999 999 950 889 988 999 982 982 1031 977 ... 999 929 943 ...



F

1860 1850 1840 1840 1750 1832 1830 1830 1742 1632 1810 1830 1800 1800 1887 1790 ... 1830 1705 1730 ...

Solidus temperature 

C

860 854 843 843 832 854 854 831 818 818 826 854 843 843 857 854 ... 854 762 854 900



F

1580 1570 1550 1550 1530 1570 1570 1528 1505 1505 1518 1570 1550 1550 1575 1570 ... 1570 1403 1570 1650

can be cast, but only with difficulty and under carefully controlled conditions. A bronze with a narrow solidification (freezing) range and good directional solidification characteristics is recommended for castings having section thicknesses greater than approximately 25 mm (1 in.). It is difficult to achieve directional solidification in complex castings. The most effective and most easily used device is the chill. For irregular sections, chills must be shaped to fit the contour of the section of the mold in which they are placed. Insulating pads and riser sleeves sometimes are effective in slowing down the solidification rate in certain areas to maintain directional solidification.

Melting and Melt Control Furnaces for melting copper casting alloys are either fuel fired or electrically heated. They are broadly classified into three categories:  Crucible furnaces (tilting or stationary)  Open-flame (reverberatory) furnaces  Induction furnaces (core or coreless)

Selection of a furnace depends on the quantity of metal to be melted, the degree of purity required, and the variety of alloys to be melted. Environmental restrictions also influence furnace selection. In addition, the melting of copper alloys also involves various melt treatment to improve melt quality. These melt treatments include:     

Fluxing and metal refining Degassing Deoxidation Grain refining Filtration

Each of these is described in the sections that follow. It should be noted that some of these process methodologies pertain not only to foundry melting and casting but also to smelting, refining, and, in certain cases, mill product operations. Fuel-Fired Furnaces. Copper-base alloys are melted in oil- and gas-fired crucible and openflame furnaces. Crucible furnaces, either tilting or stationary, incorporate a removable cover or lid for removal of the crucible, which is transported to the pouring area where the molds are poured. The contents of the tilting furnace are poured into a ladle, which is then used to pour the molds (Fig. 1, 2). These furnaces melt the raw materials by burning oil or gas with sufficient air to achieve complete combustion. The heat from the burner heats the crucible by conduction and convection; the charge melts and then is superheated to a particular temperature at which either the crucible is removed or the furnace is tilted to pour into a ladle. While the molten metal is in the crucible or ladle, it is skimmed, fluxed, and transferred to the pouring area, where the molds are poured.

1028 / Casting of Nonferrous Alloys The other type of fuel-fired furnace is the open-flame furnace, which is usually a large rotary-type furnace with a refractory-lined steel shell containing a burner at one end and a flue at the other. The furnace is rotated slowly around the horizontal axis, and the rotary movement helps to heat and melt the furnace charge. Melting is accomplished both by the action of the flame directly on the metal and by heat transfer from the hot lining as this shell rotates. These furnaces usually tilt so that they can be charged and poured from the flue opening. At the present time (2008), these furnaces are not used often because of the requirement that a baghouse be installed to capture all the flue dust emitted during melting and superheating. While these furnaces are able to melt large amounts of metal quickly, there is a need for operator skill to control the melting atmosphere within the furnace. Also, the refractory walls become impregnated with the melting metal, causing a contamination problem when switching from one alloy family to another.

Fig. 1

Typical lift-out type of fuel-fired crucible furnace, especially well adapted to foundry melting of smaller quantities of copper alloys (usually less than 140 kg, or 300 lb)

Fig. 2

Typical lip-axis tilting crucible furnace used for fuel-fired furnace melting of copper alloys. Similar furnaces are available that tilt on a central axis

Electric Induction Furnaces. In the past 30 years, there has been a marked changeover from fuel-fired melting to electric induction melting in the copper-base foundry industry. While this type of melting equipment has been available for more than 60 years, very few were actually used due to the large investment required for the capital equipment. Because of higher prices and the question of availability of fossil fuels and because of new regulations on health and safety imposed by the Occupational Safety and Health Administration (OSHA), many foundries have made the changeover to electric induction furnaces. When melting alloys in group III, fumes of lead and zinc are given off during melting and superheating. The emission of these harmful oxides is much lower when the charge is melted in an induction furnace because the duration of the melting cycle is only approximately 25% as long when melting the same amount of metal in a fuel-fired furnace. By the use of electric induction melting, compliance with OSHA regulations can be met in many foundries without the need for expensive air pollution control equipment. The two types of electric induction furnaces are the core type, better known as the channel furnace, and the coreless type. Core Type. This furnace (Fig. 3) is a large furnace used in foundries for pouring large quantities of one alloy when a constant source of molten metal is required. This furnace has a primary coil, interfaced with a laminated iron core, surrounded by a secondary channel, which is embedded in a V- or U-shaped refractory lining located at the bottom of a cylindrical hearth. Here, the channel forms the secondary of a transformer circuit. This furnace stirs and circulates molten metal through the channel at all times, except when the furnace is emptied and shut down. When starting up, molten metal must be poured into the furnace to fill up the “heel” on the bottom of the bath. Because these furnaces are very efficient and simple to

Fig. 3

Cutaway drawing of a twin-channel induction melting furnace

operate, with lining life in the millions of pounds poured, they are best suited for continuous production runs in foundries making plumbing alloys of group III. They are not recommended for the dross-forming alloys of group I. The channel furnace is at its best when an inert, floating, cover flux is used and charges of ingot, clean remelt, and clean and dry turnings are added periodically. Coreless Type. This furnace has become the most popular melting unit in the copper alloy foundry industry. In earlier years, the coreless furnace was powered by a motor generator unit, usually at 980 Hz. The present coreless induction furnaces draw 440 V, 60 cycle power and, by means of solid-state electronic devices, convert the power to 440 V and 1000 or 3000 Hz. These furnaces are either tilting furnaces (Fig. 4) or crucible lift-out units (Fig. 5, 6). A coreless induction furnace comprises a water-cooled copper coil in a furnace box made of steel or Transite. The metal is contained in a crucible or in a refractory lining rammed up to the coil. Crucibles used in these furnaces are made of clay graphite; silicon carbide crucibles cannot be used because they become overheated when inserted in a magnetic field. Clay graphite crucibles do a good job of conducting the electromagnetic currents from the coil into the metal being melted. Induction furnaces are characterized by electromagnetic stirring of the metal bath. Because the amount of stirring is affected by both power input and power frequency, the power unit size and frequency should be coordinated with the furnace size in order to obtain the optimal-sized equipment for the specific operation. In general, the smaller the unit, the higher the frequency and the lower the power input. Large tilting units are used in foundries requiring large amounts of metal at one time. These furnaces, if over 4.5 Mg (10,000 lb) capacity, operate at line frequency (60 Hz). They are very efficient and will melt large quantities of metal in a very short time if powered with the propersized power unit. Stationary lift-out furnaces are often designed as shown in Fig. 5. Here, the crucible sits on a refractory pedestal, which can be

Fig. 4

Cross section of a tilt furnace for high-frequency induction melting of brass and bronze alloys. Crucible is of clay graphite composition.

Casting of Copper and Copper Alloys / 1029

Fig. 5

Cross section of a double push-out furnace. Bilge crucibles are placed on refractory pedestals and raised and lowered into position within the coils by hydraulic cylinders.

metal in which chemical compounds or mixtures of such compounds are employed. These compounds are usually inorganic. In some cases, metallic salts are used in powder, granulated, or solid tablet form and may often melt to form a liquid when used. They can be added manually or can be automatically injected, and they can perform single or, in combination, various functions, including degassing, cleaning, alloying, oxidation, deoxidation, or refining. The term fluxing also includes the treatment of nonferrous melts by inert or reactive gases to remove solid or gaseous impurities. Fluxing practice in copper alloy melting and casting encompasses a variety of different fluxing materials and functions. Fluxes are specifically used to remove gas or prevent its absorption into the melt, to reduce metal loss, to remove specific impurities and nonmetallic inclusions, to refine metallic constituents, or to lubricate and control surface structure in the semicontinuous casting of mill alloys. The last item is included because even these fluxes fall under the definition of inorganic chemical compounds used to treat molten metal.

Types of Fluxes

Fig. 6

Foundry installation of high-frequency induction lift swing furnaces

raised or lowered by a hydraulic cylinder. This unit, also called a push-out furnace, operates by lowering the crucible into the coil for melting and then raising the crucible out of the coil for pickup and pouring. While one crucible is melting, the other crucible can be charged and ready to melt when the knife switch is pulled as the completed heat is being pushed up for skimming and pouring. The other common type of coreless induction melting is the lift swing furnace (Fig. 6). Here, the coil (and box) is cantilevered from a center post to move up or down vertically and swing horizontally about the post in a 90 arc. Because there are two crucible positions, one

crucible can be poured, recharged, and placed into position to melt, while the other is melting. When the metal is ready to pour, the furnace box is lifted (by hydraulic or air cylinder), pivoted to the side, and lowered over the second crucible. The ready crucible is then standing free and can be picked up and poured, while melting is taking place in the second furnace.

Fluxing of Copper Alloys The term fluxing is used in this article to represent all additives to, and treatments of, molten

Fluxes for copper alloys fall into five basic categories: oxidizing fluxes, neutral cover fluxes, reducing fluxes (usually graphite or charcoal), refining fluxes, and semicontinuous casting mold fluxes. Oxidizing fluxes are used in the oxidationdeoxidation process; the principal function here is control of hydrogen gas content. This technique is still practiced in melting copper alloys in fuel-fired crucible furnaces, where the products of combustion are usually incompletely reacted and thus lead to hydrogen absorption and potential steam reaction (see the section “Degassing of Copper Alloys” in this article). The oxidizing fluxes usually include cupric oxide or manganese dioxide (MnO2), which decompose at copper alloy melting temperatures to generate the oxygen required. Figure 7 illustrates the effectiveness of oxidizing fluxes in reducing porosity due to hydrogen and in improving mechanical properties for a tin bronze alloy. Neutral cover fluxes are used to reduce metal loss by providing a fluid cover. Fluxes of this type are usually based on borax, boric acid, or glass, which melts at copper alloy melting temperatures to provide a fluid slag cover. Borax melts at approximately 740  C (1365  F). Such glassy fluxes are especially effective when used with zinc-containing alloys, preventing zinc flaring and reducing subsequent zinc loss by 3 to 10%. The glassy fluid cover fluxes also agglomerate and absorb nonmetallic impurities from the charge (oxides, molding sand, machining lubricants, and so on). As with aluminum alloys, fluxes containing reactive fluoride salts (CaF2 and NaF) can strip oxide films in copper-base alloys, thus permitting entrained

1030 / Casting of Nonferrous Alloys metallic droplets to return to the melt phase. Table 5 indicates the effectiveness of this type of flux in reducing melt loss in yellow and high-tensile brass. For red brasses, however, it may not be proper to use a glassy flux cover, because such a cover will prevent or limit beneficial oxidation of the melt (see the section “Degassing of Copper Alloys” in this article). Use of a glassy cover flux can sometimes result in reduced alloy properties (Ref 2). Oxide films in aluminum and silicon bronzes also reduce fluidity and mechanical properties. Fluxes containing fluorides, chlorides, silica, and borax provide both covering and cleaning, along with the ability to dissolve and collect these objectionable oxide skins. Chromium and beryllium-copper alloys oxidize readily when molten; therefore, glassy cover fluxes and fluoride salt components are useful here in controlling melt loss and achieving good separation of oxides from the melt. Reducing fluxes containing carbonaceous materials such as charcoal or graphite are used on higher-copper lower-zinc alloys. Their principal advantage lies in reducing oxygen absorption of the copper and reducing melt loss. Lowsulfur, dry, carbonaceous flux materials should always be used with copper alloys to avoid gaseous reactions with sulfur or with hydrogen from contained moisture. However, carbonaceous materials will not agglomerate

Fig. 7

nonmetallic residues or provide any cleaning action when melting fine or dirty scrap. For this reason, a glassy cover must also be used in the latter case. Table 5 indicates the beneficial effects of a glassy cover when melting brass turnings. Melt-Refining Fluxes. It is possible to remove many metallic impurity constituents from copper alloys through the judicious use of fire refining (oxidation). According to standard free energy of reaction, elements such as iron, tin, aluminum, silicon, zinc, and lead are preferentially oxidized before copper during fire refining (Ref 3), and there is an order of preference for their removal (Fig. 8). These metallic impurities are thus rendered removable if the oxide product formed can be adequately separated from the melt phase itself. A wet cover flux such as borax is useful with fire refining because it will agglomerate the impurity metal oxides formed and minimize the metal content of the dross. The need to refine specific metallic impurities is highly dependent on and variable with the specific alloy system being refined. An alloying element in one family of copper alloys may be an impurity in another, and vice versa. In red brass (Cu-5Zn-5Pb-5Sn; alloy C83600), the elements lead, tin, and zinc are used for alloying, while aluminum, iron, and silicon are impurities. In aluminum bronzes, on the other

hand, lead, tin, and zinc become contaminants, while aluminum and iron are alloying elements. Foundries typically do little melt refining, leaving this assignment to the secondary smelter supplier of their foundry ingot. However, there may be certain instances when additional refining capability is necessary within the foundry or mill. Table 6 gives the results of fire refining a melt of C83600 with aluminum, silicon, and iron contaminants under a variety of flux covers. Fire refining (oxidation) can be used to remove impurities from copper-base melts in approximately the following order: aluminum, manganese, silicon, phosphorus, iron, zinc, tin, and lead. Nickel, a deliberate alloying element in certain alloys but an impurity in others, is not readily removed by fire refining, but nickel oxide can be reduced at such operating temperatures. Mechanical mixing or agitation during fire refining improves the removal capability by increasing the reaction kinetics. Removal is limited, however, and in dilute amounts (