Cold Regions Pavement Engineering

Citation preview

Cold Regions Pavement Engineering Guy Doré, Ph.D., Ing. Professor of Civil Engineering Laval University Quebec City, Quebec

Hannele K. Zubeck, Ph.D., P.E. Professor of Civil Engineering University of Alaska Anchorage Anchorage, Alaska

New York Chicago San Francisco Lisbon London Madrid Mexico City Milan New Delhi San Juan Seoul Singapore Sydney Toronto

American Society of Civil Engineers 1801 Alexander Bell Drive, Reston, VA 20191-4400 www.pubs.asce.org Copyright © 2009 by the American Society of Civil Engineers. All rights reserved. Manufactured in the United States of America. Any statements expressed in these materials are those of the individual authors and do not necessarily represent the views of ASCE, which takes no responsibility for any statement made herein. No reference made in this publication to any specific method, product, process, or service constitutes or implies an endorsement, recommendation, or warranty thereof by ASCE. The materials are for general information only and do not represent a standard of ASCE, nor are they intended as a reference in purchase specifications, contracts, regulations, statutes, or any other legal document. ASCE makes no representation or warranty of any kind, whether express or implied, concerning the accuracy, completeness, suitability, or utility of any information, apparatus, product, or process discussed in this publication, and assumes no liability therefore. This information should not be used without first securing competent advice with respect to its suitability for any general or specific application. Anyone utilizing this information assumes all liability arising from such use, including but not limited to infringement of any patent or patents. ASCE and American Society of Civil Engineers—registered in U.S. Patent and Trademark Office. Photocopies and reprints: You can obtain instant permission to photocopy ASCE publications by using ASCE’s online permission service (http://pubs.asce.org/permissions/requests/). Requests for 100 copies or more should be submitted to the Reprints Department, Publications Division, ASCE, 1801 Alexander Bell Drive, Reston, VA 20191-4400; email: [email protected]. A reprint order form can be found at http://pubs.asce.org/support/reprints/.

Library of Congress Cataloging-in-Publication Data Doré, Guy. Cold regions pavement engineering / Guy Doré, Hannele K. Zubeck.— 1st ed. p. cm. Includes bibliographical references and index. ISBN 978-0-07-160088-0 (alk. paper) 1. Pavements—Cold regions—Design and construction. 2. Pavements— Cold regions—Maintenance and repair. I. Zubeck, Hannele K. II. Title. TE251.D67 2009 625.80911—dc22 2008037476 1 2 3 4 5 6 7 8 9 0 DOC/DOC 0 1 4 3 2 1 0 9 8 ISBN 978-0-07-160088-0 MHID 0-07-160088-4 Sponsoring Editor Larry S. Hager

Proofreader Carol Shields

Production Supervisor Richard C. Ruzycka

Indexer Broccoli Information Management

Editing Supervisor Stephen M. Smith

Art Director, Cover Jeff Weeks

Project Manager Somya Rustagi, International Typesetting and Composition

Composition International Typesetting and Composition

Copy Editor Surendra N. Shivam, International Typesetting and Composition Printed and bound by RR Donnelley. McGraw-Hill books are available at special quantity discounts to use as premiums and sales promotions, or for use in corporate training programs. To contact a special sales representative, please visit the Contact Us page at www.mhprofessional.com. This book is printed on acid-free paper. Information contained in this work has been obtained by The McGraw-Hill Companies, Inc. (“McGraw-Hill”) from sources believed to be reliable. However, neither McGraw-Hill nor its authors guarantee the accuracy or completeness of any information published herein, and neither McGraw-Hill nor its authors shall be responsible for any errors, omissions, or damages arising out of use of this information. This work is published with the understanding that McGraw-Hill and its authors are supplying information but are not attempting to render engineering or other professional services. If such services are required, the assistance of an appropriate professional should be sought.

To Natalie, Brad, and our lovely daughters

About the Authors Guy Doré, Ph.D., Ing., is a professor of civil engineering at Laval University in Quebec City, Quebec. He earned B.S. and M.S. degrees in geological engineering and a Ph.D. in civil engineering at Laval University. Before joining academia, Dr. Doré worked at the Quebec Ministry of Transportation. During his time at MTQ, he joined the Canadian Strategic Highway Research Program (C-SHRP) in Ottawa and the U.S. Strategic Highway Research Program (SHRP) in Washington, D.C., as a visiting researcher. Dr. Doré has authored and co-authored numerous journal and conference papers in his specialty area of pavement performance under the effects of frost and thaw. He serves on several national and international committees on cold regions engineering and recently received the American Society of Civil Engineers’ Can-Am Award for his distinguished service in building relationships between engineers in Canada and the United States. He resides in Ste.-Catherine-de-la-Jacques-Cartier near Quebec City with his wife Natalie and daughters Léonie, Rosalie, and Flavie. Hannele K. Zubeck, Ph.D., P.E., is a professor of civil engineering at the University of Alaska Anchorage. She earned B.S. and M.S. degrees in civil engineering at Tampere University of Technology in Tampere, Finland, and a Ph.D. in civil engineering at Oregon State University. Prior to joining academia, Dr. Zubeck worked in her native Finland as a geotechnical engineer for engineering consulting companies and as a research engineer in bituminous pavement materials for Neste Oil. She later joined the SHRP research team at Oregon State University. Dr. Zubeck has authored numerous journal and conference papers in her specialty area of the behavior of bituminous pavements in cold regions. She currently chairs UAA’s arctic engineering online graduate program and serves on several national and international committees on cold regions engineering. She lives in Kenai, Alaska, with her husband Brad and daughters Maija and Elli.

Contents Foreword . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

xiii xv

1

Cold Regions Pavements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-1 Road Networks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-2 Pavement Surface Types . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-2-1 Hot Mix Asphalt . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-2-2 Cold Mix . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-2-3 Surface Treatments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-2-4 Gravel Surface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-2-5 Stabilized Bases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-3 Role of Pavements and Pavement Layers . . . . . . . . . . . . . . . . . . 1-3-1 Surfacing Layer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-3-2 Base Course . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-3-3 Subbase . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-3-4 Subgrade Soil . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-3-5 Special Pavement Layers . . . . . . . . . . . . . . . . . . . . . . . . . 1-3-6 Embankment Geometry . . . . . . . . . . . . . . . . . . . . . . . . . . 1-4 Design Considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Review Questions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1 1 5 5 5 6 7 7 7 8 8 9 10 10 12 12 13 14

2

Pavement Environment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-1 Temperature Regime in Pavements . . . . . . . . . . . . . . . . . . . . . . . 2-1-1 Factors Inducing Heat in the Pavement System . . . . . . 2-1-2 Factors Contributing to Heat Extraction from the Pavement System . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-1-3 Factors Contributing Either to Heat Induction or Extraction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-1-4 Thermal Balance and Thermal Cycles . . . . . . . . . . . . . . 2-2 Moisture Regime in Pavements . . . . . . . . . . . . . . . . . . . . . . . . . . 2-2-1 Phases of Water . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-2-2 Factors Contributing to Water Intake in the Pavement System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-2-3 Factors Contributing to Moisture Extraction from the Pavement System . . . . . . . . . . . . . . . . . . . . . . . . 2-2-4 Moisture Balance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-3 Stress Regime in Pavements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-3-1 Earth Pressure at Rest . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-3-2 Static Stresses Induced by Traffic Loads . . . . . . . . . . . . 2-3-3 Stresses Related to Permanent Soil Movements . . . . . .

15 15 16 17 18 19 23 23 24 33 35 37 37 38 42

ix

x

Contents

3

4

5

2-3-4 Moving Traffic Loads . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-3-5 Thermal Stresses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-3-6 Stresses Related to Frost Heave . . . . . . . . . . . . . . . . . . . . 2-3-7 Negative or Positive Pore Pressure . . . . . . . . . . . . . . . . . 2-4 Interaction with Geology and Morphology . . . . . . . . . . . . . . . . Review Questions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

43 47 48 49 51 52 53

Cold Region Pavement Performance . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-1 Thermal Cracking of Asphalt Concrete . . . . . . . . . . . . . . . . . . . . 3-2 Fatigue Cracking . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-3 Crack Deterioration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-4 Rutting of Asphalt Concrete . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-4-1 Permanent Deformation . . . . . . . . . . . . . . . . . . . . . . . . . . 3-4-2 Rutting Due to Studded Tire Wear . . . . . . . . . . . . . . . . . 3-5 Aging of Asphalt Concrete . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-6 Pavement Disintegration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-7 Potholes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-8 Frost Heaving . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-8-1 Differential Frost Action . . . . . . . . . . . . . . . . . . . . . . . . . . 3-8-2 Frost Heave Cracking . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-8-3 Frost Heaving in Granular Base Material . . . . . . . . . . . 3-9 Bearing Capacity Loss During Spring Thaw . . . . . . . . . . . . . . . . 3-10 Frost Deconstruction of Undisturbed Sensitive Clays in Seasonal Frost Conditions . . . . . . . . . . . . . . . . . . . . . . . Review Questions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

57 58 65 69 71 71 76 80 82 85 88 88 93 96 99 106 108 109

Investigation and Testing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-1 Site Investigation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-1-1 General Site Investigation . . . . . . . . . . . . . . . . . . . . . . . . 4-2 Investigation of Existing Pavements . . . . . . . . . . . . . . . . . . . . . . 4-2-1 Evaluation of Pavement Structural Characteristics Using Falling Weight Deflectometer . . . . . . . . . . . . . . . 4-2-2 Evaluation of Pavement Functional Characteristics Using Longitudinal Profile Measurements . . . . . . . . . 4-2-3 Pavement Instrumentation . . . . . . . . . . . . . . . . . . . . . . . . 4-3 Soils and Material Testing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-3-1 Testing of Bituminous Pavement Materials . . . . . . . . . . 4-3-2 Soils and Unbound Materials . . . . . . . . . . . . . . . . . . . . . Review Questions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

115 115 116 150

157 163 164 164 184 201 204

Calculation of Engineering Parameters . . . . . . . . . . . . . . . . . . . . . . . . 5-1 Air Temperature and Air Freezing and Thawing Indices . . . . . 5-2 Surface Temperature and Surface Freezing and Thawing Indices . . . 5-2-1 The n-Factor Approach . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-2-2 Radiation Index Approach . . . . . . . . . . . . . . . . . . . . . . . .

209 210 217 217 218

150

Contents 5-3 5-4 5-5

Temperature in Asphalt Concrete . . . . . . . . . . . . . . . . . . . . . . . . . Thermal Properties of Soils and Pavement Materials . . . . . . . . Freezing and Thawing Indices within the Pavement Structure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-6 Frost and Thaw Depth . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-6-1 Transmitted Freezing Index Method . . . . . . . . . . . . . . . 5-7 Frost Heave . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-7-1 Konrad’s Method for Frost Heave Prediction . . . . . . . . 5-7-2 Saarelainen’s Method for Frost Heave Prediction . . . . 5-8 Thaw Settlement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-9 Stresses and Strains in Pavements . . . . . . . . . . . . . . . . . . . . . . . . Review Questions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

220 223

6

Design Considerations and Approaches . . . . . . . . . . . . . . . . . . . . . . . 6-1 Lifetime Engineering Considerations . . . . . . . . . . . . . . . . . . . . . 6-2 Long-Term Procurement Methods . . . . . . . . . . . . . . . . . . . . . . . . 6-3 Life-Cycle Cost Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-3-1 Factors of Life-Cycle Cost Analysis . . . . . . . . . . . . . . . . 6-3-2 Calculation of Life-Cycle Costs . . . . . . . . . . . . . . . . . . . . 6-4 Pavement Management Concepts . . . . . . . . . . . . . . . . . . . . . . . . 6-4-1 Network-Level PMS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-4-2 Project-Level PMS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Review Questions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

247 247 248 251 252 255 258 258 261 262 262

7

Mix Design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-1 Mix Design Approaches . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-2 Hot Mix Asphalt Concrete . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-2-1 Material Selection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-2-2 Trial Aggregate Gradations . . . . . . . . . . . . . . . . . . . . . . . 7-2-3 Volumetric Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-2-4 Performance Tests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-3 Cold Mixes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-3-1 Material Selection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-3-2 Selection of Optimum Asphalt Residue Content . . . . . 7-3-3 Cold Mix Recycling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-4 Stabilized Bases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-5 Asphalt Surface Treatment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-6 Gravel Surfaces . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Review Questions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

265 265 266 269 278 281 286 292 293 296 298 302 305 306 307 309

8

Pavement Design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-1 Current Practice in Pavement Design in Cold Climates . . . . . . 8-1-1 Pavement Design Approaches . . . . . . . . . . . . . . . . . . . . . 8-1-2 Synthesis of Design Methods Used by Highway Agencies in Cold Climates . . . . . . . . . . . . . . . . . . . . . . .

313 313 313

225 227 228 230 230 232 232 235 239 245

315

xi

xii

Contents 8-2

Mechanistic-Empirical Pavement Design Procedure for Cold Region Pavements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-3 Selection and Design of Special Protective Features . . . . . . . . . 8-3-1 Control of Frost Heave . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-3-2 Pavement Reinforcement . . . . . . . . . . . . . . . . . . . . . . . . . 8-4 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Review Questions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

322 333 333 343 344 345 345

9

Maintenance and Rehabilitation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-1 Routine Maintenance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-2 Major Maintenance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-2-1 Surface Treatments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-2-2 Overlays . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-2-3 Maintenance of Drainage Systems . . . . . . . . . . . . . . . . . 9-2-4 Repair of Local Failures . . . . . . . . . . . . . . . . . . . . . . . . . . 9-3 Maintenance of Gravel Roads . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-4 Rehabilitation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-5 Load Restrictions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Review Questions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

349 349 352 352 354 354 354 355 358 362 367 368

10

Pavements on Permafrost . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10-1 Causes of Instability and Problem Manifestation . . . . . . . . . . . 10-2 Climate Warming and Its Effect on Permafrost . . . . . . . . . . . . . 10-3 Management of Transportation Infrastructure Built over Thaw-Sensitive Permafrost . . . . . . . . . . . . . . . . . . . . . . . . 10-3-1 Identification of Thaw-Sensitive Areas . . . . . . . . . . . . . 10-3-2 Characterization of Thaw-Sensitive Soils . . . . . . . . . . . 10-3-3 Design Considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . 10-3-4 Technical and Economical Assessment of Applicable Solutions . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10-3-5 Implementation of the Strategy . . . . . . . . . . . . . . . . . . . . 10-4 Embankment and Pavement Design over Permafrost . . . . . . . . 10-4-1 General Considerations . . . . . . . . . . . . . . . . . . . . . . . . . . 10-4-2 Protection Techniques . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10-4-3 Methods Based on Preventing Heat Intake Underneath the Embankment . . . . . . . . . . . . . . . . . . . . 10-4-4 Methods Based on Heat Extraction from the Embankment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10-4-5 Methods Based on Embankment Reinforcement . . . . . 10-4-6 Other Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10-4-7 Applicability and Relative Cost of Protection Techniques . . . . . . . . . . . . . . . . . . . . . . . . . . . Review Questions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

369 369 374

Index

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

374 375 375 375 376 376 376 377 380 380 385 391 394 396 397 398 403

Foreword

A

new book on cold regions pavement engineering is great news and addresses a real and growing need. Cold regions present unique challenges to engineers, and Guy Doré and Hannele Zubeck have the experience and expertise to meet these challenges. Cold regions cover not only a substantial geographic area that includes North America, northern Europe, the Nordic countries, Russia, and northern Asia, but cold regions are increasing in significance as oil and gas production, mining, and transportation links grow in importance. Pavement engineering and the supporting technologies must incorporate the effects of temperature extremes, materials behavior, snow and ice, variable soil conditions, long distances, limited financial resources, high costs, variable bearing capacity, and other special conditions in making planning, design, and construction decisions, as well as implementing them. In writing the first book of its kind, Doré and Zubeck focus on cold regions, but they have certainly not limited themselves. They clearly recognize these special conditions and effects, and they make a major contribution to the state of knowledge and practice. Practitioners and researchers alike in the public and private sectors as well as academia are the beneficiaries. The costs of constructing pavements in cold regions and preserving them through proper maintenance are enormous. It is vital that the required technologies are understood and applied. The timeliness and value of this book in providing a foundation for such understanding will be apparent to the reader. While the authors address the special influences and sensitivities of environment, subgrade, materials, construction, and maintenance, they place equal emphasis on basic principles. This book addresses a pressing need, but it is also noteworthy in its comprehensive treatment of the underlying fundamentals, its extensive coverage of the subject, its presentation of problem descriptions, assessments, and remedial solutions reinforced by examples, and its references to the original sources of technology development. Current reference citations are provided, of course, so that readers can pursue up-to-date details of design, maintenance, and rehabilitation. A chapter on permafrost adds to the value of this book. So what makes Doré and Zubeck so authoritative? Both authors enjoy international reputations and experience in pavement engineering, complemented by an impressive track record of research and professional accomplishments. Guy Doré, at Laval University in Quebec, is a transportation, geotechnical, and materials engineer who has worked on various parts of the U.S. Strategic Highway Research Program and spent a sabbatical

xiii

xiv

Foreword leave in Alaska. He is certainly among North America’s key players in the pavement field. Hannele Zubeck, now with the University of Alaska Anchorage, builds upon extensive geotechnical background in her native Finland to add the complementary expertise and experience. Together, Doré and Zubeck have written a book that is a must-read for everyone in the field of pavement engineering and management. After many years of working in pavement engineering and management, as a teacher, researcher, and practitioner, and after two decades of teaching a course on northern engineering, I find this book by Doré and Zubeck a most welcome and timely contribution. It has been a pleasure to read the book, and it is an honor and a privilege to offer these few comments on what the book is about, why the subject is important, what is special about the book, and what is the authoritative background for the contribution. Ralph Haas, CM, FRSC, FCAE, FEIC, FCSCE, Ph.D., P.Eng. The Norman W. McLeod Engineering Professor and Distinguished Professor Emeritus University of Waterloo Waterloo, Ontario

Preface

O

ur goal was to author a book that will prepare engineers to make the right decisions in areas where freezing temperatures, unstable soils, snow and ice, sparse population, long road mileage, and often limited funds dictate design and maintenance actions on pavement structures. We aim for Cold Regions Pavement Engineering to be utilized by practicing civil engineers serving private consultants and by public agencies holding responsibility over roads in cold regions. We complemented the text with solved examples and problems that make the book also suitable as a textbook for graduate and upper-class civil engineering students. The book is divided into 10 chapters. Chapter 1 introduces readers to characteristics of cold regions pavements and pavement terminology. Road networks and their peculiarities in cold regions are explained in order to comprehend the special considerations required for pavement design and management. Pavement surface types used in cold regions are described as well as roles of pavements and pavement layers. Engineering challenges related to pavement design, materials, construction, and maintenance resulting from the aforementioned challenges in cold regions are introduced. Environment is the main cause of pavement engineering being different in cold regions than more temperate regions, and therefore a whole chapter is dedicated to it. Chapter 2 covers pavement environment, defined as a set of physical processes related to loading and climatic factors acting on a pavement in a given geological and geomorphologic context. Theories behind pavement temperature, stress regime, and moisture fluctuation including frost action are covered in detail. Chapter 3 leads the reader into the challenges in pavement performance in cold regions. Performance of asphalt pavements is explained regarding failure modes such as cracking, rutting, and disintegration of pavement surfaces. Pavement performance of the underlying structural layers and subbase including frost heaving, bearing capacity loss during spring thaw, and frost destructuration of clays is described. Problem assessment for each failure mode is explained with suggestions for appropriate mitigation techniques. Chapter 4 covers intensively investigation and testing of pavement materials, subgrade soils, and existing pavements. Focus is made on the test methods used mainly in cold regions. The test methods are described in detail including the analysis of the test results. Calculation of engineering parameters needed for pavement design is presented in Chap. 5. These parameters include air, ground, and pavement temperatures as well as estimation of frost depth, frost heave, and thaw settlement. Determination of stresses and strains at critical pavement interfaces needed for pavement structural design is also explained.

xv

xvi

Preface Design considerations and approaches are covered in Chap. 6. Pavements are longterm products and therefore lifetime engineering concepts and long-term procurement methods are introduced. Calculation of lifetime costs is covered in detail. Pavement management concepts including pavement condition assessment, prioritizing candidate sections, the impact of funding decisions, and the feedback process are discussed. Chapters 7 and 8 cover the design of pavements in cold regions. Chapter 7 focuses on mix design of bituminous pavement layers, surface treatments, stabilized bases, and gravel surfaces. Chapter 8 includes a synthesis of current pavement design practice, a description of the state-of-the-art methodology for mechanistic-empirical design of pavements, and guidelines for the selection and design of special features for the mitigation of cold region performance problems. Maintenance and rehabilitation of cold regions pavement have their own distinctiveness. Chapter 9 presents routine and major maintenance techniques including selection of appropriate rehabilitation techniques. Examples of winter maintenance quality standards are given as well as maintenance techniques. Rehabilitation techniques due to poor bearing capacity and widening of road embankments are proposed. Seasonal load restrictions unique to cold regions pavements are portrayed. Pavements in permafrost deserve their own chapter due to the fact that protection of permafrost from thawing is the most important part of the pavement design. Chapter 10 describes the causes and manifestation of thermal degradation, management considerations, design principles, and protection techniques for pavements on permafrost. Authoring this book would have not been possible without help from several supporters. We would especially like to acknowledge Dr. Ralf Haas, an internationally known expert in pavement engineering and management, who wrote the foreword for this book. Dr. Terhi Pellinen provided tremendous help in reviewing and providing materials for several sections, particularly the section on HMA mix design. Taina Rantanen coauthored the chapter on maintenance and rehabilitation of cold regions pavements, and Isabelle Beaulac synthesized most of the information used in Chap. 10. Several other individuals have provided their support and suggestions throughout the project. These persons include Ivar Horvli, Sven Knutsson, Kauko Kujala, Seppo Saarelainen, Steve Saboundjian, Safwat Said, and Ted Vinson. ASCE’s Technical Council on Cold Regions Engineering and its Committee on Transportation and Infrastructure viewed the book as important and provided their support and encouragement for making it available for cold regions engineers. Jean Parent perfected our drafts to create fabulous illustrations. Léon van Biljon translated our Frenglish and Finglish into English. Betsy Kulamer, the ASCE Press’ acquisitions editor, always had patience with us, accompanied by friendly words, when the schedule got stretched. Laval University and the University of Alaska Anchorage provided us with invaluable support throughout the authoring, especially in the form of sabbatical leave that made writing the book possible. The most faithful and encouraging supporters, however, have been our spouses Natalie and Brad, who endured with a smile on their faces the countless hours that we spent on researching and typing the manuscript. Our daughters, Léonie, Rosalie, Flavie, Maija, and Elli, deserve a big hug for their fond encouragement and understanding during the project. Guy Doré, Ph.D., Ing. Hannele K. Zubeck, Ph.D., P.E.

CHAPTER

1

Cold Regions Pavements

C

old regions pavements are pavement structures exposed to and affected by frost, ice, and snow for significant periods of time. They are located in seasonal or perennial frost areas, often connecting sparse populations spread out over hundreds of kilometers. Figure 1-1 illustrates the areas in the northern hemisphere where pavements are exposed to significant seasonal freezing and perennial frost conditions. The road alignment may cross regions with undesirable soils that are weak at all times, weak during the spring breakup, or experience heave due to frost, causing uneven driving surfaces. Ideal materials for the pavement structural layers may not be available, thus requiring that materials be brought from far away, local materials be modified, or performance expectations are lowered. Ground movements, thermal stresses and traffic loading, including the use of studded tires, cause pavements to rut and crack more severely in cold regions than in warm regions. Pavement funding in sparse population areas may not cover the capital and operating costs required for ideal pavement performance. Optimization is needed to use funds wisely so that roads are passable, safe, and meet desired performance levels. For these reasons, pavements in cold regions are considered from a different perspective than pavements in the warmer regions, where the traffic volume often dictates the design.

1-1

Road Networks To show the unique nature of the cold regions road networks, their densities are collected (in Table 1-1) from selected states and countries. The road density is defined as the ratio of the total length of roads to the total area of the state or country. The regions with the lowest road network density, namely, Alaska (United States), Yukon (Canada), and Mongolia are all cold regions. Alaska’s road network shown in Fig. 1-2 is 440 times less dense than the average road density in the United States. The road network density in Yukon is 10 times lower than the average density in Canada. The average road network density in Canada is about seven times lower than the average road network density in the United States and 20 times lower than that of France. The road network densities in Nordic countries (e.g., Iceland, Finland, Norway, and Sweden) are about five times lower than in the United Kingdom or France. Since populations in cold regions are lower, the total road length per person is not necessarily lower in cold regions than in warm regions. Low road density does not only mean few roads in a large area, but also long distances between settlements, maintenance stations, paving plants, and other resources.

1

FIGURE 1-1 Areas in the northern hemisphere where pavements experience significant seasonal and perennial frost conditions.

Total Length of Road Network, km Alaska∗

Land Area, km2

Road Density, km/100 km2

22,720

14,772,611

0.15

4,681

478,970

0.98

49,249

1,565,000

3.15

952,000

17,075,200

5.58

Canada

901,902

9,976,140

9.04

‡

12,955

103,000

12.58

§

78,161

338,145

23.11

‡

91,180

324,220

28.12

210,760

449,964

46.84

Yukon† Mongolia

‡

‡

Russia

‡

Iceland Finland

Norway

Sweden

‡

United States

‡

6,370,031

9,629,091

66.15

United Kingdom‡

371,603

244,820

151.79

France‡

892,900

547,030

163.23

1,152,207

377,835

304.95

Japan

‡

∗

http://www.dot.state.ak.us/stwdplng/highwaydata/pub/cprm/2002cprm.pdf [certified public road mileage in centerline km as of December 31, 2002: paved roads include asphalt surface treatments (AST)]. † http://www.gov.yk.ca/facts/#LAND (1990/2000 forecast for maintained road surfaces). ‡ The CIA’s World Fact Book, http://www.cia.gov/cia/publications/factbook/geos/rs.html. § http://www.tiehallinto.fi/pls/wwwedit/docs/17702.pdf (length of public roads in 2007).

TABLE 1-1

2

Road Network Densities Ranked in Order from Sparse to Dense for a Few Selected Areas

Cold Regions Pavements

FIGURE 1-2

Road network in Alaska.

Depending on government structure and politics, areas with low road densities may not receive adequate construction and operation funding. Since road construction and operation is more expensive per kilometer in cold regions than in warm regions, due to frost and other cold conditions, the funding limitations become magnified. Another unique feature of cold regions pavements, when compared to warm regions is the share of unpaved or undeveloped roads. Statistics of pavement types for selected states and countries are shown in Table 1-2. For example, 66 percent of Alaskan roads are not paved, whereas only 10 percent of the roads in the entire United States (including Alaska) are unpaved. The same applies for most of the cold regions (excluding Norway, Sweden, and Finland). The share of unpaved or undeveloped roads is significantly larger than paved roads. While the goal of public road agencies may be to increase the percentage of paved roads to minimize the maintenance costs for increased services to local populations and for dust control, it may not always be the goal of some local populations in quiet, remote areas. The beauty of many cold areas draws large numbers of tourists, which sometimes conflicts with the local lifestyle. Paving a road may change a quiet country road into a conduit for tourist buses, bringing hundreds of daily visitors and may change commercial traffic routes. Even if the funding for cold regions pavements were increased, there would probably still be unpaved roads. The proportion of unpaved roads and the unique effects of cold climate on pavement structures make it necessary for governments in cold regions to fund local research. For example, Superpave technology does not address all the conditions in cold climates adequately. Even if the research behind the Superpave was conducted by an outstanding research team, cold regions road authorities have to create their own design manuals and cannot rely purely on research conducted in warm regions.

3

4

Chapter One Paved Roads, km HMA or PCC

Surface Treatment

Unpaved Roads, km Unpaved (Gravel)

Unknown or Undeveloped

Total, km (% Unpaved)

United States (1997)a

5,733,028

637,003

Alaska (2002)b

7,791

8,744

Canada (1999)a

318,371

583,531

901,902 (65%)

Yukon (2000)c

260

2,525

4,681 (54%)

Finland (2007)f

50,836g

27,325

78,161 (35%)

Sweden (1999)a

162,707

48,053

210,760 (23%)

Norway (1999)a

67,838

23,342

91,180 (26%)

Iceland (2003)a

3,863g

9,092

12,955 (70%)

Russia (1998)a

336,000

416,000h

200,000

952,000 (65%)

Mongolia (2000)a

1,563

1,824

45,862

49,249 (97%)

United Kingdom (1998)a

371,603

—

371,603 (NA)

Japan (1997)a

863,003

289,204

1,152,207 (25%)

France (1999)a

892,900

—

892,900 (NA)

1,897e

6,370,031 (10%) 6,185d

a

22,720 (66%)

The CIA’s World Fact Book, http://www.cia.gov/cia/publications/factbook/geos/rs.html. http://www.dot.state.ak.us/stwdplng/highwaydata/pub/cprm/2002cprm.pdf [certified public road mileage in centerline km as of December 31, 2002: paved roads include asphalt surface treatments (AST)]. c http://www.gov.yk.ca/facts/#LAND (1990/2000 forecast for maintained road surfaces). d Classified as unknown. e Bituminous surface treatment, BST. f http://www.tiehallinto.fi/pls/wwwedit/docs/17702.pdf (length of public roads in 2007). g Includes cold mixes. h All-weather gravel surface. b

TABLE 1-2

Road Networks Divided between Paved and Unpaved Roads

Cold Regions Pavements

1-2

Pavement Surface Types The most common pavement surface types in cold regions include traditional hot mix asphalt, cold mixes, surface treatments, and gravel surfaces. All of these pavement types cover thousands of kilometers of road as shown in Table 1-2. Which pavement surface is the most common depends on the region. Alaska and western Canada pave some of their roads with asphalt surface treatment (AST). Iceland and Finland have used cold mix (oil gravel), but currently use reduced temperature mixes containing soft asphalt cements. Portland cement concrete and cement treated materials are also used, to a limited extent, in seasonal frost areas. For instance, several states and provinces in north central United States and central Canada use these materials for road construction. Pavements built using cement-based materials, commonly referred to as rigid or semi-rigid pavements, can perform very well if built at the suitable locations using proper design methods, materials, construction, and maintenance techniques. Among other applications, rigid or semi-rigid pavements are known to perform well when subjected to intense heavy traffic loading. There are, however, some limitations to their use in cold climates which restricts their applicability. Their relatively high initial cost and their sensitivity to differential soil movements are two main reasons making these pavements unsuitable in cold areas where traffic volumes are low and soils are sensitive to frost action. Being rarely used in these areas, it also becomes difficult to maintain good expertise for design, construction, and maintenance of rigid or semi-rigid pavements. Without downplaying the importance of rigid or semi-rigid pavements in some specific contexts, the book will focus mainly on hot mix asphalt, on asphalt treated, and on gravel pavements which constitute the vast majority of pavements used in cold environments. The following sections describe pavement types, while selection criteria, mix design, and structural design are described in Chaps. 6, 7, and 8.

1-2-1

Hot Mix Asphalt

Hot mix asphalt (HMA) is the most common pavement type used for high traffic volume roads. It contains typically 94 to 96 percent aggregate by weight and 4 to 6 percent asphalt cement. Antistripping agents, polymer modifiers, and fillers are some of the additives occasionally used to address anticipated performance problems. HMA is produced in a centralized hot mix plant, transported to the site by trucks, spread out by pavers and compacted by rollers. The mixing temperature at the hot mix plant is based on equivalent asphalt cement viscosity of 1.75 Pa·s that allows complete coating of aggregate while preventing unnecessary heating. The mixing temperature is about 135 to 160°C for straight run asphalt cements (PANK 2000) and higher for polymermodified asphalt cements. The layer thickness varies between 50 and 150 mm. The mix design is based either on the Marshall, Hveem, Superpave (Asphalt Institute 1997; 2001) or a similar procedure.

1-2-2

Cold Mix

Cold mix differs from HMA by the mixing temperature. Cold mixes are mixed at the ambient temperature or slightly heated. Lower mixing temperatures are made possible by modifying the asphalt cement by either emulsification, addition of lighter oil components, or by using road oils or extremely soft asphalt cements. The layer thickness is typically 50 mm. The mix design is based on similar techniques as used in HMA mix design.

5

6

Chapter One

1-2-3

Surface Treatments

While chip seals and coatings are used widely as surface treatments in warm climates, bituminous surface treatment (BST) is used extensively in the Yukon and Canada as a low cost highway surface course (MacLeod 1989). Alaskans have been using the same type of treatment since 1987, calling it asphalt surface treatment (AST). The AST and BST consist of a thin layer of asphalt binder, typically high float asphalt emulsion, covered with well-graded aggregate. In comparison, single-size aggregates are used to cover the emulsion in a chip seal application. Figures 1-3 and 1-4 show AST surfaces in Alaska and Yukon.

FIGURE 1-3

Close-up of asphalt surface treatment on Taylor Highway, Alaska.

FIGURE 1-4

16 Year-old asphalt surface treatment section in Yukon.

Cold Regions Pavements

FIGURE 1-5

Gravel surface of Denali Highway, Alaska.

The advantages of surface treatment versus gravel road are dust control, improved drainage, improved driving surface and reduced maintenance. Furthermore, the dust free surface and improved driving surface, come without the costly capital outlays required for hot-mix pavements. In permafrost areas, BSTs can be re-profiled more easily than conventional hot-mix asphalt. They can be rehabilitated more often and still remain cost effective (MacLeod 2000).

1-2-4

Gravel Surface

Gravel surface is one of the most common pavement types in cold regions. While it has the lowest capital cost to construct, the maintenance costs are often higher than that for paved roads. Gravel roads need periodical grading and dust mitigation. A typical gravel surface for the Denali Highway in Alaska is shown in Figure 1-5. Gravel surface may be treated with dust control palliative (such as calcium chloride), asphalt emulsions or proprietary blends.

1-2-5

Stabilized Bases

The base course beneath the wearing course may be bound or unbound. Asphalt product stabilization is the most common binding method in cold regions due to its flexibility. The asphalt products used are asphalt cement, asphalt emulsion or high float asphalt emulsion. The asphalt content is typically lower for the base course than for the surface layer.

1-3

Role of Pavements and Pavement Layers Pavements are large-scale linear structures stretching out across various geomorphologic, geologic, and climatic environments. Pavements, acting as an interface between the traffic and the underlying soil, have a twofold role: from top-down, they distribute the load, and from bottom-up, they attenuate various geotechnical effects. The first

7

8

Chapter One

FIGURE 1-6

Pavement system and related terminology.

aspect of a pavement’s role is to distribute loads from heavy traffic, which is well known and widely documented. Each layer of the pavement system must sustain a large number of moving load repetitions and effectively transmit an acceptable level of stress to the underlying layer. The second aspect of a pavement’s role is predominant in cold climates, it is not as well known or formally taken into consideration in pavement and material design procedures. A pavement’s second role is to attenuate environment-related stresses and displacements, such as differential frost heave and postconsolidation, as well as temperature, moisture content, and negative pore pressure (matric suction) variations, in order to maintain a good structural and functional performance. In order to fulfill these essential roles, pavement systems are composed of several layers, each having a specific role to play within the structure. As illustrated in Fig. 1-6, from top-down, pavements systems are typically composed of the surfacing layer, the base course, the granular subbase and the subgrade soil. Other features of the pavement system also contribute to its structural and functional stability. The most important features are the embankment geometry and the drainage system.

1-3-1

Surfacing Layer

As discussed above, when surfacing layers are used on cold region pavements, they typically consist of hot mix asphalt, cold mix asphalt, or bituminous surface treatments. The surfacing layer plays a structural and a functional role in the pavement system. The surfacing layer is the stiffest layer and thus the most effective layer for load distribution (except for the BSTs, which are generally considered to have a negligible structural role in pavements). Surfacing layers play another structural role. By sealing the surface of the pavement, they contribute to keeping the underlying granular layers relatively dry. In doing so, they help maximizing stiffness of those layers. In that sense, BSTs will also contribute to the structural capacity of a pavement system. Surfacing layers also play an important functional role in the pavement for traveling vehicles by providing adequate adherence and ultimately (since the whole pavement system is contributing), a good drive quality. They also improve the overall appearance of the road embankment and support pavement markings.

1-3-2

Base Course

The base course is a relatively stiff layer playing, essentially, a structural role in the pavement structure. The ability of the base layer to distribute load depends mainly on the material density and moisture content. Aggregate gradation, roughness, and shape also contribute to the stiffness of the granular base material. When an open graded

Cold Regions Pavements drainage layer is used, it is also responsible for the rapid evacuation of infiltrated water at the top of the pavement system. Three primary types of material are used for the construction of base courses: dense graded granular material, stabilized granular material, and permeable granular material (Haas 1997). Dense-graded (or well-graded) gravel with grain size ranging up to 20 mm is the most common material used for the construction of base courses. Mechanical performances of these materials can be maximized by specifying a high percentage of crushed particles. Stabilized base courses can also be used to increase the stiffness of the base course or to reduce moisture and/or frost susceptibility of the material. Different products are used to stabilize the granular base material. Asphalt cement and emulsion are the most commonly used materials in cold regions, but Portland cement, flyash, hydrated lime, calcium chloride, and lignosulfonate are also used (Haas 1997). Permeable granular material can be used in conditions where rapid drainage of water infiltrating through cracks and gravel shoulders is needed. Free drainage is usually obtained by washing or sieving off fine particles from the graded granular material. The material then usually becomes vulnerable to segregation during construction and unstable under load. Thus permeable base materials are often stabilized using a low content of bituminous material or Portland cement.

1-3-3

Subbase

When pavements are constructed over moisture and/or frost susceptible soils in cold and humid environments, the granular subbase also becomes a very important layer in the system. Subbase layers are generally constituted of good quality pit run gravel. The only important requirements for subbase materials is the maximum particle size (generally around 100 mm) to allow for proper compaction and the maximum content in fine particles (generally less than 10 percent passing 0.075 mm) to limit frost and moisture susceptibility. The subbase layer, which is generally stiffer than the underlying subgrade soil, plays a role in load distribution. However, the main role of the subbase layer is to attenuate environmental effects. More specifically, the layer acts as • A drainage layer to minimize and regulate the moisture content of the pavement structure and, more specifically, the base layer. • A separation layer to intercept fine particles from the subgrade soil migrating up under hydraulic pressure during spring thaw and thereby to prevent contamination of the overlying base course. While achieving this role, part of the subbase layer will be lost over the life span of the pavement. Depending on the conditions, a specially designed granular filter layer or a geotextile might be required to help the subbase layer achieve this specific function. • A frost protection layer. The subbase will achieve this role in three different ways. Firstly, through accumulated heat, the layer will resist frost penetration and reduce the duration of frost action in the frost susceptible subgrade soil, thus reducing frost heave and thaw weakening of the pavement system. The thickness of the subbase should be designed to limit or prevent frost penetration in frost susceptible subgrade layers. Secondly, if frost heave occurs, the subbase layer will dampen any differential movement that may result from frost action. Thirdly, the subbase layer will contribute to some extent to load distribution and drainage while frost susceptible subgrade soils are thawing.

9

10

Chapter One

1-3-4

Subgrade Soil

Despite the fact that subgrade soil is not part of the pavement structure, it is an important layer of the pavement system. In fact, it is by far the most complex layer of the system. In most cases, subgrade soils are local materials ranging from barely modified natural soils to engineered filled material. In cut sections, subgrade soils are generally constituted of natural mineral soils graded and compacted in place. In fill sections, subgrade soils are generally constituted of mineral soils excavated elsewhere on the construction project and then compacted in terrain depressions in order to obtain the desired grade. Subgrade soils are often compressible as well as moisture and frost susceptible. Moreover, they are often heterogeneous which makes them prone to differential behavior. The investigation of subgrade soils and the identification of problems will be discussed in Chap. 4. When poor performance of the subgrade soil is anticipated, the design engineer needs to decide if subgrade soils should be improved or if the pavement structure should be adapted to withstand the expected problems. Subgrade improvement techniques, generally considered when granular materials are scarce and adaptation techniques are too expensive, include homogenization, stabilization, reinforcement, and replacement of the subgrade soil.

1-3-5

Special Pavement Layers

Special layers used in pavements to perform specific roles include drainage layers, separation layers, reinforcement layers, and insulation layers. Drainage layers are generally used when the normal pavement layers are not expected to effectively drain water out of the system. This situation might be caused by an expected excessive water infiltration in the pavement system or by the use of poor draining material in pavement layers. Drainage layers can consist of open graded gravel wrapped in geotextile or of a geocomposite material. As illustrated in Fig. 1-7a, typical use of drainage layers includes • Horizontal drainage layers placed underneath the surfacing layer when excessive moisture contents are expected due to seepage through cracks, joints or granular shoulders, or due to frost action. • Horizontal drainage layers placed underneath the pavement structure when excessive moisture contents are expected due to abundant precipitation or thawing of ice-rich subgrade soils. • Vertical drainage layers placed near the pavement edge in order to intercept and evacuate water infiltrating through gravel shoulders. • Horizontal capillary barriers placed below expected frost depth to reduce water migration toward frost susceptible soils. It is generally accepted that active drainage of pavement systems is beneficial for the seasonal and long-term pavement performance. However, the effectiveness and more specifically the cost effectiveness of some of these layers are often questioned. Separation layers are required when a coarse layer is placed in contact with a finegrained layer in the pavement system. Examples of incompatible materials include coarse granular subbases placed over fine-grained subgrade soils or dense-graded base

Cold Regions Pavements

FIGURE 1-7

Special layers in pavement structures.

courses placed over coarse-grained subbases. Hydraulic pressures (spring thaw, artesian pressure, or other) or gravity can lead to fine particle migration into the coarsegrained layer causing its contamination and reducing its structural capacity. Separation layers generally consist of a geotextile blanket interposed between the two incompatible layers. A granular material can also be used. In all cases, the separation material used must meet specific hydraulic, filtration, and constructability criteria in order to be effective. Reinforcement layers may be required when pavements are subjected to excessive stresses due to frost heave, spring thaw, compressible soils, or other environmental factors. Reinforcement layers typically consist of woven geotextiles, geogrids, or steel mesh. As illustrated in Fig. 1.7b, typical applications include: • Horizontal reinforcement layer at the base of pavement structures constructed over compressible soils

11

12

Chapter One • Horizontal reinforcement layer placed at the bottom of the base course to improve confinement of the granular material and bearing capacity during spring thaw • Horizontal reinforcement layer placed within or at the bottom of the bound layer to resist tensile stresses caused by temperature or frost heave The cost effectiveness of these applications still needs to be demonstrated. Insulation layers are often used in conditions where an excessive differential movement caused by frost action or excessive weakening during spring thaw is expected. As illustrated in Fig. 1.7c, the insulation layer consists of a horizontal layer of low thermal conductivity material. Extruded polystyrene is the most commonly used material for pavement insulation. Several other materials such as expanded polystyrene, sulfur foam, polyurethane foam, expanded clay, polystyrene-concrete mix’s, tire chips, sawdust, tree barks, and peat are used for pavement insulation material. The depth of the insulation layer should be carefully calculated considering the facts that the benefit of the insulation is maximized if the layer is near the surface (better protection of the pavement system) while the risk of poor mechanical performance and the risk of differential icing at the pavement surface decrease with depth. In addition, the cost of placing an insulation layer in an existing pavement increases considerably with depth. Pavement insulation is discussed further in Chap. 8.

1-3-6

Embankment Geometry

The road embankment also plays an important role in pavement performance. Granular materials constituting base and subbase layers generally have a stress dependant behavior under load resulting in a stiffness increase with confinement stress. Considering this typical behavior, embankments with narrow shoulders and steep slopes do not provide good confinement conditions to pavement structural layers resulting in larger deformations under loading and ultimately greater permanent deformation. Increasing embankment width and reducing slopes will result in a more stable embankment. In conclusion, despite the fact that pavements appear to be simple structures, they are complex multilayer systems where every layer plays an important role. Moreover, as discussed in the following chapter, pavements are in complex interaction with environmental factors among which traffic and climate are the most important.

1-4

Design Considerations Pavements are long term products, therefore lifetime engineering approaches starting with the investment planning and decision making should be applied. Lifetime engineering considerations in the integrated design, management, and maintenance planning, and in recovery and reuse of construction materials are all vital for wise use of resources. For a cold region pavement to technically perform according to the desired level of service, four criteria need to be considered: design, materials, construction, and maintenance. These components are as links in a chain. When the weakest link fails, the entire chain fails. Therefore, proper lifetime design considering available capital, desired performance, operation and maintenance funding, access to the site, local traffic, subgrade soils, available construction materials, construction equipment, skilled labor, and

Cold Regions Pavements weather conditions need to be applied. Due to changing conditions, cookie-cutter designs seldom work in cold regions and should not be used. Technologies transposed from warmer or more developed regions should not be implemented without validation through pilot projects. Availability of funding and funding sources affect pavement design. Ideally, unlimited funding would ensure long-lasting pavement structures with low maintenance needs. This situation, however, seldom occurs. Typically available funds are stretched over a multitude of projects. Pavement management methods are used to prioritize and optimize the road sections needing rehabilitation or reconstruction in relation to the costs of the individual projects. In cases where capital funding is more available than operation and maintenance funding, low maintenance pavements should be designed. On the other hand, if funding for capital projects is low, and there is more financial support for operation and maintenance, then roads with lower capital cost to perform under constant maintenance operations should be designed. After proper materials have been specified in the design phase, quality control and assurance should be carried out before and during construction to ensure that the materials do not become the weakest link. The materials need to resist not only the construction and traffic loading without breaking apart, but also environmental stresses such as freeze-thaw cycling, thermal stresses, extremely cold and sometimes hot temperatures. Construction in cold regions differs from warm regions due to the daylight, weather, short construction season, and a multitude of challenges due to long distances. Sometimes, no land or water access exists to the construction site. Excavation and compaction of frozen soils may be an issue. Paving, when the weather is cold or wet and when hauling distances are long may be a real challenge. Clearly written contract documents and end-result specifications with physically measurable values are recommended. As new materials or systems are introduced to market, special construction equipment or methods may be needed to meet the specifications. Cultural and ecological considerations are a vital part of the lifetime engineering methodology, and their importance increases in cold regions with indigenous populations and fragile ecosystems. For example, the selection of the wearing course is affected by sociological and ecological considerations. A paved surface would minimize road dust and its ecological impact, but also have an aesthetic impact changing the character of an area due to increased mobility and subsequent traffic.

Review Questions 1-1. Identify and briefly discuss three factors contributing to the complexity of designing and building roads in cold regions.

1-2. What is (a) the road network density and (b) percentage of paved roads in your area? 1-3. What are the main factors to consider for the selection of a pavement surface type? 1-4. What are the characteristics of a granular subbase and what are the main functions of the layer in a cold region context?

1-5. layer.

Identify three ways to reinforce pavement structures using a synthetic reinforcement

13

14

Chapter One

References Asphalt Institute (1997). MS-2 Mix Design Methods for Asphalt Concrete and Other Hot Mix Types, 6th ed. Asphalt Institute, Lexington, Ky. Asphalt Institute (2001). SP-2 Superpave Mix Design, 3d ed. Asphalt Institute, Lexington, Ky. Haas, R. (1997). Pavement Design and Management Guide. Transportation Association of Canada, Ottawa, Canada. MacLeod, D. R. (1989). A BST Management System for Yukon Highways. Public Works Canada, DIAND Technical Services, Government of Yukon Community & Transportation Services. MacLeod, D. R. (2000). “BST Management Systems in the Yukon Territory,” Technology for Alaskan Transportation, vol. 25, no. 4. PANK (2000). Finnish Asphalt Specifications. Finnish Pavement Technology Council, PANK, Helsinki, Finland.

CHAPTER

2

Pavement Environment

T

he terms environment and environmental factors have been used widely and rather loosely to describe pavement conditions. For the purpose of pavement engineering, pavement environment can be defined as a set of physical processes related to climatic factors acting on a pavement in a given geological and geomorphologic context. Thus, it involves the interaction of climatic factors, soils, and land morphology. Pavements have their own climate. There are several similarities and interactions between pavement climate and atmospheric climate. They both can be characterized by their temperature, their level of humidity, and by acting pressures. They are also affected by daily and seasonal variations of these parameters, by their spatial distribution, and their interactions with the system’s environment (Oliver 1973). As opposed to atmospheric climate characterized by physical processes active in a gaseous environment, pavement climate occurs in a porous mineral system governed by its own physical laws (Oliver 1973). Pavement climatic factors are in constant interaction with material properties and external loading on the system. Pavement climate can be characterized by its temperature, moisture, and pressure regimes. This chapter includes a summary description of important environmental effects on pavement systems. Basic equations describing relevant phenomena are presented in simple form to help the reader understand the role played by different contributing factors. Readers interested in the development and the use of these equations may need to refer to the references cited.

2-1 Temperature Regime in Pavements The temperature regime of a pavement-soil system is controlled by the boundary conditions of the system, but is also affected by energy available within the system. The temperature at the bottom of the system is practically constant throughout the year and roughly equal to the mean annual surface temperature; at that point, the temperature conditions correspond to the steady-state balance between the geothermal heat flux and the annual average heat loss in the atmosphere. Negligible temperature variations are usually observed at depths of about 10 m in a pavement-soil system. At the top of the system, temperature varies considerably between summer high and winter low temperatures. Temperature at the surface of the pavement-soil system is the result of a complex balance at the pavement-air interface. Figure 2-1 illustrates the main factors affecting pavement thermal regime.

15

16

Chapter Two

± 6

+

+ 7

4a

1

3

±

+

2

Factors contribution to heat intake Factors contributing to heat extraction Factors contributing either to heat induction or extraction

1. 2. 3. 4. 5. 6. 7.

4b

5

Solar radiation Geothermal heat Emitted radiations Convection and turbulence Latent heat of fusion Evaporation/condensation Heat exchange with precipitations

FIGURE 2-1 Summary of the factors affecting temperature regime in pavements (highlighted factors are considered most frequently in pavement engineering).

2-1-1

Factors Inducing Heat in the Pavement System

Solar Radiation Solar radiation is electromagnetic energy emitted by the sun and needs no support for its propagation. It is this parameter that has the largest effect on pavement surface temperature (Dysli et al. 1997). The amount of solar radiation reaching the earth’s surface is a function of the seasonal variations of the length of a day and the angle of incidence with the surface of the earth. The latter factor affects the distance traveled in the atmosphere and the resulting diffusion, reflection, and absorption by atmospheric particles. For a given amount of solar radiation reaching the pavement surface, part of it is absorbed and the rest is reflected back into the atmosphere. The ratio between the radiation reflected by a surface and the total radiation reaching the surface is termed the “albedo” of the surface. The albedo of a normal bituminous pavement surface is approximately 15 percent (10 to 18 percent) which means that 85 percent of solar radiation is absorbed by the surface. Furthermore, the albedo of a packed snow-covered pavement surface is around 55 percent (40 to 60 percent), thus reducing the absorbed radiation to 45 percent. The albedo of a pavement surface covered by fresh snow can reach up to 80 percent.

Geothermal Heat A large amount of heat, accumulated in the earth’s core and crust, flows toward Earth’s surface and is dissipated into the atmosphere. It is called geothermal heat. As a general

Pavement Environment

FIGURE 2-2 Thermal regime fields (trumpet and whiplash curves) in pavement systems for (a) seasonal frost and (b) permafrost conditions.

rule, whenever there is a thermal gradient, heat flow will occur. This mechanism is described by Fourier’s equation: qG = − kTG

(2-1)

According to Eq. (2-1), the heat flux in homogeneous soil, qG, is proportional to the thermal gradient, TG. The proportionality constant, k, is the thermal conductivity of the soil. The slope between the mean annual surface temperature (MAST) and the temperature at the center of the earth is the average geothermal gradient that governs the heat flux at the surface of the earth. For practical pavement engineering considerations, the area of interest is the depth where no significant temperature variation occurs. The average geothermal heat flux at the surface of the pavement system is proportional to the thermal gradient TG shown in Fig. 2-2. The average thermal gradient near the pavement surface increases considerably during winter, thus augmenting the geothermal heat flux at the surface of the cooling pavement system. However, during summer the geothermal gradient is nullified by a steeper opposed thermal gradient resulting from the warming of the pavement surface and as a result, the heat flux is inverted. Heat is thus accumulated in the pavement system until the following cooling cycle. This phenomenon is analogous to water flowing from a river affected by tidal forces into the sea. When the tide is low (low energy corresponding to a cold surface), the current is swift and the flow is high. When the tide is high (warm surface), the gradient and the current are inverted in the mouth of the river and water accumulates until the next low tide.

2-1-2

Factors Contributing to Heat Extraction from the Pavement System

Emitted Radiations All surfaces emit energy, qer, in the form of electromagnetic radiation. A perfect radiator emits radiation, the intensity of which is proportional to the fourth power of its temperature (T) as predicted by the Stefan-Boltzmann equation: qer = σ T 4

(2-2)

17

18

Chapter Two where s is the Stefan-Boltzman constant (5.67 × 10−8W/m2·K4). Pavement surfaces emit long-wave and infrared radiations mostly during the night when the pavement surface is often warmer than the air. Adapted to heat exchange between pavement surface and atmosphere, Eq. (2-2) can be written as

(

qer = σε s Ts 4 − Ta4

)

(2-3)

where es is the surface emissivity, which, for asphalt pavements is in the range of 0.90 to 0.95, and Ts and Ta are, respectively, the temperature of the surface and the atmosphere.

Convection and Turbulence Another important mechanism contributing to heat extraction from pavement surfaces is convection. Convection requires the support of moving fluid such as air. In order for heat to be extracted from the pavement surface, it must first be transferred to a thin film of air by conduction and radiation. The difference of temperature between the heated thin film of air and the rest of the air mass induces a difference of pressure, which will in turn induce a motion in the air mass. Heat will then move to another location and be dissipated. In the presence of wind-induced turbulence, the motion of the fluid and its dispersive capacity will be increased. As indicated by Eq. (2-4), heat extraction by convection, qc, is thus a function of the temperature difference between the surface, the fluid and wind speed. qc = hc (Ts − Ta )

(2-4)

where Ts and Ta are surface and air temperature and hc is the convection coefficient (W/m2·K). The coefficient hc is in turn a function of the surface drag coefficient, air density, specific heat of air, air Prandtl number and wind speed. Considering the difficulty of measuring or determining these parameters, McAdams (1954), cited by Zarling and Braley 1988, has proposed the following dimensional relationship to determine the convective heat transfer coefficient for a smooth surface as a function of wind speed: hc = 5.678 + 1.056U

(2-5)

where U is wind speed in km/h and hc is in W/m2·K. From Eq. (2-5), it can be inferred that the convective heat flux will double if wind speed goes from 0 to 5 km/h.

2-1-3

Factors Contributing Either to Heat Induction or Extraction

Latent Heat of Fusion One of the basic principles of chemistry is that a system always tends to oppose changes imposed to the system. Phase change of water is not an exception as heat is released during freezing and absorbed during thawing. The quantity of heat released or absorbed during the phase change is termed the latent heat of fusion. Latent heat of fusion, L, is constant at 334 kJ/kg or 334 MJ/m3 for freezing or thawing water. For soils or pavement materials, the latent heat of fusion, Ls, can be obtained from Ls =

ω ρd L 100 ρW

(2-6)

Pavement Environment where w is the gravimetric water content of soil or pavement material, rd and rw are the densities of dry soil and water, respectively.

Evaporation/Condensation Following the same principle for liquid-solid, vapor-liquid phase change will also absorb or release heat. Water accumulated in asphalt and gravel surface pores tends to evaporate in warm and dry conditions, absorbing heat and consequently impeding surface warming. Humidity present in air is likely to condense on cool pavement surfaces releasing heat and impeding surface cooling. Heat released or absorbed by the evaporation/condensation process, qe, can be approximated as qe = nw × h fg

(2-7)

where nw is the evaporative mass flux (kg/s·m2) and hfg is the heat of vaporization of water (2257 kJ/kg). The amount of water involved in the process is, however, generally small. For instance, Dysli (1991) has reported that phase change due to pavement deicing has modified surface temperature by less than 1°C for a period of about 20 min. Heat exchange resulting from the evaporation/condensation process is thus considered to be negligible for surfaced pavements, but can perhaps be significant for unsurfaced roads.

Heat Exchange with Precipitation Precipitation in the form of rain or snow is likely to affect pavement surface temperature, if different from the temperature of the precipitation. Heat will be transferred between the surface and the precipitation by conduction and the effectiveness of the process depends on the difference in temperature and quantity of precipitation. Heat exchange resulting from the contact between pavement surface and precipitation can be significant at the scale of an event, but can be considered negligible over a long period.

2-1-4 Thermal Balance and Thermal Cycles As illustrated in Fig. 2-1, absorbed solar radiation, Qsr, and geothermal heat, Qg, contribute to the heat intake at the air-pavement interface. The main contributors to heat extraction are emitted radiation, Qer, and air convection and turbulence, Qac. Latent heat of fusion, L, evaporation/condensation, Qe, and precipitations, Qp, can either contribute to induction or extraction of heat from the pavement surface depending on the prevailing thermal conditions. The thermal balance of the surface of the pavement can thus be obtained by Qsr + Q g − Qer − Qac ± Qe ± L ± Qp = 0

(2-8)

Qg, Qe, L, and Qp can have a significant contribution to short-term thermal variations of the pavement surface. It is however generally accepted (Dysli et al. 1997; Zarling and Braley 1988; Pavlov 1976) that for the establishment of long-term (seasonal) thermal balance these factors are not significant and can be neglected. For practical considerations, thermal balance equation can thus be written as Qsr − Qer − Qac = 0

(2-9)

19

20

Chapter Two

FIGURE 2-3 Schematic illustration of a diurnal thermal balance cycle at the pavement surface and resulting surface temperature.

All these factors undergo important temporal variations causing the balance to be different than 0. When the balance is positive, pavement surface tends to accumulate heat and its temperature increases. When the balance is negative, pavement surface tends to lose heat and its temperature decreases. As a result, pavement surface temperature follows two typical behavior cycles: diurnal and seasonal. Diurnal cycles: During the day, solar radiation is generally high and emitted radiation is relatively low. Surface temperature tends to increase until the end of daylight. Maximum surface temperature will be observed on clear sunny days without wind. At night, solar radiation becomes negligible and emitted radiation increases. The thermal balance then becomes negative and pavement surface cools down. A typical diurnal cycle is conceptually illustrated in Fig. 2-3. Seasonal cycles: During summer months, the high position of the sun in the sky maximizes the quantity of radiation reaching the surface of the earth. The overall diurnal thermal balance tends to be positive and heat builds up at the pavement surface. In contrast to summer conditions, the sun is low on the horizon during winter months. Consequently, absorption is low and diffusion of solar radiation in the atmosphere is maximized. As a result, the overall diurnal thermal balance is negative and the pavement surface temperature cools down. Figure 2-4 illustrates a typical seasonal cycle. Temperature regime in pavement systems will evolve between a stable bottom temperature and a continually changing surface temperature. The envelope of temperature conditions at any given depth in the pavement system is referred to as the “trumpet curve.” Figure 2-2 shows typical trumpet curves for seasonal frost conditions (Fig. 2-2a)

Pavement Environment

FIGURE 2-4 Schematic illustration of a seasonal thermal balance cycle at the pavement surface and resulting surface temperature.

and permafrost conditions (Fig. 2-2b). Maximum and minimum surface temperatures as well as frost and thaw depth can readily be obtained from these representations. Temperature variations at the surface of the pavement will cause the temperature regime curve to swing within the limits defined by the trumpet curve. These curves represent “snapshots” of temperature conditions at one point in time within the system and are often referred to as the “whiplash curves.” Typical spring and fall whiplash curves are illustrated in Fig. 2-2a. The shape of the curve depends on the boundary conditions at the surface and at the base of the pavement system. The shapes also depend on several factors specific to the physical properties of soils and pavement materials. These factors include: • Thermal conductivity, k: As defined in Eq. (2-1), thermal conductivity is the proportionality constant between heat flux in homogeneous soil, qG, and the thermal gradient, TG. Thermal conductivity represents the capacity of a material to transport heat by conduction. The thermal conductivity of soils and pavement materials increases as dry density increases and as degree of saturation increases. • Heat capacity: It represents the ability of soils or materials to accumulate heat. It is defined as the amount of heat required to raise the temperature of a unit quantity of soil or material by 1°C. Mineral particles and interstitial water contribute to the heat capacity of soils. When the unit quantity is a volume (1 m3), the property is termed the volumetric capacity (Cv) and when it is a mass (1 kg), it is termed the massic heat capacity (Cm). • Moisture affinity: Moisture in soils and pavement materials has an important influence on the thermal regime. The quantity of water present in soils directly affects its thermal conductivity and its heat capacity. It will also generate or absorb an important quantity of heat (latent heat of fusion, L) during phase change. Moisture regimes in pavement systems will be discussed in Sec. 2-2. Figure 2-5 is a schematic illustration of the effect of these factors on temperature regime within the pavement system. Figure 2-5a and b represent, respectively, a cooling

21

22

Chapter Two

FIGURE 2-5 Factors affecting the thermal regime in (a) cooling and (b) warming pavement system composed of asphalt concrete surfacing layer (1) underlain by granular soil (2).

and a warming bilayer system composed of asphalt concrete over granular soils. Temperatures at the top (Tt) and at the base (Tb) are imposed boundary conditions. Following the thermodynamic principle of energy conservation, energy entering a system plus energy generated by the system must be equal to energy leaving the system plus energy stored within the system. If no energy is generated or stored in the system, the heat flux in both layers will be equal and the temperature regime will be governed by the Fourier heat conduction law [Eq. (2-1)]. The permanent temperature regime illustrated by the short-dashed line will rapidly be reached and the thermal gradient will be inversely proportional to the thermal conductivity k of the layer. Soils and pavement materials have the capacity to generate and store heat. When the system is cooling (Fig. 2-5a), energy stored (C) will be progressively released and will impede cooling of the system. This is illustrated by a warmer temperature regime illustrated by the solid line in Fig. 2-5a. With heat being generated in Layer 2, the amount of heat leaving the layer will be greater than the amount of heat entering the layer, as illustrated by steeper gradients near the top of the layer than at the bottom of the layer. The same principle applies to a warming pavement system. Figure 2-5b illustrates that heat is stored in the system during the warming process. The result is a cooler temperature regime. Due to the heat loss within the system, the amount of heat leaving the system is less than the amount of heat entering the system. This is illustrated by a steeper gradient at the bottom of the system than at the top of the system. For soils and pavement materials with moisture available in pores or in ice lenses, energy is also available or storable in the form of latent heat of fusion. The process is illustrated in Fig. 2-5a and b through the phase change of a thin zone of high water content within Layer 2. As indicated by the long-dashed line, latent heat of fusion will increase the heat available in a cooling system, thus increasing internal temperature within and above the high moisture content zone. Here again, since the amount of heat leaving the system is greater than the amount of heat entering the system, the thermal gradients at the top of the system are steeper than at the bottom of the system. In a

Pavement Environment warming system, the latent heat of fusion is acting in the opposing direction absorbing some of the heat flowing through the system. Temperatures are thus reduced and thermal gradients at the bottom of the system are gentler. The latent heat exchange process is, however, limited to a short period of time during which phase change is occurring. Nevertheless, when moisture contents are high, latent heat of fusion is a major factor limiting frost/thaw penetration in pavement systems.

2-2

Moisture Regime in Pavements Moisture conditions in pavement systems ensue from the amount of water in soils and pavement materials, its movements, the form under which it occurs, and the phase under which it operates. Like moisture in the air, soil moisture can occur in gaseous, liquid, and solid forms, each significantly modifying the soils properties (Oliver 1973).

2-2-1

Phases of Water

The gaseous phase, water vapor, is always present in unsaturated soils. The humidity of the air in soil pores is always close to 100 percent (Oliver 1973). Vapor can play an important role in water transportation in unsaturated soils. The liquid phase is the most important phase of water for pavement engineering. Liquid water can be found under different forms in soils and pavement materials. These forms are free or gravitational water, capillary water, hygroscopic water, and chemically bound water. Free water is not subjected to any significant force resulting from the interaction between water and soil particles and can move without restraint in soil pores under acting forces such as gravitation or suction. Capillary water is the portion of water held by surface tension forces in continuous films around soil particles in capillary interstices. Surface tension exists at air-water or water-ice interfaces. It results from the fact that water molecules at the surface of a water body would be unstable if they were not submitted to a very high tensile pull along the surface of the liquid. The thin layer of water submitted to high tensile forces is termed the “contractile skin” (Fredlund and Rahardjo 1993) and controls the behavior of capillary water. Capillary water can move in soil pores under acting forces including capillary force. Hygroscopic water is a thin layer of water bound by chemical attraction to polarized clay particles. In the bound water layer, water molecule tends toward a pseudo-crystalline structure (Dysli 1991), which considerably reduces the mobility of water particles in a plane perpendicular to the surface of the particle. Hygroscopic water cannot move in soils under the forces typically acting in a pavement-soil environment. The only form of water that cannot be removed from soil at high temperature is water chemically bound to soil minerals such as hydrated oxides. In addition to binding forces acting on interstitial water, another important factor controlling water state and mobility in soils and pavement materials is the continuity of the mobile water film in an unsaturated soil matrix. As described by Dysli (1991), unsaturated soil can be classified in three different states: Lenticular water regime where water films are discontinuous, making the hydraulic conductivity null, but allowing for effective vapor transportation through communicating air voids.

23

24

Chapter Two

FIGURE 2-6

Water regime in soils.

Funicular water regime where water and air films are distinct and continuous, making possible water and vapor transportation through the soil matrix. Occluded air regime where vapor transport is not possible, but hydraulic conductivity is relatively effective as it tends toward saturated soil conditions. Solid phase of water begins to develop in soils and pavement materials when the temperature is cold enough for a period of time to allow phase change of interstitial water. Two types of ice are found in soils: interstitial ice and segregation ice. In the first case, water freezes in soil pores. In the case of saturated soils, phase change of the interstitial water can cause a volume change that can reach 1.09 times the soil porosity. For segregation ice, the freezing process involves transportation of water from the warm side of the freezing front toward the segregation front. Thus, it involves an increase in water content and a volume increase of the freezing soil. As for temperature, water regime in soils is governed by boundary conditions with many similarities. As shown in Fig. 2-6, at the bottom of the system, the conditions are relatively stable and correspond to the groundwater table where soils are saturated and water is at atmospheric pressure. At the top of the system, moisture conditions vary widely between saturated conditions and dry conditions as a result of surface characteristics and climatic events. In a typical pavement system, surface conditions may vary widely depending on the longitudinal and transverse position on the pavement. As indicated in Fig. 2-7, numerous water sources may affect water conditions and, consequently, moisture regime in pavement systems. On the other hand, other factors contribute to moisture removal from the pavement system. These factors are listed in Fig. 2-7 and described below.

2-2-2

Factors Contributing to Water Intake in the Pavement System

Capillary Rise Capillary forces will draw up water from the groundwater table zone into a zone called the capillary fringe. Capillary rise varies considerably depending on soil characteristics. The concept of capillary rise has often been demonstrated using an ideal

Pavement Environment

FIGURE 2-7

Factors affecting moisture regime in pavements.

capillary tube in which the rise hc is inversely proportional to the diameter of the tube according to hc =

4T cos α dγ w

(2-10)

where T is the surface tension at the interface of water and air, a is the angle of contact between the water meniscus and the capillary tube, d is the tube diameter, and gw is unit weight of water. Based on the principle that capillary rise is a function of the size of the openings, Hazen (1930) has proposed a formula to estimate capillary rise (hc, mm) in soils based on the effective particle diameter D10 (mm): hc =

C eD10

(2-11)

where C is a constant ranging between 10 and 50 mm2 and e is the void ratio.

Lateral Moisture Transfer Lateral moisture transfer is likely to occur when the groundwater table in the surrounding areas is higher than underneath the pavement. As illustrated in Fig. 2-8a, the resulting hydraulic gradient will induce a water flow underneath the pavement. In addition to the moisture regime, the water mass is likely to modify the temperature regime and to induce hydraulic pressures in the pavement system. A special case of lateral moisture transfer is illustrated in Fig. 2-8b and involves an artesian aquifer. In that specific case, hydraulic pressures and upward water flow can be generated underneath the pavement.

25

26

Chapter Two

FIGURE 2-8 Lateral moisture transfer in pavement systems from (a) adjacent high water table and (b) from artesian aquifer.

These phenomena are likely to reduce the bearing capacity and the stability of the highway embankment. Frost depth may be reduced by the latent heat of fusion accumulated in interstitial water. Frost heave is, however, likely to increase considerably if soils are frost susceptible, due to the more effective water movement toward the freezing front. It is possible to compute the flow rate, ql, into a pavement system from lateral moisture transfer on the basis of Darcy’s law: ql = k h i

(2-12)

where kh is the horizontal hydraulic conductivity of the soil and i is the hydraulic gradient. A comprehensive description of the method for computing lateral moisture flow can be found in Garber and Hoel (1997) and Moulton (1980).

Infiltration of Water from Precipitations Infiltration is one of the main sources of water in pavement systems above the capillary fringe. Water from liquid or solid precipitations (assuming it melts on the pavement surface due to the temperature of the pavement surface or action of deicing chemicals) is likely to seep into the pavement structure through cracks in the bound surfacing layer and through unprotected gravel surfaces such as shoulders and side slopes (case 3a on Fig. 2-7). The following empirical relationship is proposed by the Federal Highway Administration (FHWA) (Johnson and Chang 1984) to estimate the infiltration rate into cracked surfaced pavements: N Wc  qic = I c  c + + Kp  W WCs 

(2-13)

where qic is the design infiltration rate (m3/day·m2 of drainage layer), Ic is the crack infiltration rate = 0.22 m3/day·m linear of crack (recommended empirical value), Nc is the

Pavement Environment number of contributing longitudinal cracks or joints, Wc is the length of contributing transverse cracks (m), W is the width of granular base subjected to infiltration (m), Cs is the spacing of transverse cracks or joints (m), and Kp is the rate of infiltration (m3/day·m2) through uncracked pavements (can be assumed to be 0 for dense asphalt concrete). A significant amount of water will also seep into the pavement structure through granular shoulders and unprotected embankment slopes. From equations proposed in the literature (Johnson and Chang 1984) to estimate the runoff from rainfall, Eq. (2-14) can be derived to estimate the infiltration in pavement structures: qi = (1 − C)iA

(2-14)

where qi is the rate of infiltration (m3/s·m2), i is the average rainfall intensity (m/s), A is the infiltration surface (m2), and C is the runoff coefficient that varies from 0.4 to 0.6 for gravel surfaces or shoulders and from 0.5 to 0.7 for embankment slopes (Johnson and Chang 1984). The lower values should be used for flat slopes and permeable soils, while the higher values should be used for steep slopes and impermeable soils. Runoff water will reach the bottom of the ditch where more infiltration is likely to occur before the remaining water is evacuated from the pavement system. A special case of infiltration is likely to occur during winter and spring thaw periods. As illustrated by case 3b in Fig. 2-7, the presence of ice and snow accumulation on the shoulder may block surface drainage, causing water stagnation on gravel shoulders. This phenomenon combined with the possible presence of a frozen layer within the pavement structure may cause water to accumulate in the granular base near the pavement surface. This situation and the resulting problems it may cause to pavements will be discussed further in Chap. 3.

Frost Action Frost action is one of the important sources of excess water within the pavement system. It is widely accepted that three conditions are required for frost heave to occur in a pavement system: (1) the temperature must be sufficiently cold for a long enough period to allow for phase change of the interstitial water, (2) water must be available and allowed to flow freely to the freezing front and, (3) freezing soil must be frost susceptible. Cold region pavements are, to varying degrees, subjected to these three conditions and are likely to experience frost heaving during winter. 1. Temperature: In cold regions temperatures are low enough to induce freezing of the pavement system. Figure 1-1 illustrates areas where pavements are subjected to significant freezing. In North America, the northern half of the United States and most of Canadian territory are subjected to temperatures that are cold enough to induce frost penetration underneath pavement structures. In Eurasia, Russia, China, Mongolia, Finland, Sweden, Norway, and mountainous parts of central Europe are also subjected to substantial frost penetration. 2. Moisture: As indicated above, moisture is generally available in pavement systems through capillary rise from the water table, infiltration of moisture from precipitation and lateral moisture transfer. All these sources of water are likely to supply frost action in pavements.

27

28

Chapter Two 3. Frost susceptibility: This is a soil property that describes the ability of water to flow toward a growing ice lens behind the freezing front. High negative pressures are generated at the base of the ice lens letting water flow in a partly frozen zone of soil between the freezing front and the ice lens. Fine-grained soils are generally more frost susceptible while coarse-grained soils are less. Frost susceptibility of soils will be further discussed in this section as well as in Chap. 4. During winter, frost penetrates pavement materials and subgrade soils. While progressing in the pavement structure, frost causes interstitial water to expand and can also cause some segregation ice to form in the unbound granular materials. Although these phenomena are not considered to cause excessive frost heave in pavement granular material, their significance comes from the fact that they cause the materials to loosen. Heaving of pavement surfaces reaching 10 mm has been observed on experimental test sites (Doré 2004) before the frost front reached the subgrade soil. When the frost front reaches the frost-susceptible subgrade soils, water is sucked toward the segregation front where ice lenses are formed. Heave of the pavement surface, resulting from the segregation ice formation can reach and even exceed 150 mm for climatic conditions prevailing in northern countries. The importance of this phenomenon for cold region pavements warrants a detailed description. Soils at rest with uniform temperature are in thermodynamic equilibrium, which means they are in thermal equilibrium, in chemical equilibrium and in mechanical equilibrium (Henry 2000). As discussed in Sec. 2-1, during the cooling process leading to pavement freezing, a thermal gradient is induced in the system, breaking the thermodynamic equilibrium and causing the system to react and try to regain equilibrium. Work by Taber, Beskow, Everett, Miller, Loch, and Gilpin (summarized in Henry 2000) has led to the current understanding of frost heave mechanism. The process is complex and involves the combined effect of thermal and chemical potential at particle/ice/ water interfaces acting against the mechanical contact between particles. The following paragraphs attempt to describe in a simple way the frost heave mechanism. As shown by experimental data in Fig. 2-9, when a cooling soil mass reaches freezing temperature (0°C for solute-free water), ice begins to form in pores and the unfrozen water content begins to decrease. Temperatures slightly below 0°C are required to force the initiation of ice crystal formation (nucleation) in the pores. The latent heat generated by the nucleation then raises the temperature near the freezing point before it starts to decrease again. Unfrozen water content then decreases progressively as temperature drops until it reaches a residual level of approximately 3.5 percent at about −2°C. At that point, most of the free water and capillary water is frozen and only hygroscopic water remains unfrozen in the soil matrix. The unfrozen water content then decreases very slowly with decreasing temperature. Transposed to a soil column subjected to freezing under a temperature gradient, this situation creates three distinct zones in the column. The lower zone is characterized by temperatures above the freezing point and interstitial water is completely unfrozen. The intermediate zone, comprised between the freezing temperature and the residual level, is characterized by partly frozen interstitial water. Free water and ice coexist in this zone, but their relative proportion changes with temperature. Finally, in the top part of the column, soil water is mostly frozen leaving only a relatively low proportion of hygroscopic water unfrozen. The presence of a zone where free water and ice coexist is the basis of most recent theories on frost heaving mechanisms.

Pavement Environment

FIGURE 2-9 Unfrozen water content as function of temperature: (a) experimental data (Doré et al. 2004) and (b) transposition to soil column subjected to freezing under thermal gradient.

The zone where free water and ice coexist can be several tens of centimeters thick in typical freezing pavement situations. It is a place of significant thermodynamic instability. As illustrated in Fig 2-10, the freezing front corresponds to the lowest temperature at which ice can form in the pores. As indicated by the phase diagram for water, at lower temperatures (i.e., higher in the freezing column), ice exerts more pressure on water

FIGURE 2-10

Partly frozen soil layer in freezing soils and acting pressures.

29

30

Chapter Two and occupies more space in the pores thus creating smaller radii ice/water interfaces. The water film in contact with ice, referred to as the contractile skin (Fredlund and Rahardjo 1993), acts like a membrane in tension and is opposing ice pressure by exerting a tensile stress conferring a negative pressure to interstitial water. This situation has many similarities with water rise in a capillary tube. A narrow capillary will cause the air/water interface meniscus to have a short radius and will increase the negative pressure in the water film causing a high rise in the tube. Suction in frozen soils can be described and quantified by the thermodynamic theory and using the generalized Clausius-Clapeyron equation (Ladanyi and Shen 1989): pi pw − π ∆T − = −L Tf ρi ρw

(2-15)

where pi and pw are ice and hydrostatic pore water pressure, ri and rw are ice and water densities, p is the osmotic pressure associated with leachable solutes, L is the latent heat of fusion, Tf is the freezing temperature, and ∆T is the freezing point depression. Using appropriate values in Eq. (2-15) (i.e., ri = 916.8 kg/m3, rw = 1000 kg/m3, L = 334 kJ/kg, and Tf = 273.15 K), neglecting p for pure water and assuming constant ice pressure in the partly frozen zone, the differential water pressure (dpw/dT) can be estimated to be (Ladanyi and Shen 1989): dpw kPa = 1220 dT K

(2-16)

The differential pressure existing between the top and bottom of the partly frozen soil zone creates conditions favorable to water flow from the bottom of the zone toward the top. Two other conditions must exist for water to flow toward an eventual ice lens. The partly frozen zone must remain permeable and water must be removed at the end of the flow path. Similarly in the capillary tube, no flow will occur in the tube unless water is removed from the top of the water column. If, for example, it is removed through contact with blotting paper, water will flow up the tube to replace the extracted water. Permeability of the partly frozen zone of the freezing soil column varies from unfrozen soil permeability at the bottom to essentially no permeability at the top. Significant permeability is thus likely to be found in the bottom portion of the partly frozen zone of the soil column. The permeability of a partly frozen soil will remain significant, if a continuous film of free water continues to exist between the ice and hygroscopic water within soil pores. This is a soil characteristic that is essentially related to its frost susceptibility. Water removal in a sealed system (mostly frozen soil at the top of the column) needs to occur within the system. This occurs when the hydraulic pressure caused by the negative pressure gradient in unfrozen water exceeds the overburden pressure in the partly frozen zone, in other words, when the effective stress becomes null (Henry 2000). Soil grains are thus separated and an ice lens begins to grow, hence removing water from the underlying partly frozen soil. If all these conditions exist in the frozen zone and assuming saturated conditions, Darcy’s law of water flow in soils should apply: v = kf ×

dpw dx

(2-17)

Pavement Environment where v is the rate of flow in the partly frozen zone and kf is the effective hydraulic conductivity in the partly frozen soil. Consequently, the zone of interest in the partly frozen soil layer is a relatively thin layer of soil comprised between the growing ice lens (segregation temperature, Ts) and the freezing front (freezing temperature, Tf). This layer of soil has been termed the frozen fringe (Miller 1972). Note that the frozen fringe is a much thinner layer than the partly frozen zone described above. Experimental data has shown that the thickness of the frozen fringe was in the range of a few millimeters (Loch and Kay 1978, quoted by Ladanyi and Shen 1989). Within the frozen fringe most of the conditions described above exist and water migration to the growing ice lens is possible. Above the ice lens, the low hydraulic conductivity of the partly frozen soil and the lack of external source of water restrict water movement to a limited redistribution of water within the partly frozen zone. Another important aspect of the frost heave mechanism is that ice lens formation generates a large quantity of latent heat, which will oppose congelation. Ice lenses will continue to grow only if latent heat is effectively removed from the system. As discussed in Sec. 2-1, the effectiveness of the system to remove heat is described by Fourier’s law of heat transfer by conduction [Eq. (2-1)]. Rewording the three basic conditions for occurrence of frost heave given at the beginning of the section, it can be stated that (1) removal of heat (temperature), (2) removal of water through ice lensing (frost susceptibility), and (3) supply of water are required for an effective frost heave mechanism (Henry 2000). Frost heave in soils is, thus, the result of the combined action of heat and moisture transfer in freezing soils. For aforementioned reasons and as illustrated in Fig. 2-11, freezing soils in a pavement system are subjected to a thermal gradient. As a result the temperature differential induces a negative pore water pressure gradient, but also creates variable hydraulic conductivity conditions. Based on this understanding of the frozen fringe conditions and assuming that the validity of Fourier’s law for heat transfer, Clausius-Clapeyron’s equation for pressure conditions, and Darcy’s law for water flow all were valid in the frozen fringe

FIGURE 2-11 Thermodynamic conditions in frozen fringe (modified from Konrad and Morgenstern 1983, with permission of National Academies Press).

31

32

Chapter Two conditions. Konrad and Morgenstern (1980) have developed the segregation potential concept to model one dimensional frost heave in soils. According to the model, the frost heave rate (v) is directly proportional to the thermal gradient (TG) in the frozen fringe following Eq. (2-18): v = SP × TG

(2-18)

where SP is the proportionality constant also termed the segregation potential. In this relationship, the term TG is related to pressure gradient through the Clausius-Clapeyron equation and is thus an expression of the driving force imposed by the pavement thermal regime. The term SP is a material- and site-condition-specific characteristic that translates the propensity to frost heaving. It can be seen as the hydraulic conductivity of the frozen fringe under a given thermal gradient. The segregation potential can thus be considered as a mechanistic frost susceptibility index. Considering a system subjected to freezing under fixed surface temperature (Tt) and bottom temperature (Tb) (Fig. 2-12) the segregation ice formation process can be summarized as follows: • A first ice lens will be initiated at a shallow depth. The temperature gradient in the frozen fringe is initially very steep, resulting in a very high rate of heat extraction. Despite the fact that the resulting negative pressure gradient is very high, the ice lens formation mechanism rapidly becomes ineffective due to the rapid cooling of the system, which reduces the hydraulic conductivity of the frozen fringe. • The freezing front progresses downward and a new ice lens is initiated at a location where the effective stress becomes null. The first ice lens might continue to grow but only with redistributed water available between the two ice lenses. • As the freezing front progresses downward, the thermal gradient is reduced and the net cooling of the frozen fringe is reduced. Thus, the growth of ice lenses is slower but lasts longer. Ice lenses become thicker and are more widely spaced due to the thicker frozen fringe. • Near steady-state conditions, the ice lens will grow as long as the system can effectively extract heat from the frozen fringe.

FIGURE 2-12 Schematic illustration of rhythmic ice lens formation (modified from Konrad and Morgenstern 1980, with permission from Canadian Geotechnical Journal).

Pavement Environment The last important consideration for moisture intake in the pavement system resulting from frost heave is that moisture is accumulated in solid form over a relatively long period of time and is released in a relatively short time during spring thaw. The large quantity of water released by melting ice lenses is likely to make pavements unstable during spring thaw. In this case, there is a coupled heat transfer-consolidation problem. The severity of the problem is related to three important factors: 1. The quantity of water accumulated in ice lenses per unit thickness of soil 2. The rate at which water is released (or the rate of progression of the thaw front) 3. The rate at which water is evacuated by the consolidation process The thaw-weakening problem will be discussed further in Chap. 3.

2-2-3 Factors Contributing to Moisture Extraction from the Pavement System Evaporation Evaporation is a rather important factor of moisture extraction from pavement systems. Obviously, the presence of an impervious surfacing layer will strongly reduce the amount of evaporation over a large proportion of the pavement surface. However, as for water infiltration in pavements, effective evaporation can still occur on gravel shoulder and embankment slopes as well as through pavement cracks. The rate of evaporation at the surface of a pavement system is a function of several factors. Two conditions are needed for effective evaporation: 1. Temperature must be sufficiently high to favor water vaporization and to supply the phase change reaction. 2. Vapor must be effectively transported away from the vaporization front. The second condition involves several factors including vapor pressure in the air, wind speed, and surface roughness. Wilson has formulated Eq. (2-19) to calculate the evaporation rate from unsaturated soil surfaces (Fredlund and Rahardjo 1993): Ev =

ΓQn + ηE Γ + ηA

(2-19)

where Ev is the vertical evaporative flux (mm/day), Γ is the slope of the vapor pressure versus temperature curve, Qn is the heat budget of all net radiations, h is a psychrometric constant, E = (0.35 + 0.051W)·uav (B − A), W is wind speed, uav is the vapor pressure of the air above the evaporating surface, B is the inverse of relative humidity in air, and A is the inverse of the relative humidity at the soil surface. Even if evaporation is relatively small in pavements due to the presence of the surfacing layer, the pavement system is significantly affected by evaporation and drying of adjacent soils. Moisture will be transferred laterally from the pavement system to dryer soils in an attempt to reach equilibrium.

Pavement Drainage The moisture regime of a system is, of course, strongly affected by all sources of moisture inputs. It is also strongly affected by the capacity and the effectiveness of the

33

Next Page

34

Chapter Two moisture outlet of the system. The drainage system of a pavement generally includes some or all of the following: • Surface drainage • Internal pavement drainage • Collectors and evacuation systems As discussed in Sec. 2-2, water from precipitation will either seep into the pavement system or flow at the surface toward the nearest water collector. An effective surface drainage involves an impervious surface and an appropriate cross-slope. Runoff surface water will eventually reach a collector and be evacuated. Infiltrated water will percolate vertically in the pavement structure until it reaches a low permeability surface, such as the interface between the subbase and fine grained subgrade soils or an internal drainage layer. In the first case, water will partly seep into the subgrade soil, but will also flow along the interface toward a collector (ditch or internal drain) if an adequate cross-slope has been prepared during pavement construction. If water is intercepted by a drainage layer (such as those illustrated in Fig. 1-7a), it will flow in the drainage layer following the effective drainage gradient. In the case of pavement drainage, the drainage gradient is not necessarily the slope between the location of the drop of water to be drained and the collector as would be the case for saturated soils. Pavement layers and subgrade soils immediately underneath the pavement are generally unsaturated and characterized by negative pore pressures. Thus gravity is not the only driving force acting on water. The driving potential or hydraulic head (hw) acting on the water phase of unsaturated soils is equal to (Fredlund and Rahardjo 1993): hw = y +

uw ρw g

(2-20)

where y = gravitational head and uw /rw g is the pressure head with uw = pore water pressure, rw = density of water, and g = gravitational acceleration. The hydraulic gradient, i, in the x direction for an unsaturated soil, thus, becomes i=

dhw dx

(2-21)

This can lead to situations where water tends to flow in different directions than expected, as illustrated in Fig. 2-13. Despite the fact that these systems are designed to extract moisture from pavements, drainage collectors and layers can sometimes feed moisture into unsaturated pavement systems. The presence of air in pores of unsaturated soils also affects the effective hydraulic conductivity of these soils. Permeability of the water phase, kw, in unsaturated soils decreases as the level of saturation decreases and as the matric suction increases. Using the following relationship, kw can be estimated (Fredlund and Rahardjo 1993): k w = k s Sδe

(2-22)

where ks is the permeability of the saturated soil, Se is the effective degree of saturation, d is an empirical constant = (2 + 3l)/l, l is the pore size distributions index that is the slope of the relationship between the effective degree of saturation (Se) and matric suction (ua − uw).

Pavement Environment

FIGURE 2-13 Unexpected flows in pavements due to unsaturated moisture regimes.

Given the above considerations, Darcy’s law for unsaturated soils becomes dhw (2-23) dx where vw is the flow rate of the water phase in unsaturated soils. The effectiveness of pavement drainage is also considerably affected by the effectiveness of the collector and evacuation systems. If these systems have an adequate capacity (dimension and slope), runoff water and water drained out of the pavement system will be effectively removed. If not, water will tend to stagnate within the system resulting in unfavorable moisture conditions. vw = − k w

2-2-4

Moisture Balance

All the factors described in this section will contribute to a constantly evolving moisture balance in the pavement system. Figure 2-14 illustrates a typical moisture balance at two levels in the pavement structure, based on experimental data collected from the Dickey

FIGURE 2-14 Moisture balance at two levels in pavement system (modified from Janoo and Greatorex 2002).

35

36

Chapter Two Lake test site in Montana by Janoo and Greatorex (2002). The top half of Fig. 2-14 illustrates the moisture balance in the base material (depth of 292 mm), while the bottom half illustrates the moisture balance in the subgrade soil (depth of 584 mm). In both cases, the evolution of moisture balance with time can be divided into four distinct seasons: • During fall, heavier precipitations combined with lower evaporation at the surface of the pavement system cause moisture content to increase slightly during the months of October and November. • At the end of November, freezing temperatures cause interstitial water to freeze. For pavement engineering purposes, this can be considered as the beginning of winter. At the beginning of the winter season, unfrozen water contents of the base material and of the subgrade soils drop close to 0 percent. This process occurs slightly later and more progressively for the subgrade soil than for the granular base material. The trend observed is the result of the limitations of measuring instruments that typically measure unfrozen water content in soils instead of total water content. In reality, the total water content in soils and granular materials will typically increase during winter as a result of water and brine infiltration in pavements as well as the formation of segregation ice. The evolution of the total water content (in solid and liquid forms) will typically evolve from the highest content measured prior to freezing to the peak observed in early spring thaw as indicated by the dotted lines in Fig. 2-14. The pattern of water intake during winter for subgrade soils is well documented through several seasonal pavement monitoring projects (Palolahti et al. 1993; Imbs and Doré 2003; and several others). An increase of volumetric water content from about 8 to about 35 percent is observed in the silty sand subgrade of the Dickey Lake test site (Fig. 2-14) and is probably associated with segregation ice formation. The increase of volumetric moisture content from about 5 percent to more than 20 percent in the granular base material cannot be explained by the sole effect of water infiltration in the granular base layer. It is likely the result of a combined effect of frost heave, vapor transport, and water infiltration in the base material. The pattern of moisture intake could, thus, vary between patterns (a) and (b) in Fig. 2-14. More research is needed to fully understand the behavior of granular pavement materials submitted to frost action. • Spring is the season when moisture balance varies the most. During spring, water accumulated during winter is released over a relatively short period of time as indicated by the sharp increase in unfrozen water content in the base and subgrade soils. At the same time, precipitation and melting of snow accumulated on the shoulders and on the slopes of the embankment contribute to high moisture contents in the system through infiltration. Excess water in the system will progressively drain out of the system in the late spring and early summer seasons. • During summer, the residual excess moisture in the pavement system will continue to drain. Moreover, evaporation will become more and more effective as the embankment surface becomes warm and dry. Under these conditions, the moisture contents in the granular materials and subgrade soils will decrease progressively during the summer months. Intense precipitation may affect the moisture balance temporarily, as evident for the month of June in Fig. 2-14. The effect of that specific precipitation event is more apparent in the subgrade soil than in the pavement base.

Pavement Environment Moisture regime in pavements has an important effect on its response to temperature variations and to external stresses. Good pavement performance requires low moisture contents with minimum fluctuations thorough the year.

2-3

Stress Regime in Pavements Pavements are engineered structures designed to sustain stresses. Pavement engineering relies therefore on a good understanding of pavement stresses and on their quantification. Stresses acting on pavements can be classified as static or cyclic stresses. They can further be classified as load induced or environmentally induced stresses as shown in Table 2-1. Static stresses, described in Secs. 2-3-1 and 2-3-2, induced in the pavement system by loads, include the earth pressure at rest and the weight of equipment or vehicles standing on the pavement. The latter case should be considered in situations where vehicles are likely to be moving slowly or stopped for short or long periods. These include parking areas and intersections for roadways and aprons, taxiways and runway extremities for airport pavements. Cyclic stresses, covered in Secs. 2-3-4 to 2-3-7, induced in the pavement system include stresses induced by moving traffic loads and by seasonal environmental factors:

2-3-1

Earth Pressure at Rest

Geostatic stresses caused by the weight of soils and pavement materials are important to consider when assessing stresses in the pavement materials. They represent the initial stress state at a given point in the system prior to external loading. Since mechanical properties of soils and unbound pavement materials are stress dependent, the correct assessment of geostatic stresses is required for the mechanical analysis of the pavement system. As illustrated in Fig. 2-15, effective earth pressure at rest is the result of the summation of two factors: (1) stresses associated with the weight of soils and materials

Static stresses

Induced by loads

Earth pressure at rest Static traffic loads

Cyclic stresses

Induced by long-term environmental and soil effects

Stresses related to consolidation or other permanent soil movements

Induced by loads

Moving traffic loads

Induced by seasonal environmental factors

Thermal stresses Stresses related to frost heave Negative or positive pore pressure

Highlighted factors are considered most frequently in pavement engineering.

TABLE 2-1

Summary of Stresses Acting on Pavements

37

38

Chapter Two

FIGURE 2-15

At rest earth pressures in pavements.

above the considered point in the pavement system and (2) pore water pressure. The first factor is usually expressed in terms of vertical stress sv:

σ v = ∑ γ i hi

(2-24)

in terms of horizontal stress (at rest) sh:

σ h = K 0 ∑ γ i hi

(2-25)

θ = (1 + 2K 0 )∑ γ i hi

(2-26)

or in term of total stress (at rest) q:

where gi is the unit weight of soil or pavement material in layer i, hi is the thickness of layer i, K0 is the coefficient of lateral earth pressure at rest (Das 2002 or other geotechnical engineering reference). Pore pressure, u, can be determined as u = γ w hw

(2-27)

where gw is the unit weight of water and hw is the difference between the depth of the point considered in the analysis and the surface of the water table. The effective stress s ′ can be obtained from

σ′ =σ −u

(2-28)

Note that the effect of negative pore pressure in unsaturated soils above the water level has not been considered in the previous equations. This specific aspect will be discussed in Sec. 2-3-7.

2-3-2

Static Stresses Induced by Traffic Loads

The evaluation of the stress distribution caused by traffic loads into the multilayer pavement systems is a complex problem. For practical reasons, pavement engineers

Pavement Environment

FIGURE 2-16

Stress-strain behavior of an elastic system.

have reduced the problem to the simplest possible form. Amongst the simplifications commonly used, the assumption of elastic behavior of pavement materials is very convenient. Since a pure elastic behavior is not time dependent, this assumption allows engineers to use a static load to compute stresses into a pavement structure. According to the basic law of the theory of elasticity, referred to as Hooke’s law, strain e in an elastic material is proportional to the applied stress s. The proportionality constant is termed the modulus of elasticity E and is a basic material property for material engineering. Hooke’s law can be formulated as follows:

σ = Eε

(2-29)

As opposed to the viscoelastic behavior also commonly used in pavement mechanics, strains will develop instantaneously in an elastic system subjected to a pulse load and will remain constant until the load is removed. The two behaviors are illustrated in Fig. 2-16. Static stress analysis, is thus, a practical way to assess stresses and strains in a pavement system under a stationary load as well as under a moving load considering the hypothesis of elastic behavior. The latter hypothesis has proven to be realistic in the context of pavements in “normal” operating conditions subjected to loads that are relatively small compared to the failure load (Brown 1993). Like most engineering materials, pavement materials and soils have the ability to distribute loads. Thickness and stiffness of the material will both contribute to load distribution. The simplest mathematical representation of load distribution with depth was given by Boussinesq in 1885 for a homogeneous half space. Considering a load uniformly applied on a flexible circular plate, the vertical stress sz and the radial (horizontal) stress sr at depth z, under the center of the plate of radius a can be obtained from (Ullidtz 1987):   z3 σz = σ0 1 − 2 2 1.5  ( ) a + z  

(2-30)

39

40

Chapter Two and

σr =

σ0 2

  z3 2 z(1 + µ )  (1 + 2 µ ) − (a 2 + z 2 )0.5 + (a 2 + z 2 )1.5   

(2-31)

where s0 is the uniform pressure applied on the plate and m is Poisson’s coefficient. Despite the fact that the contact area of a truck tire has an oval-type shape, the hypothesis of a load transmitted by a circular flexible plate is generally considered to be reasonable for the analysis of load induced stresses in pavements (Peattie 1962). Boussinesq equations also allow calculating vertical and radial strains as well as vertical displacements in the homogeneous half space. These equations are developed further in Sec. 5-9. The resulting load distribution under the center of the plate in the homogeneous mass is illustrated in Fig. 2-17. A detailed description of the Boussinesq method and equations can be found in Ullidtz (1987). Pavement systems are obviously far from being homogeneous masses. The Boussinesq approach thus has limited applications, such as in the analysis of load distribution in multilayer pavement systems. A mathematician by the name of Burminster (1943a and b) formulated a complete theory on the calculation of stresses and strains in a multilayer elastic system based on the following principal assumptions: • Each layer is considered to be composed of homogeneous and isotropic material • The layers have a finite thickness, but are considered infinite in the horizontal direction • The multilayer system is resting on an infinite half space, that is, the last layer of the system has an infinite thickness • The layers are perfectly bound at their interfaces • The conditions for continuity at the interfaces are satisfied

FIGURE 2-17 Stress distribution in a homogeneous mass under a 150-mm radius circular plate for m = 0.35.

Pavement Environment The outcome of Burminster’s work is a complex set of mathematical functions allowing for the calculations of stresses and displacements at any point in space on the multilayer system subjected to loading. The vertical and the radial stress equations are given in Eqs. (2-32) and (2-33) as an example of Burminster’s theory.

σz =

∂  ∂ 2φ  (2 − µ )∇ 2φ − 2   ∂z  ∂z 

(2-32)

∂  2 ∂2φ  µ∇ φ − 2   ∂z  ∂z 

(2-33)

σr =

where z is the depth, m is Poisson’s coefficient, ∇ is a Laplacian used in compatibility equations, and f is a Bessel stress equation computed for each layer of the system. Burminster’s publications also include equations for the calculation of radial, shear and bulk stress, vertical and radial strain as well as vertical displacement. Due to the complexity of the mathematic formulation of Burminster’s equations, several methods based on charts and tables have been published (Jones 1962; Peattie 1962; and others). Figure 2-18 illustrates Burminster’s solution for the vertical stress distribution under the center of a circular plate of radius a in a two-layer system for different elastic modulus ratios E1/E2. The slope of the lines in Fig. 2-18 illustrates the importance of the stiffness of the top layer and the effectiveness of load distribution on the bottom layer of the system. Figure 2-19 is inferred from Fig. 2-18 and illustrates how successive layers of decreasing stiffness distribute stresses in the pavement system. It illustrates how the load, initially applied over the contact area between the tire and the pavement, is distributed over much wider areas at pavement layer interfaces. Figure 2-19 illustrates the interfaces in the pavement structure at which the stresses or the strains are considered to be critical to pavement performance. These interfaces and

FIGURE 2-18 Distribution of vertical stresses in a two-layer system for different E1/E2 ratios (modified from Burminster 1958, quoted by Huang 2004; reprinted by permission of Pearson Education, Inc., Upper Saddle River, N.J.).

41

42

Chapter Two

FIGURE 2-19 Stress distribution in pavement system and stresses at critical interfaces associated with resulting pavement performance problems.

the resulting performance problems associated with excessive stresses or strains are as follows: • Vertical stresses or strain near or at the surface of the pavement, associated with the development of rutting due to permanent deformation of the bituminous wearing course • Horizontal stresses or strains at the bottom of bound layers associated with the development of fatigue cracking • Vertical stresses or strains at the surface of the subgrade soil associated with the development of structural rutting Calculation of stresses and strains at these critical interfaces is the basis for mechanistic empirical pavement design and pavement analysis methods. These methods will be described in Chap. 8.

2-3-3

Stresses Related to Permanent Soil Movements

Important static stresses can be induced in pavement systems by permanent soil movements such as differential settlement, soil creep, and solifluction. In permafrost conditions, consolidation resulting from the degradation of ice-rich permafrost is a widespread soil movement problem. These types of soil movements typically induce large distortions of the pavement surface and subsequently, tensile stresses develop in bound pavement layers. If the movements are slow, the stresses may be relaxed by the viscoelastic behavior of the asphalt bound materials. If the movements are rapid or if the material is brittle due to cold temperature or aging, cracks that traverse to the surface of the pavement are likely to occur.

Pavement Environment

2-3-4

Moving Traffic Loads

The aforementioned practical considerations often lead to the assumption of static conditions for the analysis of traffic loads. It is, however, generally agreed that traffic loading is a complex phenomenon that would ideally require analysis using dynamic considerations. When applied to pavement loading, dynamic effects can be classified in three main categories: 1. The evolution of stresses in pavement structures subjected to loading by a moving wheel 2. Time-dependent material response 3. Stresses induced by an oscillating load

Evolution of Stresses Each element of the pavement is subjected to a stress history that evolves as a function of the position of loading wheel relative to the loaded element of soil or pavement material. As indicated in Fig. 2-20, as the load approaches the considered element, the principal stresses s1 and s3 are acting obliquely on the element. As a result (illustrated in Fig. 2-20b), a shear stress that is arbitrarily indicated as positive is generated. The shear stress increases as the load approaches a certain point where the effect of the oblique loading decreases and becomes null when the load is immediately above the element. The shear stress then becomes negative when the load moves away from the element. The negative shear stress increases as s1 becomes oblique and it later diminishes as the load moves away from the element. The horizontal and vertical stresses keep increasing until the load is above the element and decreases as it moves away from it. This phenomenon, referred to as the rotation of the principal stresses due to the

FIGURE 2-20 Stresses under a moving wheel (a) on a soil or unbound material element within the pavement system and (b) at the bottom of the bound surfacing layer.

43

44

Chapter Two . passage of a wheel, is described by several authors (Blazejowski et al. 1996; Lekarp and Dawson 1998; Barksdale et al. 1998). It is believed to have an important effect on the mechanical performance of soils and unbound layers in pavements. Stresses in the bound layer(s) also evolve considerably as the loading wheel passes above a given point on the pavement. As indicated in Fig. 2-20a, the asphalt-bound layer bends under the wheel load, inducing tensile stresses at the bottom of the layer. On each side of the wheel, the bound layer is subjected to inverse bending, inducing compressive stresses at the bottom of the layer. As a result, a wheel approaching a given element at the bottom of the asphalt-bound layer will first induce compressive stresses (represented by positive values in Fig. 2-20a) followed by a sharp increase in tensile stress. The maximum value of tensile stress occurs when the load is at the vertical of the considered element. The pattern described will be inversely reproduced as the wheel moves away from the loaded element.

Time-Dependent Material Response Assumption that materials have an elastic behavior makes them insensitive to loading time. Thus, the elastic theory fails to explain permanent deformation in pavement structures. It also fails to explain the effect of wheel speed on strains illustrated in Fig. 2-21. According to literature data compiled by Coulombe (2002), strains under a slowly moving or stopped vehicle can be more than twice those under a vehicle moving at 70 km/h. More complex rheological models have to be used to explain material responses under a moving load. Figure 2-22 illustrates two models used in pavement engineering to represent the rheological behavior of pavement materials. Figure 2-22a illustrates Burger’s viscoelastic model commonly used to model the response of materials such as asphalt bound materials. The elastic element (E) of the model will cause the material to deform instantly under a loading pulse. The viscoelastic (VE) elements will induce a delayed elastic response characterized by a decreasing rate of strain with time. The viscous (V) element will in turn accumulate strain linearly with time as long as the load is maintained on the element. When load is removed from the viscoelastic material, the

FIGURE 2-21 Effect of vehicle speed on (a) horizontal strains at the bottom of the asphalt layer and (b) vertical strains at the top of subgrade soils (modified from Coulombe 2002).

Pavement Environment

FIGURE 2-22 Rheological models used in pavement engineering: (a) Burger’s viscoelastic model and (b) Rowe’s visco-elasto-plastic model.

elastic response is immediate and the viscoelastic response is delayed in time. The viscoelastic element will restore all the accumulated strain if unloading time permits, while the displacement accumulated by the viscous element is irrecoverable allowing for permanent deformation to accumulate in the system with load repetitions. The equation of the Burger’s model is the following (Huang 2004):

ε=

σ E0

  t  t  σ   1 + T  + E  1 − exp  − T   0 1  1   

(2-34)

where e is total strain, s is applied stress, E0 and E1 are the elastic modulus of the elastic element (E) and the elastic component of the viscoelastic element (VE), respectively, T0 and T1 are relaxation times T = l/E, where l is viscosity and t is time. Figure 2-22b illustrates a rheological representation of a visco-elasto-plastic model presented by Rowe (Rowe et al. 1995, cited by Brown 1997). The model was used by Rowe to compute dissipated energy related to fatigue cracking and permanent deformation in dense bitumen macadam. The model is also likely to adequately represent the behavior of unbound granular materials. The elastic component of the visco-elastoplastic model will cause an important instantaneous deformation upon loading of the system. The visco-plastic (VP) element of the system will then linearly accumulate

45

Chapter Two deformation as long as the load acts on the element. The elastic deformation will be recovered instantaneously when load is removed leaving a residual visco-plastic permanent strain. Climatic conditions have an important effect on the rheological behavior of pavement materials. The elastic component will dominate the viscoelastic behavior of asphalt bound materials at low temperatures while the viscous component will have a significant effect on material response at high temperatures. Similarly, for relatively dry soils or unbound granular materials, elastic and plastic behaviors are dominant, while viscous behavior will prevail for soils or materials at or approaching saturation.

Stresses Induced by Oscillating Load The third aspect of dynamic loading of pavement structures is related to stresses induced by an oscillating load. If pavement surfaces were perfectly uniform, the load applied by a moving wheel would remain constant with time or distance on the pavement. In reality, pavements are not uniform. As illustrated in Fig. 2-23, irregularities in longitudinal profiles induce oscillations of the suspended masses of a vehicle, causing variations in the loads transmitted to the pavement structure. Dynamic loads caused by oscillating vehicle mass can vary substantially depending on surface roughness, speed, and characteristics of the vehicle, such as suspension type. Equation (2-35) (OECD 1988) represents the load equivalence law adapted for uneven pavements: P  Ni = α s  Ns  Pi 

γ

(2-35)

where Ni is the number of applications of a given load Pi causing pavement failure, Ns is the number of applications of a standard load Ps causing pavement failure, g is the “aggressiveness” factor generally considered as 4 for flexible pavements, and a is the dynamic load coefficient equal to Wd /Ws (see Fig. 2-23).

Dynamic load Wd Wheel load

46

FIGURE 2-23

Constant (static) load Ws

Dynamic load induced by an oscillating mass.

Pavement Environment Experimental studies in Belgium (OECD 1988) have shown that typical values for the coefficient a are 1.06 for fairly even surfaces, 1.24 for average surface conditions, and 1.54 for uneven road surfaces. Bad surface conditions can, thus, lead to a reduction of allowable load applications on a pavement reaching 50 percent. Results from an experimental study at the CAPTIF facility in New Zealand (Steven et al. 1999) showed that the dynamic load coefficient for parabolic spring suspension doubles when speed increases from 20 to 45 km/h, while it is not affected by speed for multileaf spring suspensions. Uneven loading resulting from vehicle oscillation is also an important cause of differential permanent deformation causing pavement roughness (Steven et al. 1999; Ullidtz 2002).

2-3-5 Thermal Stresses Thermal stresses are environmental stresses caused by diurnal and shorter-term temperature variations in the bound pavement layer that is restricted from contracting. As the temperature gets colder, a thermal stress starts to develop gradually. Figure 2-24 illustrates the development of thermal stress under constant cooling rate. The stress does not increase linearly with temperature at warm temperatures (close to 0°C) due to asphalt cement’s viscoelastic behavior that allows partial relaxation of stresses. At a certain transition temperature depending mainly on the binder properties, the asphalt concrete starts to behave as a pure elastic material and the thermal stress increases linearly with the temperature. When the thermal stress reaches pavement tensile strength, cracking occurs. The stress at which cracking occurs in the field is termed “cracking strength” and the corresponding temperature “cracking temperature.” The development of the thermal stress can be modeled with Eq. (2-36): T2

σ thermal = α ∫ SdT

(2-36)

T1

FIGURE 2-24 Typical stress versus temperature relationship (modified from Jung and Vinson 1994; reproduced with permission of TRB, from Transportation Research Record: Journal of the Transportation Research Board, No. 1417, Transportation Research Board of the National Academies, Washington, D.C., 1993, Figure 1, p. 13).

47

48

Chapter Two where a is the linear thermal contraction coefficient, S is temperature and loading time dependent stiffness of the asphalt concrete, T1 is the initial temperature, and T2 the final temperature. The material properties, a and S, in Eq. (2-36) can be measured or estimated using values reported in the literature. Asphalt concrete has two distinct contraction coefficients depending on the temperature range. At cold temperatures where the thermal stresses develop, a typical value for the contraction coefficient is 2.93 × 10−5/°C ( Jones et al. 1968). Zeng and Vinson (1998) have measured values for the linear contraction coefficient varying from 1.89 to 3.33 × 10−5/°C. The mixture stiffness is traditionally determined using binder stiffness and volumetric mixture properties (Christensen et al. 2003; Bonnaure et al. 1977; Heukelom and Klomp 1964). The binder stiffness can be predicted using loading time, temperature step, and binder properties (Van der Poel 1954; McLeod 1976). Currently, the binder stiffness is measured using the bending beam rheometer (BBR) and used to predict the pavement thermal stress through complex algorithms performed by specific software (AASHTO Provisional Standards 2003). The thermal stress restrained specimen test (TSRST) described in Chap. 4 can be used to directly measure the thermal stress as a function of temperature. In the TSRST a thermal stress is developed as illustrated in Fig. 2-24 by keeping an asphalt aggregate mixture specimen at a constant length while cooling it down at a standard rate. The experiment captures the mixture’s fracture strength and the fracture temperature that are generally representative of field conditions. As the thermal stress develops in a pavement slab, a resisting stress develops simultaneously that opposes the contraction of the pavement slab. The maximum resisting stress that can be developed is the shear strength at the interface of the contracting slab and the underlying layer and can be represented as s = sv tanf + c

(2-37)

where sv is the vertical stress [Eq. (2-24)], f is the friction angle, and c is the cohesion at the interface (Zubeck and Vinson 2007). When the thermal stress exceeds the maximum mobilized restraint stress, or restraint strength, the pavement slab is able to contract and the thermal stresses are relaxed as further explained in Chap. 3.

2-3-6

Stresses Related to Frost Heave

As described in Sec. 2-3-3, important stresses can be induced in pavement systems by soil movements. The phenomena described in Sec. 2-3-3 are long-term irreversible soil movements. Similar movements can occur as a result of cyclic phenomena such as frost heave and thaw consolidation. Frost heave is rarely uniform and significant differential movements can be induced in pavements. Two main mechanisms, identified by Doré (2002), can act on pavements. The first one, referred to as random differential heaving, is mainly associated with variations of soil properties along the highway corridor. The resulting distortions tend to increase pavement roughness during winter and, by forcing pavements to bend upward, can also cause pavement cracking. The second type of differential frost heaving occurs along the transverse axis of the pavement and is the result of the variation of pavement geometry and snow accumulation on pavement sides. Both mechanisms are likely to force pavements to bend upward. The following

Pavement Environment model has been proposed by Doré et al. (1999) to estimate strains caused by upward bending of pavements caused by differential frost action:

εf =

χ tan θ L + (∆h tan θ)

(2-38)

where ef is the flexural strain caused by differential heaving, ∆h is the difference of heaving between two points separated by a distance L on the pavement. c and q are parameters describing the geometry of the flexion in the pavement structure (more details are given in Chap. 3). Assuming an elastic behavior of the asphalt concrete, stresses induced by the bending action can be calculated using Eq. (2-29), but if viscoelastic behavior is expected then Eq. (2-34) should be used. Differential frost heaving will be further discussed in Chap. 3.

2-3-7

Negative or Positive Pore Pressure

Water can be a major source of problems in pavements. Therefore, drainage is considered to be an essential precaution in pavement engineering. As a result of good design and drainage practice, pavement systems are mostly constituted of unsaturated soils and materials. The relationship between the degree of saturation and pore pressure is shown in Fig. 2-25, where pore pressure is represented by the balance between air pressure ua and water pressure uw or matric suction (ua − uw) in the pore. Matric suction is one of the two components of the total suction in unsaturated soil pores (the second component, osmotic suction, is not discussed in this book). The relationship is soil specific and is referred to as the soil-water characteristic curve. From the saturation state at atmospheric pressure (Point 1 in Fig. 2-25), a certain level of pressure is needed to force air penetration into soil pores (Point 2). This pressure is referred to as the air-entry pressure and is denoted (ua − uw)e. As air occupies more space in the pores (Point 3), pore water resists the intrusion of air by opposing an increasing tensile force through the contractile skin at the air-water interface. This phenomenon is similar to the one described

FIGURE 2-25

Relationship between the degree of saturation and matric suction in soils.

49

50

Chapter Two earlier for ice intrusion into soil pores (Sec. 2-2). Assuming that pore air is at atmospheric pressure, pore water pressure is a negative value (or matric suction) that tends to increase with decreasing level of saturation. Increasing pore air volume forces water to retreat in exiguous spaces in the pores (Point 4) reducing considerably the radius of curvature of the air-water interface and, thus, increasing the negative pressure in pore water as described by Eq. (2-39) (Fredlund and Rahardjo 1993). (ua − uw ) =

2Ts Rs

(2-39)

where Ts is the surface tension at the air-water interface and Rs is the radius of curvature of the air-water interface. Note that in Fig. 2-25, the path followed by a drying soil is not the same as the path followed by a wetting soil. The hysteresis in the soil-water characteristic curves is mainly due to nonuniform pore size diameter and distribution, the different contact angles between an advancing and a receding interface, as well as the presence of entrapped air in a wetting soil (Fredlund and Rahardjo 1993). The soil-water characteristic curves of two soils are transposed to a typical pavement situation in Fig. 2-26. In the illustrated case, the two unbound granular layers of the pavement are assumed to have the same water retention characteristics. Point 1 of Fig. 2-26 corresponds to the water table in the subgrade soil. At that point, soil is saturated and water is in balance with atmospheric pressure. Point 2 corresponds to the top of the capillary fringe. The soil is still saturated, but is affected by negative pore pressures corresponding to the air-entry pressure. At Point 3, fine-grained soils are in contact with pavement granular materials. The suction at the interface is in balance, meaning that

FIGURE 2-26 Saturation and matric suction regimes in a pavement system in dry (solid line) and damp (dashed line) conditions.

Pavement Environment the two types of soils need to coexist at different levels of saturation. At Point 4 in the pavement system, granular materials are at a relatively low level of saturation, which induces considerable matric suction in pavement materials. Values of matric suction ranging from 10 to 150 kPa have been measured in pavement granular bases by Perera et al. (2004). Matric suction has an important effect on stress state in soils and pavement materials. Bishop in 1959 (cited by Fredlund and Rahardjo 1993), has proposed the following expression of effective stress s′ in unsaturated soils:

σ ′ = (σ − ua ) + χ (ua − uw )

(2-40)

where s is the total normal stress, ua and uw are air and water pore pressures, c is a parameter varying between 0 and 1 and related to the degree of saturation. From Eq. (2-40), it is easy to see the importance of matric suction in the stress state of a soil element in the pavement system. For example, assuming a matric suction of 70 kPa, c = 0.7 and ua = 0, the second term of Eq. (2-40) would have a dominant effect on the effective stress (49 kPa) as compared to the first term for a soil element located 1.0 m below the surface (s ≈ 20 kPa). Since rigidity of soils and pavement materials is stress dependent, this situation translates into an increased resilient modulus. Doucet and Doré (2004) have proposed the following empirical relationship to relate saturated resilient modulus of granular base materials to unsaturated resilient modulus: ∆Mr sat = −8, 700(ua − uw ) − 17 , 000

(2-41)

where ∆Mrsat is the increase in resilient modulus from the saturated state to a given unsaturated state (kPa), (ua − uw) is matric suction (kPa). Based on Eq. (2-41), the resilient modulus of a saturated base material would increase by 157,000 kPa for a 20-kPa matric suction, which is typically measured for granular materials at levels of saturation around 20 percent. Obviously, climatic conditions and pavement characteristics are going to have an important effect on matric suction fluctuations in the pavement system. As indicated by the dashed line on Fig. 2-26, heavy precipitation and spring thaw conditions can cause near-saturation conditions in pavements. Matric suction is, thus, likely to be strongly reduced in those conditions. Positive pore pressures are also likely to occur for soils or materials decompacted by frost heave. The effective stresses and rigidity of pavement materials are consequently strongly reduced.

2-4

Interaction with Geology and Morphology As discussed throughout this chapter, temperature, moisture, and stresses are closely interrelated and constitute the environment in which pavement structures must perform. Pavement environment is also affected by several other factors, among which the most important are the geology and morphology of the surrounding areas. The morphology of the terrain in which the pavement is built has an important influence on the pavement environment. In high terrain, the water table is likely to be low and drainage is usually effective. In those conditions, water content within the pavement system tends to remain relatively low and constant through the seasons. Consequently, frost action is reduced by the reduction of the effective flow gradient

51

Chapter Two toward the ice lens. Moreover, effective stresses are high due to high suction levels in pavement materials. Pavements built in high lands are also more exposed to wind, increasing heat extraction by convection and are also likely to be more exposed to solar radiations. Pavements built in low land areas are likely to be affected by poor drainage conditions and high water tables. As a result, frost action is expected to be more severe and effective stresses in pavement materials are expected to be reduced. They are less affected by wind, but shading effects can reduce solar radiations on the pavement surface. These pavements can also be exposed to extreme low air temperature conditions due to the effects of temperature inversions (Dysli 1991). Pavements built in sloped areas are generally easy to drain, but they are more likely to be affected by seepage and artesian flow conditions. These conditions can lead to severe frost or stability problems. The orientation of the slope can considerably affect pavement surface temperature and its variation along the road. The geology of the surrounding terrain is another important factor affecting pavement environment. Soil interaction with pavement environment has been discussed throughout this chapter. The main characteristics acting on temperature, moisture, and stress regimes in pavements are permeability, stiffness, frost susceptibility, and moisture sensitivity of subgrade soils. The other factors of major concern for pavement engineers are homogeneity and uniformity of soils. Most of the problems found for pavements in cold climates are related to the lack of homogeneity or uniformity along the road corridor. Poor homogeneity within a soil deposit will lead to differential frost penetration and frost heaving. It also leads to uneven mechanical properties and behavior of the entire pavement structure. Lack of uniformity is a widespread characteristic of geological formations and deposits. Contacts between different soil deposits or between bedrock and soil deposits as well as soil and rock stratification are often the cause of localized differential mechanical, thermal, and hydric behavior. These interactions cannot be overemphasized. Proper characterization of soil properties and conditions prior to pavement construction or rehabilitation will significantly reduce the risk of poor performance of the pavement structure in any given environment. Pavement performance problems related to soil characteristics will be further discussed in Chap. 3. Soil investigation and characterization approach and techniques will be described in Chap. 4.

Review Questions

2m

2-1. For the conditions given below, determine the vertical effective stress at point A.

2m

γd = 14.5 kN/m3

3m

γd = 15.3 kN/m3 γsat = 17.7 kN/m3

2.5 m

52

γsat = 16.9 kN/m3 A

Pavement Environment 2-2. A load of 1100 kPa is applied uniformly on a circular plate with a diameter of 3.5 m. Find the vertical and horizontal stress at a depth of 3 m in a homogenous soil for which Poisson’s coefficient is 0.35. Use Boussinesq’s stress distribution theory.

2-3. Based on experimental results, a dynamic load of 20 kN circulating on an uneven flexible pavement causes failure when it is applied 30,400 times. If the load is reduced to 6 kN, how many applications will cause a failure?

2-4. A saturated subgrade soil has a resilient modulus of 600 MPa. A matric suction of 17 kPa is measured in the same soil at an unknown unsaturated state. What is the resilient modulus in these conditions?

References AASHTO Provisional Standards (2003). June 2003 Edition, American Association of State Highway and Transportation Officials. Barksdale, R. D., Alba, J., Khosla, N. P., Kim, R., Lambe, P. C., and Rahman, M. S. (1998). “Laboratory Determination of Resilient Modulus for Flexible Pavement Design,” NCHRP Web Doc 14, TRB. Bishop, A. W. (1959). “The Principle of Effective Stress,” lecture delivered in Oslo, Norway, in 1955, published in Technisk Ukeblad, vol. 106, no. 39, pp. 859–863. . Blazejowski, K., Nilsson, R., Hopman, P., and Sybilski, D. (1996). “Visco-Elastic Analysis of Typical Polish Flexible Pavements Using VEROAD,” Proceedings of the Second International Conference for Durable and Safe Road Pavements, Kielce, Poland. Bonnaure, F., Gest, G., Gravois, A., and Uge, P. (1977). “A New Method of Predicting Stiffness of Asphalt Paving Mixtures,” Proceedings of the Association of Asphalt Pavement Technologists, White Bear Lake, Minn., vol. 46. Brown, S. F. (1993). Structural Analysis of Pavements, Ehrola and Turumen (eds.), University of Oulu, Publications of road and transport laboratory, 22 Oulu, Finland, pp. 259–273. Brown S. F. (1997). “Achievements and Challenges in Asphalt Pavement Engineering,” Proceedings of the Eighth International Conference on Asphalt Pavements, International Society of Asphalt Pavements, White Bear Lake, Minn. Burminster, D. M. (1943a). “The Theory of Stress and Displacements in Layered Systems and Applications to the Design of Airport Runways,” Proceedings of the 23rd Annual Meeting of Transportation Research Board of the National Academies, Washington, D.C., vol. 23, pp. 126–144. Burminster, D. M. (1943b). “The General Theory of Stresses and Displacements in Layered Soils Systems,” Journal of Applied Physics, vol. 16, no. 2, pp. 89–96. Burminster, D. M. (1958). “Evaluation of Pavement Systems of the WASHO Road Test by Layered Systems Method,” Bulletin 177, Transportation Research Board of the National Academies, Washington, D.C., pp. 26–54. Christensen, D. W., Pellinen, T., and Bonaquist, R. (2003). “Hirsch Model for Estimating the Modulus of Asphalt Concrete,” Journal of the Association of the Asphalt Paving Technologists, vol. 72, pp. 97–121. Coulombe, C. (2002). “Effet de la vitesse des véhicules sur les paramètres de conception des chaussées souples en milieu municipal (Effect of vehicle speed on flexible pavement design parameters in municipal context),” Essai de Maîtrise, département de génie civil, Université Laval. (in French)

53

54

Chapter Two Das, B. M. (2002). Principles of Geotechnical Engineering, Brooks/Cole, Pacific Grove, Calif. Doré, G. (2002). “Cold Region Pavements,” Journal of Glaciology and Geocryology, vol. 24, no. 5, pp. 593–600. Doré, G. (2004). “Development and Validation of the Thaw Weakening Index,” International Journal of Pavement Engineering, vol. 5, no. 4, pp. 185–192. Doré, G., Konrad, J. -M., and Roy, M. (1999). “Deterioration Model for Pavements in Frost Conditions,” Transportation Research Record 1655, Transportation Research Board of the National Academies, Washington, D.C., pp. 110–117. Doré, G., Pierre, P., Bilodeau, J. P., and Abdelrazik Idriss, A. (2004). “Développement d’un essai simple et rapide pour l’estimation du potentiel de ségrégation (Development of a simple and rapid test for the estimation of the segregation potential),” Rapport GCT-2004-12. (in French) Doucet, F., and Doré, G. (2004). “Module des matériaux granulaires c-ltpp (Resilient modulus of C-LTPP granular materials),” Proceedings of the 57th Canadian Geotechnical Conference, Canadian Geotechnical Society, Quebec City, October 24–27. Dysli, M. (1991). Le gel et son action sur les sols et les fondations. Presses Polytechniques et Universitaires Romandes, Lausanne. Dysli, M., Lunardi, V., and Stenberg, L. (1997). “Related Effects on Frost Action: Freezing and Solar Radiation Indices,” Ground Freezing 1997— Frost Action in Soils, Knutsson (ed.), Proceedings of an International Symposium, 15–17 April 1997, A. A. Balkema, Rotterdam. Fredlund, D. G., and Rahardjo, H. (1993). Soil Mechanics for Unsaturated Soils, Wiley InterSciences, New York, New York. Garber, N.J., and Hoel, L. A. (1997). Traffic and Highway Engineering, PWS Publishing Company, Boston, Mass. Hazen, A. (1930). “Water Supply,” American Civil Engineering Handbook, Wiley, New York. Henry, K. (2000). “A Review of the Thermodynamics of Frost Heave,” ERDC/CRREL report TR-00-16, U.S. Army Corps of Engineers. Heukelom, W., and Klomp, A. J. G. (1964). “Road Design and Dynamic Loading,” Proceedings of the Association of Asphalt Pavement Technologists, White Bear Lake, Minn., vol. 33. Huang, Y. H. (2004). Pavement Analysis and Design, 2d ed., Prentice-Hall, Upper Saddle River, N.J. Imbs, C., and Doré, G. (2003). “Méthode d’évaluation des effets du dégel et de son évolution sur les différentes couches d’une structure routière (A method for the evaluation of the effect of thawing and it’s evolution on pavement layers),” Research report GCT03-01, Laval University, Civil Engineering Department, Quebec City, Canada. Janoo, V., and Greatorex, A. (2002). “Performance of Montana Highway Pavements during Spring Thaw,” Report FHWA/MT-02-006/8155, Federal Highway Administration and Montana Department of Transportation. Johnson, F. L., and Chang, F. F. M. (1984). “Drainage of Highway Pavements,” Federal Highway Administration, Publication No. FHWA-TS-84-202. Washington, D.C. Jones, A. (1962). “Tables of Stresses in Three-Layer Elastic Systems,” Bulletin 342, Transportation Research Board of the National Academies, Washington, D.C., pp. 76–214. Jones, G. M., Darter, M. I., and Littlefield, G. (1968). “Thermal Expansion-Contraction of Asphaltic Concrete,” Proceedings of Association of Asphalt Pavement Technologists, White Bear Lake, Minn., vol. 37.

Pavement Environment Jung, D., and Vinson, T. (1994). “Thermal Stress Restrained Specimen Test to Evaluate Low-Temperature Cracking of Asphalt-Aggregate Mixtures,” Transportation Research Record 1417, Transportation Research Board of the National Academies, Washington, D.C. Konrad, J. -M., and Morgenstern, N. R. (1980). “A Mechanistic Theory of Ice Formation in Fined-Grained Soils,” Canadian Geotechnical Journal, no. 17, pp. 473–486. Konrad, J. -M., and Morgenstern, N. R. (1983). “Frost Susceptibility of Soils in Terms of Their Segregation Potential,” Proceedings Permafrost: Fourth International Conference, National Academy Press, Washington, D.C., pp. 660–665. Ladanyi, B., and Shen, M. (1989). “Mechanics of Freezing and Thawing in Soils,” Proceedings of FROST ’89, Technical Research Center of Finland, Espoo, Finland. Lekarp, F., and Dawson, A. (1998). “Modeling Permanent Deformation Behaviour of Unbound Granular Materials,” Construction and Building Materials, Elsevier, vol. 12, no. 1, pp. 9–18. Loch, J. P. G., and Kay, B. D. (1978). Water Redistribution in Partially Frozen, Saturated Silt under Several Temperature Gradients and Overburden Loads, J. Soil Sci. Soc. of Amer., 43, 3, pp. 400–406. McAdams, W. C. (1954). Heat Transmission 3d ed., McGraw-Hill, New York. McLeod, N. W. (1976). “Asphalt Cements: Pen-Vis Number and Its Applications of Moduli of Stiffness,” ASTM Journal of Testing and Evaluation, vol. 4, no. 4. Miller, R. D. (1972). “Freezing and Heaving of Saturated and Unsaturated Soils,” Highway Research Records, No. 393, Transportation Research Board of the National Academies, Washington, D.C., pp. 1–11. Moulton, L. K. (1980). “Highway Subsurface Drainage,” Federal Highway Administration, Publication No. FHWA-TS-80-224, Washington, D.C. OECD (1988). “Heavy Trucks, Climate and Pavement Damage,” Organisation for Economic Co-operation and Development, Paris. Oliver, J. E. (1973). Climate and Man’s Environment: An Introduction to Applied Climatology, John Wiley & Sons, Hoboken, N.J. Palolahti, A., Slunga, E., Saarelainen, S., and Orama, R. (1993). “Sulavan Maan Kantavuus (Elastic Stiffness of Thawing Soils),” Helsinki University of Technology, Faculty of Civil Engineering and Surveying, p. 99. Pavlov, A. V. (1976). “Heat Transfer of the Soil and the Atmosphere at Northern and Temperate Latitudes,” CRREL Draft Translation 511, U.S. Army Corps of Engineers. Peattie, K. R. (1962). “A Fundamental Approach to the Design of Flexible Pavements,” Proceedings of International Conference on the Structural Design of Asphalt Pavements, University of Michigan, Ann Arbor, Mich., pp. 403–411. Perera, Y. Y., Zapata, C. E., Houston, W. N., and Houston, S. L. (2004). “Moisture Equilibria beneath Highway Pavements,” Transportation Research Board 2004 Annual Meeting CD-ROM. Rowe, G. M., Brown, S. F., Sharrock, M. J., and Bouldin, M. G., 1995, “Visco-elastic analysis of hot mix asphalt pavement structures,” Transp. Res. Record No. 1482, Transp. Res. Board, Washington, D.C., pp. 44–51. Steven, B. D., de Pont, J. J., Pidwerbesky, B. D., and Arnold, G. (1999). “Accelerated Dynamic Loading of Flexible Pavements at CAPTIF,” Proceedings of International Conference on Accelerated Pavement Testing, Paper GS2-3, Reno, Nevada. Ullidtz, P. (1987). Pavement Analysis, Elsevier Science, New York.

55

56

Chapter Two Ullidtz, P. (2002). “Analytical Tools for Design of Flexible Pavements,” Keynote address at the Ninth International Conference on Asphalt Pavements, International Association of Asphalt Pavements, White Bear Lake, Minn., http://www.asphalt.org/ (July 2, 2007). Van der Poel, C. (1954). “A General System Describing the Viscoelastic Properties of Bitumens and Its Relation to Routine Test Data,” Journal of Applied Chemistry, May 1954. Zarling, J. P., and Braley, W. A. (1988). “Embankment Design and Construction in Cold Regions,” section 3, Geotechnical Thermal Analysis, E. G. Johnson, A. Phukan, and W. H. Haas (eds.), ASCE, Reston, Va. Zeng, H. Y., and Vinson, T. S. (1998). “Thermal Contraction of an Asphalt Concrete Mixture,” Proceedings of the Ninth International Conference on Cold Regions Engineering, ASCE, Reston, Va. Zubeck, H. K., and Vinson, T. S. (1996). “Prediction of Low Temperature Cracking of Asphalt Concrete Mixtures with Thermal Stress Restrained Specimen Test Results,” TRR No. 1545, Transportation Research Board, National Research Council, Washington, D.C. Zubeck, H. K., and Vinson, T. S. (2007). “Prediction of HMA Low Temperature Crack Spacing Using TSRST Results,” Proceedings of the Eighth International Symposium on Cold Regions Development, Finnish Association of Civil Engineers, Helsinki, Finland.

CHAPTER

3

Cold Region Pavement Performance

P

avements are built to provide a safe and comfortable ride for road users. Fulfilling this vital role, referred to as the functional role of the pavement, involves that the surface must be smooth and skid resistant. The ability of a pavement to play its functional role is often referred to as the serviceability. Distortions of the pavement surface reduce pavement smoothness. These distortions can be caused by differential movements, crack deterioration, and raveling of the pavement surface. The loss of skid resistance is generally the result of the wear of the surface texture or of the presence of distortions, which can affect vehicle dynamics and cause water accumulation at the pavement surface. Pavements also need to be cost-effective assets, which implies that they must perform or, in other words, provide an adequate level of service over a reasonable period of time. Performance is directly linked to the structural adequacy of the pavement structure. Traffic action will cause fatigue cracking and permanent deformation on weak pavements. Other factors related to climate will intensify damage caused by traffic or cause other damages specific to mechanisms triggered by climatic factors. The mechanisms involved in pavement deterioration in cold climates needs to be fully understood in order to apply good engineering principles when selecting pavement materials, designing pavement structures or analyzing an existing structure needing maintenance or rehabilitation. Cold region pavements are subjected to intense loading by climatic and environmental factors which are, in good part, responsible for the seasonal and the longterm loss of structural and functional capacity of the pavement. These factors also intensify the damaging effect of heavy loads acting on the pavement structure. The main deterioration mechanisms for pavements in cold climates can be grouped in those that are acting in the asphalt-bound materials and those that are acting in the unbound layers and subgrade soils. The factors acting on the asphalt-bound materials include: • Thermal contraction or fracture (Sec. 3-1) • Fatigue (Sec. 3-2) • Crack deterioration (Sec. 3-3) • Rutting due to lack of stability or wear (Sec. 3-4) • Aging (Sec. 3-5)

57

58

Chapter Three • Pavement disintegration caused by the action of water, salt, and frost within the asphalt-bound layer (Sec. 3-6) • Potholes (Sec. 3-7) The factors acting on unbound layers and subgrade soil are • Volume change and more specifically differential volume change caused by frost heave (Sec. 3-8) • Bearing capacity loss during spring thaw (Sec. 3-9) • Frost destructuration of sensitive clays (Sec. 3-10) • Thaw consolidation of frozen soils in permafrost regions (this problem is described in Chap. 10) These deterioration modes are described in the following sections. A first subsection, labeled “problem description,” includes general descriptions of the problems including, when available, a mechanistic explanation in order to facilitate identification and quantification of the action of contributing factors. A second subsection describes available techniques to assess the problem when applicable. Finally, a third subsection includes a brief discussion on mitigation techniques applicable in cold environment contexts.

3-1 Thermal Cracking of Asphalt Concrete Problem Description Thermal cracking of asphalt pavements displays itself as fairly straight cracks perpendicular to the direction of the road (Fig. 3-1). In some cases the cracking progresses with time to the extent that the crack spacing becomes smaller than the width of the road.

FIGURE 3-1

Thermal transverse cracking.

Cold Region Pavement Performance

FIGURE 3-2

Thermal block cracking.

Cracks then start to form parallel to the direction of the road and form blocks with the transverse cracks as shown in Fig. 3-2. Low-temperature cracking are generally initiated in the asphalt bound layer, but can also be initiated in the underlying frozen pavement layers or subgrade possessing tensile strength due to the binding effect of pore ice. Figure 3-3 shows a crack initiated below the asphalt concrete layer that expands over the sidewalk, highway, bicycle path, and in-between green area.

FIGURE 3-3

Thermal cracking initiated below the HMA layer.

59

60

Chapter Three The cracks decrease the riding quality of the pavement and allow water and deicing agents to penetrate the pavement structure causing frost-related problems, which shortens the pavement service life. Annual crack sealing is necessary as the cracks reopen during winters due to pavement contraction at the vicinity of the crack. It is difficult to prevent the cracks from reflecting through a new overlay. Thermal cracking is often divided into low-temperature cracking and thermal fatigue cracking. Low-temperature cracking occurs when temperature drops fast below −16 to −35°C and the thermal stress in the pavement exceeds its tensile strength. The development of thermal stresses is described in Sec. 2-3. Thermal cracking is also observed in climates where these kinds of cold temperatures never occur. In these cases, the thermal cracking is typically assumed to be thermal fatigue cracking, which occurs due to the diurnal temperature cycling. This section describes low-temperature cracking that is the most typical form of thermal cracking in cold regions.

Problem Assessment Cooling temperature causes thermal stress to develop in hot-mix asphalt (HMA), but only if the pavement slab is constrained from contracting. The two conditions that have to be satisfied for low-temperature cracking to occur for a given HMA could be termed “cold” and “constraint.” If one of these does not exist, there is no low-temperature cracking. The thermal stress given in Eq. (2-36) with the associated cracking temperature (see Fig. 2-26) could be considered representing the “cold.” It is the on/off switch of the initial low-temperature cracking. If the temperature change in the field is not sufficient in magnitude and rate to cause the thermal stress to reach the tensile strength of the pavement (assuming adequate restraint) no cracks occur. This means that the pavement properties affecting its cracking temperature are the most important factors influencing low-temperature cracking for a given climate. Once cracking occurs, i.e. the two conditions are satisfied, the “constraint” controls the crack spacing. The spacing is mainly affected by the pavement’s tensile strength (temperature and loading time dependent), and the shear strength at the pavementbase interface (Zubeck and Vinson 2007). The best estimate for the tensile strength at the cracking temperature is obtained from the thermal stress restrained specimen test (TSRST, see Chap. 4). The cooling rate in the TSRST should equal the maximum cooling rate in the field. A standardized test method to determine the constraint conditions in a routine mix design is yet to be developed. Factors influencing lowtemperature cracking of asphalt pavements are listed by Vinson et al. (1996). These factors and their effect in cracking through either “cold” or “constraint” are summarized in Table 3-1. Binder properties are often used to predict the low-temperature cracking resistance of a pavement mixture, as they dictate the cracking temperature. The property that is crucial for low-temperature cracking is binder’s stiffness at low temperatures. Soft binders have better cracking resistance than hard binders. Penetration at a low temperature, from 4 to 15°C (ASTM D5), Fraass Breaking Point (DIN 52012), bending beam rheometer test and direct tension tester (AASHTO T 313 and T 314) can be used to assess the binder consistency at low temperatures. The degree of aging of the binder also affects the cracking temperature as the binder becomes stiffer with time. Different binders age at different rates, and therefore, an aging rate needs to be considered. Aging of binder during the hot mix production and in service can be simulated in the laboratory using the thin film oven test, TFOT, (ASTM D 1754), rolling thin film oven test, RTFOT, (AASHTO T 240), or pressure aging vessel, PAV, (AASHTO R-28).

Cold Region Pavement Performance

Factor

Effects through

Change in Factor

Effect of Change on Cracking

Material factors

Asphalt cement

Cold

Consistency increases

Increases

Aggregate type

Constraint

Durability decreases, absorptiveness increases

Increases

Environmental factors

Temperature

Cold

Minimum air temperature decreases

Increases

Rate of cooling

Cold

Increases

Increases

Pavement age

Cold/ constraint

Increases

Increases

HMA thickness

Cold/ constraint

Increases

Decreases

Shear strength between the HMA layer and base course

Constraint

Increases

Increases

Subgrade type

Cold/ constraint

Indirect effect through associated microclimates and ground thermal cracking

Increases/ decreases

Cold/ constraint

Increases

Increases

Pavement structure geometry

Construction flaws

TABLE 3-1

Factors Affecting Low-Temperature Cracking

Two approaches have been proposed to assess or predict low-temperature cracking. The first one is based on the cracking temperature alone; the lower the cracking temperature is, the better the cracking resistance. The methods based on cracking temperature evaluate the cracking tendency of a mixture without assessing crack spacing or cracking frequency. Hills and Brien (1966), reported by Vinson et al. 1996, estimate the cracking temperature by first calculating the thermal stress using Eq. (2-36) with measured values of thermal contraction and stiffness. The cracking temperature is obtained at the intersection of the thermal stress curve and the curve for measured tensile strength of the mixture as illustrated in Fig. 3-4. The AASHTO PP42-02 (AASHTO 2003) uses the same principle. Bending beam rheometer test results are used to calculate a relaxation modulus master curve and subsequently the thermal stress. The calculated thermal stress is then compared to the failure stress from the direct tension test to determine the critical cracking temperature. Simulation methods, such as TSRST, can be used to directly determine the cracking temperature of an asphalt mixture. The second approach predicts the crack spacing either by statistically derived predictive models (Hajek 1971; Ehrola 1986; Haas et al. 1987) or by a mechanistic approach (Zubeck and Vinson 1996; Timm and Voller 2003; Konrad and Shen 1997).

61

62

Chapter Three

FIGURE 3-4 Estimating the fracture temperature of asphalt concrete (after Hills and Brien 1966, reported by Vinson et al. 1996).

Hajek (1971) developed a statistical model for a cracking index based on 42 observations from pavements in Ontario and Manitoba, Canada. The cracking index is the number of cracks per 152 m section of two-lane roadway. The model in its current version (Huang 2004) is given in Eq. (3-1): I = 30.3974 + (6.7966 – 0.8741 · h + 1.3388 · a) log (0.1Sbit) – 2.1516 · d – 1.2496 · m + 0.06026 · Sbit log (d)

(3-1)

where I = cracking index, Sbit = stiffness modulus of the original asphalt cement in kg/cm2 as determined by McLeod’s method for loading time of 20,000 s and for winter design temperature, a = age of pavement in years, m = winter design temperature in −°C (use only positive values), d = dimensionless subgrade code 5-sand, 3-loam, and 2-clay, h = combined thickness of bituminous layers in inches (see Example 3-1). Example 3-1 Predict the crack spacing after 2 and 10 years assuming that all transverse cracks are full cracks (extending across the entire two-lane pavement) for a 51-mm-thick pavement with the following design parameters: stiffness of the original asphalt cement 310 kg/cm2, winter design temperature −25°C, sand subgrade.

Cold Region Pavement Performance Solution With Sbit = 310 kg/cm2, a = 2 years, m = 25°C, d = 5 and h = 2 in, from Eq. (3-1) I = 30.3974 + (6.7966 – 0.8741 · 2 + 1.3388 · 2) log(31.0) – 2.1516 · 5 – 1.2496 · 25 + 0.06026 · 310 · log(5) = 10.48 Spacing = 152.4 m/10.48 = 14.54 m. Answer: 15 m. Similarly for a = 10 years, Eq. (3-1) yields I = 20.45, and spacing = 5.76 m. Answer: 6 m.

In order to study the crack spacing with a mechanistic model, Zubeck and Vinson (1996, 2007) suggest performing an equilibrium analysis that considers the forces affecting an asphalt concrete slab (see Fig. 3-5a). For the analysis, stresses given in Eqs. (2-36) and (2-37) are transferred into forces. A thermal force for a pavement slab with width, W, and thickness, D, becomes Fthermal = sthermal WD

(3-2)

The opposing force when the shear strength is mobilized can be calculated as Fresisting = (sv tanf + c)Wx

(3-3)

FIGURE 3-5 (a) Forces affecting pavement slab, (b) stresses in pavement slab before and after cracking (after Zubeck and Vinson 2007).

63

64

Chapter Three where sv = g D, g is the unit weight of the HMA, f = friction angle and c is the cohesion at the interface (see Fig. 3-5a), W = width of the slab and x is the distance from the slab edge toward to the center of the slab (≤0 x ≤ ½ slab length). The thermal force that is associated with the cracking of the pavement can be determined as Ffailure = (sthermal) failure WD = Pts(T, t)WD

(3-4)

where (sthermal)failure is the thermal stress that equals the temperature and loading time dependent pavement tensile strength Pts(T, t) obtained from the TSRST (the cracking strength at the cracking temperature). For cracking to occur Fthermal = Fresisting = Ffailure

(3-5)

Then, using Eqs. (3-3) and (3-4) Pts(T, t) WD = (g D tan f + c)Wx

(3-6)

The crack spacing can be evaluated with the scenarios illustrated in Fig. 3-5b. The temperature drops and the thermal stress together with the opposing stress develops. At the vicinity of the slab edges, where the maximum possible opposing stress is small, the slab is able to contract and consequently the thermal stress will be released. When the temperature gets cold enough for the thermal stress to equal the tensile strength of the pavement, cracking will occur within the fully restrained area (see Fig. 3-5b). The minimum possible crack spacing occurs, when the initial crack forms at the point closest to the slab edge where the thermal stress equals the strength of the pavement. The minimum spacing can be solved from Ffailure = Pts(T, t)WD = (g D tanf + c)W Spacingmin Spacing min =

Pts(T, t) ⋅ D γ D tan φ + c

(3-7) (3-8)

The maximum possible spacing occurs, when the initial crack occurs at a location where the residual thermal stress after cracking equals the strength of the pavement. The maximum spacing is Spacingmax = 2 Spacingmin

(3-9)

The crack spacing of the pavement will stay between these two values depending on the location of the initial cracks. Additional cracking will not occur, as the entire new slab is now able to contract at cold enough temperatures, and the thermal stress will not reach the pavement’s tensile strength. However, as the pavement ages, it becomes more brittle. Consequently, the tensile strength with temperature is affected leading to the possibility of future cracking (Zubeck and Vinson 2007). The determination of crack spacing using the TSRST results is demonstrated in Example 3-2. Example 3-2 Laboratory tests were conducted for an asphalt-aggregate mixture. The following results were obtained: Unit weight was 24 kN/m3. The TSRST gave fracture strength of 4.6 MPa at a fracture temperature of −28°C. A direct shear test was conducted for a 50-mm-thick asphalt concrete slab on 20-mm maximum aggregate size dense graded base course to be used on the paving project. The test results indicated no cohesion and a friction coefficient (tan f) of 2.4 at the asphalt concrete and base

Cold Region Pavement Performance course interface. What is the anticipated range of cracking for a 50-mm pavement layer if the predicted coldest pavement temperature is (a) −32°C and (b) −25°C? Solution (a) Since the TSRST fracture temperature of −28°C is warmer than the predicted pavement temperature of −32°C the pavement will crack. The minimum crack spacing from Eq. (3-8) is Spacingmin = Pts(T, t)/(g ⋅ tanf) Spacingmin = 4600 kPa/(24 kN/m3 ⋅ 2.4) = 80 m The maximum spacing from Eq. (3-9) is Spacingmax = 2 Spacingmin = 2 ⋅ 80 m = 160 m The cracking will occur at intervals from 80 to 160 m. The average crack spacing will approach 80 m as the pavement ages. (b) Since the TSRST fracture temperature of −28°C is colder than the predicted pavement temperature of −25°C, the thermal stress does not reach the tensile strength of the pavement. Consequently no cracking will occur until the pavement has aged significantly.

Remedial Solutions The best remedy to avoid low-temperature cracking is the use of mixtures with low TSRST fracture temperatures and high fracture strengths. Use of soft asphalt cements may lead to rutting in the summer time due to permanent deformation, especially on roads with a high volume of traffic. Therefore, binder selection is often a compromise that depends on peak temperatures (during winter and summer) and the traffic volume. Polymer-modified asphalts can be developed to have both good rutting and low-temperature cracking resistance and are recommended for high-traffic roads. For low-traffic roads, soft asphalt cement is recommended; its grade should be selected based on the climate zone (see Chap. 7). Lowtemperature cracking can also be reduced by increasing the thickness of the HMA layer.

3-2

Fatigue Cracking Problem Description Fatigue cracking is often called alligator cracking because its closely spaced pattern is similar to the pattern of an alligator’s skin (Fig. 3-6). It is a fracture phenomenon caused by a repeated application of tensile strains that are less than the strength of the material. In a fatigue process, microscopic flaws in a material grow in size under repeated loading, becoming more densely concentrated until visible flaws or cracks develop. The visible cracks then propagate through the material. Fatigue cracking is made worse by inadequate pavement drainage. The HMA layers experience high strains when the underlying layers are weakened by excess moisture and consequently fail prematurely in fatigue (Roberts et al. 1996). This is an important factor in cold regions, where pavements become saturated regularly during the spring thaw. Doré and Savard (1998) report that most of the fatigue cracking in Quebec, Canada, occurs during spring. Not only are the deflections larger, but also the HMA layer is still cold and consequently more brittle. In thin pavements, cracking starts at the bottom of the asphalt layers and propagates upward. In thick pavements, bending of pavement layers is reduced eventually to the level that crack initiation is restrained and no bottom up fatigue cracking occurs.

65

66

Chapter Three

FIGURE 3-6

Fatigue cracking.

In recent years a term, perpetual pavements, has been introduced to describe these pavements (Newcomb et al. 2001). However, thick pavements may suffer top-down cracking, that is, cracking which starts from surface layer and propagates downward. At what HMA thickness and traffic levels does the bottom-up fatigue cracking transform to top-down cracking in cold regions is yet to be determined. Fatigue cracking is one of the common forms of pavement distress and is incorporated in the pavement design process. The following sections describe the assessment and remedial solutions for bottom-up cracking that is the most prominent fatigue cracking type in cold regions.

Problem Assessment The fatigue characteristics of asphalt mixture are usually expressed as relationships between the initial stress or strain and the number of load applications to failure. They are determined using repeated flexure (beam fatigue test), direct tension or diametral tests, performed at several stress or strain levels (Tayebali et al. 1992). The fatigue behavior of a specific mixture can be characterized by the slope and relative level of the stress versus the number of load repetitions to failure and can be defined by a relationship given in Eq. (3-10) (Huang 2004; Monismith et al. 1985): f2

 1  1 N f = f1      ε t   E1 

f3

(3-10)

where Nf is the allowable number of load repetitions to prevent fatigue cracking, et is the tensile strain at the bottom of the asphalt layer, E1 is the elastic modulus of asphalt layer, f1, f2, and f3 are constants determined from laboratory fatigue tests with f1 modified

Cold Region Pavement Performance to correlate with field performance observations. Several values are suggested for f1, f2, and f3. An example of the fatigue failure criterion [Eq. (3-10)] is the method given by the Asphalt Institute [Eq. (3-11)] (Huang 2004). This stress-controlled model was developed using the beam fatigue test for a standard mixture having 5 percent air-void content and 11 percent effective binder volume. N f = A ⋅ 0.00432 ⋅ C ⋅ 796ε t−3.291 E∗

−0.854

(3-11)

where A = 18.4 (factor that accounts for differences between laboratory and field conditions), C is a volumetric correction factor [see Eqs. (3-12) and (3-13)], et is the initial tensile strain and |E∗|is the dynamic modulus. C = 10 M

(3-12)

  Vbeff M = 4.84  − 0.69    Va + Vbeff

(3-13)

where Vbeff is effective binder volume, %, and Va is air-void content, % (see Example 3-3). Example 3-3 Fatigue tests were performed using 400 × 100 × 50 mm3 beams resting on a flexible polymer base by two-point loading at 15°C. The test results are given in Table 3-2. Develop an equation relating the number of repetitions to failure similar to Eq. (3-11). Plot the measured and predicted Nf values versus the measured strain. Use log-log scale on axis.

Strain, µm/m

N

Force, kN

E, MPa

1030

1933

1.5

3000

785

2846

1.1

3100

720

11783

1.2

3500

700

5776

1.3

3900

680

16400

1

3200

620

28491

0.9

3100

610

28863

0.8

2800

520

71244

0.85

3800

340

443660

0.75

5000

270

1923705

0.65

5000

1130

307

1.7

3200

Source: Spoof 1992.

TABLE 3-2

Fatigue Test Results for Example 3-3

67

68

Chapter Three

FIGURE 3-7 The relationship between strain and number of repetitions to failure for Example 3-3.

Solution Write out Eq. (3-10) as Eq. (3-14) and run a regression analysis in which Y = log(Nf), X1 = log(et), and X2 = log(E). log( N f ) = log( f1 ) + f 2 log(ε t ) + f 3 log(E)

(3-14)

The results of a linear regression analysis by using the “least squares” method gives the following values: the intercept log(f1) = −8.926, the least square estimators f2 = −6.525 for log(et) and f3 = −2.173 for log(E). Solving for f1 = 10−8.926 gives a value of 1.185 × 10−9. The equation then becomes N f = 1.185 × 10−9 εt−6.525 (E)−2.173 where the unit for strain, et, is mm/m and the unit for modulus, E, is MPa. The relationship between strain and number of repetitions to failure is given in Fig. 3-7.

Fatigue test can be conducted using constant stress or strain. In the constant stress test, the strain is increased with the number of repetitions in order to keep the stress constant. In the constant strain test, the stress or load is decreased with the number of repetitions to keep the strain constant. There is not a complete agreement as to whether real pavement systems should be designed using stress- or strain-controlled fatigue relationships. It is usually assumed that the constant stress test is applicable to thick pavements, where the HMA layer is more than 150 mm thick and is the main loadcarrying component. For thin pavements (3 percent). Asphalt content also has a significant effect on fatigue response. It should be as high as possible without risking rutting resistance of the mixture. Stiff asphalt cements and dense graded mixtures should be used for thick pavements and soft asphalt cements and more open-graded mixtures for thin pavements. The temperature affects the fatigue life through mixture stiffness that decreases with increasing temperature. As a consequence, the fatigue life of thick pavements decreases and the fatigue life of thin pavements increases with increasing temperature (Rao Tangella et al. 1990).

Cold Region Pavement Performance

Remedial Solutions Because of the aforementioned stress/strain conditions for thick and thin pavements, it is better to have stiff HMA layer for thick pavements, and soft HMA layer for thin pavements to minimize fatigue cracking. The mix designer needs to know the thickness of the HMA layer in order to optimize the composition of the mixture to withstand rutting, thermal and fatigue cracking (Pellinen 2001). Drainage of the roadway becomes imperative as moisture reduces the strength of the pavement layers and consequently increases the risk for cracking. Proper quality control during the construction, especially for pavement thickness, is also important in reducing the cracking tendency (Roberts et al. 1996).

3-3

Crack Deterioration Once cracks are initiated by thermal contraction or traffic action, pavement deterioration is accelerated. The reduction of the layer stiffness in the vicinity of the crack combined with weakened base material caused by water infiltration amplifies pavement damage caused by truck traffic. As a result, secondary cracks are initiated and the main crack tends to become faulted and depressed.

Problem Description As illustrated in Fig. 3-8a, an uncracked asphalt concrete layer is very effective in distributing the load to the underlying base layer. Once a crack appears in the layer (Fig. 3-8b), the stress distribution pattern is strongly affected by the ineffective load transfer between the two faces of the crack. In the presence of a load near the crack and in the complete absence of friction between the two faces of the crack the load distribution effectiveness of the asphalt layer is reduced by a factor of almost two. The problem is worsened by the presence of water seeping into the granular base through the crack (see Sec. 2-2). The deflection of the surface is thus excessive in the vicinity of the crack leading, under the repetitive action of wheel loads, to the development of spalling of

FIGURE 3-8

The process of crack deterioration.

69

70

Chapter Three

FIGURE 3-9

Deteriorated cracks.

the crack faces, the formation of secondary cracks and to the accumulation of differential permanent deformation (Fig. 3-8c, d and Fig. 3-9). Crack deterioration contributes to increased pavement roughness and to reduced pavement structural capacity.

Remedial Solutions Two categories of actions can be taken to mitigate the effect of pavement cracking. In all cases, the action must be taken before significant crack deterioration occurs, involving the loss of structural capacity in the vicinity of the crack (Fig. 3-8c and d). The first type of action is a preventive strategy and it involves sealing the crack early after crack initiation. Crack sealing prevents the intrusion of water and incompressible materials into the crack; thus, reducing further deterioration of the crack. Crack openings should be wide enough to allow penetration of the sealant, but should also be small enough to allow good friction and load transfer between the two faces of the crack. The National Center for Asphalt Technology recommends openings between 6 and 12 mm (Roberts et al. 1996). Two approaches can be taken for crack sealing. The first involves routing the top part of the crack to create a reservoir to contain the sealant (Masson 2001). The second involves bridging the crack with sealant poured onto the pavement surface forming a 3- to 4-mm-thick and 40- to 50-mm-wide band (Pouliot 2003). Both techniques appear to give good results (Roberts et al. 1996; Pouliot 2003; Masson 2001). The second category of action is applicable when the crack has reached a certain level of deterioration without important structural damage to the pavement. It involves localized repair of the pavement in the vicinity of the crack. This is generally done by partial or complete removal of the asphalt concrete in the damaged area and filling of the gap with hot mix asphalt. If the repair is followed by an overlay, special attention needs to be given to prevent crack reflection in the new asphalt layer. Several approaches can be used to control reflective cracking. They include acting on the cause of crack reflection, using techniques or materials, which will reduce stress concentration above the existing crack. For example, stress absorbing membranes, bitumen-rich asphalt concretes, and granular layers have been used with variable success as stress absorption layers between existing and new asphalt concrete layers.

Cold Region Pavement Performance

3-4

Rutting of Asphalt Concrete Rutting manifests itself as depressions of the wheel paths as a result of traffic load. Except for intersections, it does not increase the longitudinal roughness of the road significantly, but may still become a safety hazard due to its effect on lateral maneuverability of vehicles and possible risk of hydroplaning on ponding water. Ruts decrease the structural capacity of the pavement due to decreased layer thickness and changed properties. When the rut depth exceeds a level where the serviceability of the road starts to decrease, the road needs rehabilitation. Rutting of roads in cold regions has several sources. The rutting may be limited in the asphalt layer, where it is caused either by permanent deformation or wear by studded tires. The rutting may also result from permanent deformation in the unbound structural layers or in the subgrade. This kind of rutting occurs mostly due to bearing capacity loss during the spring thaw. It is included in the pavement structural design and is described in Chap. 8. The following sections describe rutting in the bound HMA layer due to permanent deformation and wear by studded tires.

3-4-1

Permanent Deformation

Problem Description Permanent deformation is a result of initial densification and subsequent plastic deformation of the HMA with an increased number of load applications. The volume of the HMA decreases during densification due to reduced air voids in the mixture. After the air voids drop below a mixture specific limit (e.g., from 1 to 2 percent), plastic flow starts to occur. During plastic flow volume does not change anymore (assuming incompressible material); instead rutting forms when the mixture flows from the wheel paths to the small upheavals beside the wheel paths (Fig. 3-10). The relationship of

FIGURE 3-10

Plastic deformation at a bus stop.

71

72

Chapter Three

FIGURE 3-11 Division of permanent deformation into the initial densification and plastic deformation (Saarela et al. 1993).

the densification to total deformation for a laboratory tested sample is illustrated in Fig. 3-11 (Saarela et al. 1993). As the stiffness of the mixture decreases with increasing temperature and loading time, the permanent deformation occurs during the summer months especially in areas of slow or standing traffic, such as intersections and loading areas. Deformation can be visually differentiated from wear-related rutting by smooth dark asphalt concrete surface and by the presence of upheavals between wheel paths. Permanent deformation is a severe problem in almost all cold regions, as the pavement temperature often rises higher than the air temperature.

Problem Assessment The permanent deformation of HMA mixes is affected by material properties, mix design, and in-service conditions. The factors and their effects on rutting resistance summarized in Table 3-3 are by and large universally accepted tendencies. The only controversial effects are the aggregate gradation and maximum size. Gap-graded and stone matrix asphalt (SMA) mixtures are reported to have both higher (Sousa et al. 1991) and lower (Saarela 1993) plastic deformation rates than dense-graded mixtures. The mechanistic or mechanistic-empirical pavement design procedures limit rutting of asphalt pavement system to a tolerable level by controlling the maximum vertical strain or stress at the surface of the subgrade. These methods do not prevent rutting by permanent deformation in the HMA layer, and a separate analysis may be needed. Two analytical procedures have evolved to predict the amount of rutting in the HMA layer, namely, layer-strain predictive methodology and closed-form viscoelastic analysis. In addition, statistically derived models exist to predict the permanent strain (Sousa et al. 1991). The layer-strain method predicts rut depth using laboratory test results and linear or nonlinear elastic theory. Each layer of the pavement structure is divided into sublayers, i. The stress state is calculated at the center of each sublayer directly beneath the wheel load using elastic analysis. With the average stress state, the corresponding axial plastic strain can be determined from laboratory test results. The total rut depth for a

Cold Region Pavement Performance

Factor Aggregate

Change in Factor

Effect of Change in Factor on Rutting Tendency

Surface texture

Smooth to rough

Decreases

Gradation

Gap graded to continuous

Increases/ decreases

Shape

Rounded to angular

Decreases

Size

Increase in maximum size

Increases/ decreases

Binder

Stiffness*

Increases

Decreases

Mixture

Binder content

Test/field conditions

Increases

Increases

†

Air-void content

Increases

Increases

Voids in mineral aggregate (VMA)

Increases

Increases‡

Method of compaction

§

§

Temperature

Increases

Increases

State of stress/ strain

Increase in tire contact pressure

Increases

Load repetitions

Increase

Increases

Water

Dry to wet

Increase if mix is water sensitive

∗

Refers to stiffness at temperature at which rutting propensity is being determined. Modifiers may be utilized to increase stiffness at critical temperatures, thereby reducing rutting potential. † When air contents are less than about 3 percent, increase in air voids reduces the rutting potential. ‡ It is argued that very low VMAs (e.g., 1.1

>1.25

>1.5

High

>1.75

>2.00

>2.25

Source: MTQ, 2007, unpublished pavement engineering course material.

TABLE 4-15

∆IRI Values (m/km) Used by Quebec Ministry of Transportation as Indicator of Frost

Susceptibility

Advanced analysis of longitudinal profiles can lead to valuable information to support research and advance analysis of pavement condition. As described in more details in Sayers and Karamihas (1998), pavement profiles are signals that can be decomposed in sinusoids of different wavelengths. The signal can thus be filtered to highlight the effect of distortions of specific wavelengths. Public domain software is available for advanced analysis of pavement profiles. Among other, “ROADROUGH” (University of Michigan Transportation Research Institute) and PROVAL (Federal Highway Administration) offer several profile viewing, filtering, and analysis functions. Analysis of pavement profiles based on wavelength content can provide insight on causes and consequences of roughness deterioration. As a general rule, short wavelength distortions (3 m) are typically the results of problems occurring at greater depths and tend to affect mainly user comfort. The 10-m-long distortions observed on winter profile illustrated in Fig. 3-21 is a good example of the manifestation of a problem (frost action) occurring at depth between 1.5 and 2.0 m in the pavement system. When used to study the effects of frost action in pavements, advanced analysis of profile and roughness data can provide interesting insight on the development and the cause of winter roughness. Figure 4-33 illustrates the result of research work done on two sections of Highway 367 in Quebec, Canada (Fradette et al. 2005). In this study, profile filtering has been used to highlight the effects of two different frost heave mechanisms acting on pavements. The study is based on profile measurements made every second week during winter and every week during spring on several test sections. As a first step, the raw profile is used to compute IRI (nonfiltered). The raw profile is then filtered to remove wavelengths smaller than 3 m (smoothed) and longer than 3 m (antismoothed). IRI is then recalculated using smoothed and antismoothed profiles. Nonfiltered and filtered IRI values are plotted as a function of time in conjunction with frost and thaw penetration which were also monitored during the study. It can be observed on Fig. 4-33a that nonfiltered IRI follows closely the pattern of antismoothed IRI. Moreover, IRI increase occurs very early during winter, and, in a large proportion, before frost depth reaches the frost susceptible subgrade soil. Similarly, IRI decreases rapidly while thaw progresses through the granular pavement structure. These observations suggest that the pavement is strongly affected by crack heaving problems as described in Sec. 3-8-3. Crack heaving typically cause short wavelength distortions that can cause an important increase in roughness during winter as indicated by

I n v e s t i g a t i o n a n d Te s t i n g

FIGURE 4-33 Using advanced profile analysis techniques in studying the development of winter roughness on two test sections in Canada. [Fradette et al. 2005 (Figure 4, p. 140, and Figure 5, p. 141) with permission from the Transportation Research Board.]

the observed ∆IRI of more than 2 m/km. Figure 4-33b illustrates a totally different frost heave pattern. In this case, non-filtered IRI follows closely the pattern of smoothed IRI and is somewhat independent of the evolution of anti-smoothed IRI. The dominant effect of long wavelengths in roughness development suggest that differential frost heaving from frost-susceptible subgrade soil (as described in Sec. 3-7-1) is the dominant mechanism on that pavement section. This hypothesis is supported by the fact that winter roughness develops mostly at the end of winter, when frost penetration reaches the frost-susceptible soil (ML).

4-2-3

Pavement Instrumentation

Pavement instrumentation can be considered as the “ultimate” way to gather information on pavement condition and response to load and climate induced stresses. Pavement instrumentation involves installation of sensors at strategic positions in the pavement during pavement construction or through boreholes in existing pavements. It also involves intensive data collection, treatment, management, and analysis. The type of instruments that can be used to collect information in pavements can be divided in two categories: load response measuring sensors and climate response sensors. Figure 4-34 illustrates the type of sensors that can be used in pavements and Table 4-16 summarizes information relevant to installation and the use of these sensors.

163

164

Chapter Four

FIGURE 4-34 Load response sensors and climate response sensors commonly used in pavement engineering (see Table 4-16).

4-3

Soils and Material Testing Materials used in cold region pavement structures generally include different types of unbound granular materials, asphalt concrete, and asphalt-stabilized granular materials. Together with the characteristics of traffic and the properties of a road foundation, the properties of pavement materials and the characteristics of the pavement layers are key parameters for pavement design. Mechanistic-empirical pavement design and analysis methods use pavement material properties to compute strains transmitted to the pavement foundation and to assess pavement damage in terms of fatigue, rutting, and roughness. This section will describe laboratory testing techniques with emphasis on characterization techniques related to cold region performance of soils and pavement materials. The following subsections will be included: • Testing of asphalt concrete (thermal cracking, stripping, fatigue, rutting, and durability tests) • Testing of pavement materials (Stiffness, durability and stability) • Soil testing (Stiffness and stability) General description of the investigation/testing technique with proper references to detailed descriptions and standards are included, as well as useful information for the interpretation and engineering use of the properties, allowable values, and links with relevant design methods.

4-3-1 Testing of Bituminous Pavement Materials Testing of bituminous pavement materials is needed in order to avoid deterioration of the bound pavement layer due to rutting, cracking, or disintegration as described in Chap. 3. The tests aid in material selection, volumetric mix design, and verification of the performance of pavement mixtures. The use of the tests is further explained in Chap. 7, where mix design of asphalt-aggregate mixtures is discussed.

Type of Sensor, Principle, and Application 1.

Horizontal (tensile) strain gauges Measurement of tensile strain using variable electric resistance or fibre optic based technologies, at the bottom of asphalt bound layers

2.

Vertical (compressive) strain gauges Measurement of compressive strain in unbound pavement layers and/or subgrade soil. Typically uses technologies based on induction loops or displacement sensors

3.

Multidepth deflectometer Measurement of displacements and vertical strains at selected levels. Typically uses technologies based on displacement sensors. If design of system allows, can be used to monitor frost heave

4.

Pressure gauges Measurement of vertical stress at selected level in unbound layers or in subgrade soil. Typically uses pressure induced by a fluid in a flat cell

Installation

Data Collection

• Generally installed on top of the granular base prior to paving • Gauges installed prior to paving often experience excessive stress due to high temperatures and compaction operations • Some gauge models can be retrofitted in existing asphalt layers

• Data should be collected under a moving reference vehicle (known characteristics) • Data collected during short periods at high sampling rate (≥500 Hz) to capture peak strains. This involves the use of advanced signal conditioner and data acquisition technology • The speed of the vehicle and the position of the wheel on the sensor need to be carefully controlled and measured as they have an important effect on sensor output • For seasonal variation assessment, the data collection should be repeated several times with the same reference vehicle (see recommendation for seasonal FWD testing in Table 4-13)

• Can be installed after placement and compaction of the layer for a new construction • Can possibly be retrofitted through a borehole if installation depth allows adequate installation • Precautions should be taken to assure proper compaction around the gauge • Installed through a borehole in existing pavement structures • Adequate reconstitution of pavement structure and material density between anchor plates is critical • For systems anchored on the wall of the borehole, good contact needs to be establish and precautions need to be taken to avoid disintegration of the wall • Generally installed at the desired level prior to placement of the overlying layer during pavement construction

165

TABLE 4-16 Synthesis of Information on Sensors Commonly Used in Pavement Engineering (Continued)

166 Type of Sensor, Principle, and Application 5.

Frost tube Measurement of frost and thaw depth. Uses a methylene blue solution to track phase change through a change in color of the solution

6.

Resistivity probes Measurement of levels of phase change using resistivity variations between copper rings installed along a rod. Resistivity of soil and unbound materials changes drastically with phase change of pore water

7.

Thermistor strings Measurement of temperature using temperature sensitive resistor technology. Thermistor strings are composed of several thermistors mounted on low thermal conductivity rod to allow for the measurement of thermal regimes and phase change in pavements

Installation

Data Collection

• Installed through a borehole in existing pavement structures • Requires a protective cover for manual access to the tube • Should ideally be installed at the center of the road to capture maximum frost depth • The casing of the tube needs to be carefully anchored underneath the maximum expected frost depth to avoid frost jacking • The casing of the tube needs to be sealed to avoid freezing of water inside the casing

• Manual reading only • Frequent reading needs to be done to capture frost and thaw evolution. Reading semi weekly during winter and weekly during spring thaw is adequate for most applications • Requires adequate signing and often lane closure for safe reading of the tube

• Installed through a borehole in existing pavement structures • The probe needs to be carefully anchored underneath the maximum expected frost depth to avoid frost jacking

• Manual reading most of the time, but can also me automated • Reading can be done from a junction box located on the side of the road • Frequent reading needs to be done to capture frost and thaw evolution. Reading biweekly during winter and weekly during spring thaw is adequate for most applications • Readings can be influenced by infiltration of brine from deicing chemicals through pavement cracks • Can be read manually but ideal for data logging • Several readings a day (>6) are recommended to capture daily variations of surface temperatures • Using 0°C as phase change indicator might be misleading if pavement materials and soil are contaminated by deicing chemicals

8.

TDR antennas Assessment of volumetric moisture content through measurements of dielectric constant of unbound materials and soils based using “time domain reflectometry” technology

9.

Suction probes Measurement of negative pore pressure (suction) in unsaturated pavement materials and soils. Can be related to moisture content through soil moisture retention characteristic curve

10.

Piezometer Measurement of water table level and of positive pore water pressure

TABLE 4-16 (Continued)

• Installed during construction of a new pavement structure or retrofitted in an existing pavement through a borehole • When installed in a borehole, the main difficulty is to properly assess dry density of the material surrounding the antenna. Errors in dry density estimation can lead to important errors in moisture content readings

• Manual or automated reading. • Measurements on a daily basis during spring thaw and on a weekly basis during the rest of the year are recommended • Contamination by deicing chemical can affect Tonnage distribution roster (TDR) readings

• Installed during construction of a new pavement structure or retrofitted in an existing pavement through a borehole

• Manual or automated reading • Measurements on a daily basis during spring thaw and on a weekly basis during the rest of the year are recommended

• Installed through a borehole in existing pavement structures • The tube needs to be carefully anchored underneath the maximum expected frost depth to avoid frost jacking

• Manual (observation of water level) or automated (water pressure) reading • If monitored manually, requires adequate signing and often lane closure for safe reading • Monitoring not possible when frost is present in pavement

167

168

Chapter Four

Tests for Material Selection Selection of the binder for hot mix asphalt (HMA) mixtures can be conducted using results from empirical tests described in Table 4-17 or from tests that measure fundamental properties used in performance based grading systems described in Table 4-18. Tests that aid in selection of aggregates for HMA are described in Table 4-19. Tests specific for cold regions that allow use of winter traction devices include aggregate toughness testing that simulates wear by studded tires in wet conditions. The Nordic abrasion test described in Table 4-19 is widely used in the Nordic countries and Alaska. Figure 4-35 shows the test equipment for the Nordic abrasion test.

Compacted Mixture Sample Preparation and Conditioning Gyratory compactor is used to prepare mixture samples for determining the mechanical and volumetric properties of asphalt-aggregate mixtures. The prepared samples simulate the density, aggregate orientation, and structural characteristics obtained in actual roadway when proper paving mix placement procedure is used (AASHTO T312). Adequate weight of aggregate is mixed with the binder at appropriate binder content and mixing temperature or adequate weight of mixture is collected from a plant to produce a sample of a desired size. The mixture is prepared typically at a temperature at which the kinematic viscosity of the unaged binder is 170 mm2/s. The temperature can be determined from the bitumen test data chart shown in Fig. 7-4. In the same way, compaction temperature is selected using 280 mm2/s equivalent viscosity. Before compaction the mixture is conditioned in an oven according to the mix design requirements. AASHTO R30 requires conditioning of 2 h in a forced-draft oven at the mixture’s compaction temperature for volumetric mix design test samples, and conditioning of 4 h at 135°C for test samples prepared for mechanical property testing. After the conditioning, the mixture is placed in a 150-mm-diameter mold that is positioned in the gyratory compactor at an angle (see Fig. 4-36 for a schematic picture of the test setup). A loading ram actuates a compaction pressure that is maintained during the test, while the bottom plate is rotated at a constant rate. The resulting kneading action is continued until a predetermined number of gyrations is achieved. The height of the specimen is recorded for each gyration and used in calculation of the specimen volume and uncorrected specimen bulk unit weight. After compaction the sample is extruded from the mold and its bulk specific gravity is determined. Using the determined mixture bulk (Gmb) and maximum specific gravities (Gmm) the relative density of the sample (Gmmx) at any gyration (x) can be determined from Eq. (4-15): %Gmmx =

Gmb hm × 100 % Gmm hx

(4-15)

where hm = height of the extruded specimen and hx = height of the specimen after x gyrations. The compacted mixture samples may then undergo long-term mixture conditioning. AASHTO R30 calls for five days conditioning in a forced-draft oven at 85°C.

Tests for Volumetric Mix Design The HMA volumetric mix design is covered in Chap. 7. In order to determine the volumetric mix design parameters, such as air voids in the mixture, specific gravities of the materials and the mixture need to be measured. The required specific gravities and their test methods are described in Table 4-20.

Test

Illustration

Description

Application

Penetration at 5°C and 25°C, measured in 0.1 mm ASTM D5

A needle with a weight is released and let penetrate 5 s into the asphalt sample at the test temperature

Consistency, stiffness, penetration grade classification

Softening Point, °C ASTM D36

Two discs of bitumen cast in brass rings are heated in a bath at 5°C/min while supporting a steel ball. The softening point is the temperature at which the two discs soften enough to allow each ball to fall 25 mm.

Drainability, plastic deformation, transition from viscoelastic to viscous behavior

Viscosity, 60°C, Pa·s ASTM D2171 AASHTO T202

A fixed volume of binder is placed in a viscometer that is placed in bath at the test temperature. The binder is let flow and the time is measured for the binder to flow past the two marks in the viscometer. The kinematic viscosity is then calculated using the time and a conversion factor. The dynamic viscosity at 60°C is determined similarly, but the binder is drawn up by means of vacuum

Plastic deformation, AC grade classification

Viscosity, 135°C, mm2/s ASTM D2170 AASHTO T201

169 TABLE 4-17

Binder Tests for Penetration Grade and AC Grade Specifications (Continued)

Constructability

170 Test

Illustration

Description

Application

Fraass breaking point, °C DIN 52012

A 0.5-mm-thick binder film is spread on a steel plate. The plate is cooled at 1°C/min and bent and straightened once a minute. The temperature at which the sample breaks is the Fraass breaking point in °C

Flexibility/brittleness at low temperatures

Thin film oven test ASTM D 1754 AASHTO T179

A 3-mm-thick film of asphalt cement is placed on a pan that is slowly rotated. The horizontal rotating does not cause the asphalt film to flow exposing more surface area. The oven has intake and outlet vents for natural air exchange. After the tests, mass change is reported

Aging simulation, work safety, smoking. The residue from these tests is used for additional rheometric tests (see Tables 7-2 and 7-3)

Rolling thin film oven test (RTFOT) ASTM D 2872 AASHTO T240

35 g of asphalt cement flows along the walls of a horizontally placed cylinder that is attached into a vertically rotating frame. The flowing asphalt exposes new surface continuously. Air is blown once into to the cylinders during each rotation. After the tests, mass change is reported

Flash point, °C ASTM D92 AASHTO T48

70 mL of binder is poured into a test cup. The temperature of the specimen is increased, while at specified intervals a test flame is passed across the cup. The flash point is the lowest liquid temperature at which the test flame causes vapors of the test specimen to ignite

Safety

Solubility in trichloroethylene ASTM D2042

A 2-g sample is dissolved in trichloroethylene and filtered. The insoluble material is washed, dried, and weighted. The percentage of soluble matter is calculated

Purity

TABLE 4-17

(Continued)

171

172 Test

Illustration

Description

Application

Viscosity measured with rotational viscometer, Pa·s ASTM D4402 AASHTO T316

A tube of asphalt cement is heated at the test temperature in a cylindrical chamber shown. A specified spindle is submerged into the sample and rotated. The torque required to keep the spindle spinning with the specified rate is measured, and the viscosity is calculated with the torque and the spindle dimensions

Constructability

Dynamic Shear (DSR), kPa AASHTO T315

The standard test procedure measures the complex shear modulus (G*) and phase angle (d) of asphalt binders with parallel plate test geometry. Binder sample is cast between parallel metal plates. One of the plates is oscillated with respect to the other at preselected frequencies and rotational deformation amplitudes. The complex shear modulus is the ratio of total shear stress to total shear strain. The phase angle, d, is related to the time lag between the shear and responding strain

Fatigue, plastic deformation

Bending beam rheometer (BBR) AASHTO T313

A small creep load is applied to a binder at a specified temperature and the deformation as a function of time is measured. The creep stiffness is calculated with the applied load and the beam dimensions

Direct tension tester AASHTO T314

A small “dog bone”-shaped asphalt cement sample is pulled at a slow, constant rate until it fails. The elongation at failure is used to calculate the failure strain, which indicates if a binder behaves in a brittle or ductile manner at low temperatures

Pressure aging vessel test (PAV) AASHTO R28

RTFOT residue (see Table 4-17) is exposed to high pressure and temperature for 20 h. The pressure aging apparatus consists of the pressure aging vessel shown and a forced draft oven. Three 50-g PAV samples are prepared on a pan for each binder tested and aged under 2070 kPa pressure either at 90, 100, or 110°C

1

Low temperature cracking1

Long-term in-service aging simulation

173

The critical cracking temperature method given in the AASHTO MP 1a and PP42 is obtained by first estimating the development of thermal stress using the BBR test results and then comparing the stress curve with binder’s tensile strength obtained from the DDT.

TABLE 4-18

Binder Tests for Performance Based Specifications (AASHTO MP 1a, 2003)

174

Chapter Four Test Method

Description

Application

Soundness ASTM C88 AASHTO T104

Aggregate sample is repeatedly immersed in saturated solutions of sodium or magnesium sulfate followed by oven dr ying. During the dr ying phase, the salt precipitates in the permeable pores of the aggregate. On reimmersion, the salt rehydrates and exer ts internal expansive forces that simulate the expansive forces due to freezing water. The test result is total percent loss over various sieve inter vals for a specified number of cycles (e.g., 5 cycles)

Durability— disintegration due to in-service weathering

Nordic abrasion EN 1097-9 (CEN 2006) ATM 312 (AKDOT&PF, 2005)

An aggregate sample (passing 16.0 mm sieve and retained on 11.2 mm sieve) is rotated in a standard mill (see Fig. 4-35) with 7 kg of 15-mmdiameter steel balls and 2 L of water for 1 h at 90 rpm. After the test the sample is sieved through a 2-mm sieve, and the ball mill value is defined as the percent passing the 2-mm sieve

Toughness— disintegration due to mechanical degradation during construction and in service

Los Angeles abrasion method ASTM C131 or 535 AASHTO T96

The test measures degradation of coarse aggregates (>2.36 mm) resulting from abrasion, impact and grinding in a rotating steel drum. The drum contains a specified number of steel spheres and shelf plates that pick up and drop the sample and the spheres. After the specified number of revolutions, the aggregate is sieved to measure the degradation as percent loss of material

Percentage of fractured faces ASTM D5821

The mass percent of coarse aggregate (>4.75 mm) with one or more fractured faces

Uncompacted void content of fine aggregate AASHTO T304

A sample of washed and dried fine aggregate (5. The particles are evaluated in a proportional caliber that divides the particles pass/no pass of the ratio requirement

TABLE 4-19 Test Methods Used in HMA Aggregate Evaluation

Angularity

Shape

Strength of asphalt concrete

I n v e s t i g a t i o n a n d Te s t i n g Test Method

Description

Application

Clay content (sand equivalent test) ASTM D2419 AASHTO T176

A sample of fine aggregate is mixed with a flocculation solution and agitated to loosen clayey fines. After a settling period, the cylinder height of suspended clay and settled sand is measured. The sand equivalent value is computed as the ratio of the sand to clay height, expressed as a percentage

Purity

Deleterious materials ASTM C142 AASHTO T112

The test measures the mass-% of contaminants, such as clay lumps, shale, wood, mica and coal in the blended aggregate. An aggregate sample is wet-sieved over specified sieves. The mass percent of material loss is reported as the percent of clay lumps and friable particles

TABLE 4-19 (Continued)

FIGURE 4-35 Nordic abrasion test equipment.

FIGURE 4-36 Schematic of gyratory compaction device.

175

176

Chapter Four Test Method

Description

Specific gravity of coarse aggregate ASTM C127 AASHTO T85

A sample of aggregate is immersed in water for 15 h, removed, surface dried, and weighed. The sample is subsequently weighed while submerged. Finally the sample is oven dried and weighed. The specific gravities are calculated as illustrated in Fig. 4-37

Bulk and apparent specific gravity of fine aggregate ASTM D128 AASHTO T84

Approximately 1 kg of fine aggregate is oven dried, covered with water and let stand 15–19 h. The sample is then dried on a flat sur face under a current of warm air and frequent stirring. A cone test is conducted periodically to determine when sur face-dr y condition is reached: a standard cone mold with its large diameter down in loosely filled with the fine aggregate and then lightly tamped with 25 tamper drops. The mold is lifted ver tically. When the aggregate ceases holding its molded shape, it is sur face dr y. At this point 500 g of aggregate is placed in a pycnometer with water and agitated to remove all air bubbles. The pycnometer is filled with water to its capacity, and the specific gravities are determined as illustrated in Fig. 4-38

Specific gravity of asphalt cement ASTM D70 AASHTO T228

The asphalt cement sample is placed in a pycnometer and weighed. The volume of the sample is obtained by filling the container level full of water and weighing in the air. The specific gravity of the asphalt cement is obtained with the same principle as illustrated in Fig. 4-38

Specific gravity of mineral filler ASTM D854 AASHTO T100

A sample containing natural moisture or oven dried is placed in a pycnometer and covered with distilled water. Entrapped air is removed either by applying a vacuum of 13.33 kPa or by gently boiling for at least 10 min while occasionally rolling the pycnometer. The pycnometer is filled to its capacity with distilled water, and the specific gravity is obtained as illustrated in Fig. 4-38 using the equation for apparent specific gravity. If a sample with natural moisture was used, the oven-dr y weight of the sample is determined at the end of the test

Theoretical maximum specific gravity of loose pavement mixture ASTM D2041 AASHTO T209

A sample of oven-dry and loose paving mixture is placed in a vacuum vessel and covered with water. Vacuum is applied for 15 min to gradually reach a suction pressure of 3.7 kPa and then gradually released. The volume of the sample is determined by filling the container full of water and weighing in the air, and the maximum specific gravity is obtained as illustrated in Fig. 4-38 using the equations for apparent specific gravity

Bulk specific gravity of compacted asphalt mixture ASTM D1188 / D2726 AASHTO T166

A sample of compacted mixture is oven dried and weighed. The sample is subsequently weighed while submerged. Finally the sample is surface dried and weighed. The specific gravities are calculated as illustrated in Fig. 4-37

TABLE 4-20

Specific Gravity Tests for Volumetric Mix Design

FIGURE 4-37 Principle of determination of specific gravity of paving materials with immersion method.

FIGURE 4-38 Principle of determination of specific gravities of paving materials with pycnometer method.

177

Chapter Four

Tests for Moisture Damage Moisture damage in pavements is difficult to predict, although important in cold regions. Several test methods exist to evaluate pavements’ resistance against moisture damage, but none of them has obtained vast popularity. The tests could be divided in tests using loose asphalt-aggregate mixture and in tests using compacted asphaltaggregate mixture specimens. Common test methods are described in Table 4-21. The advantages of the tests on loose mixtures are simple equipment and procedures. However, the test results are qualitative and subject to interpretation. The test results also ignore the effects of traffic, climate, and mixture properties (Solaimanian et al. 2003).

Description

Static immersion test AASHTO T182

100 g of oven-dry aggregate (6.3–9.5 mm) is coated with bitumen at a specified mixing temperature and cured for 2 h at 60°C. After curing, the sample is mixed with spatula until it has cooled to room temperature. The aggregate is then immersed in distilled water at 25°C. After 16–18 h, the bitumen-aggregate mixture is evaluated under water. The total area of retained bituminous film is estimated visually being below or above 95%; below 95% denoting “failure” and above 95% denoting “passing”

Boiling test ASTM D3625

Rolling bottle

Loose mixture—screening tests

Test Method

Modified Lottman AASHTO T283

Immersion compression ASTM D1075 AASHTO T165

Marshall immersion test

TABLE 4-21

Compacted specimen

178

250 g or coated aggregate is immersed in boiling water. The water is brought back to boiling and maintained for 10 min. The water is decanted after cooling and the aggregate sample is spread on white paper towel. The amount of stripping is determined by visual inspection Aggregate chips are coated with binder and placed in glass bottles filled with water at 20°C. The bottles are then slowly rotated. The coverage is visually estimated after 5, 24, 48, and 72 h. An example of test results is shown in Fig. 4-39 Six samples are prepared to a target air void ratio of 7%. Three of the samples are tested as dry, and the other three after conditioning that exposes them to the effects of moisture. The conditioning consists of partial vacuum saturation followed by a freeze cycle and a 24-h thaw cycle in warm water at 60°C. The moisture sensitivity is the ratio of the average tensile strengths of the conditioned subset divided by the average tensile strengths of the control subset Eight samples are prepared to a target air void ratio of 6%. Four of the samples are tested as dry, and the other four after conditioning that exposes them to moisture. The conditioned samples are immersed in water at 49°C for 4 days or at 60°C for 24 h. The samples are then moved to a water bath at 25°C for 2 h and tested for compressive strength (deformation rate of 1.27 mm/min per 25 mm of height). The index of retained strength is reported The conditioning phase of the test is identical to the one used for immersion compression test. However, Marshall stability is used instead of compressive strength

Moisture Sensitivity Tests (Continued)

I n v e s t i g a t i o n a n d Te s t i n g Test Method

Description

Environmental conditioning system ECS

A specimen (102 mm in diameter and in height with 7.5% target air voids) is subjected to static immersion saturation for 5 min, enclosed within a membrane and placed in a resilient modulus (MR) test setup. Water at room temperature is circulated through the specimen for 1 h, after which the vacuum is released and the reference MR is measured. The specimen is then conditioned for 6 or 18 h by allowing water at 60°C flow through the sample while a compressive cyclic load is applied to the specimen. After 6 h, the circumference of the sample is measured. The process is stopped if the circumference has increased more than 2%, and the material is considered moisture susceptible. Otherwise the conditioning is continued for the remaining 12 h after which the specimen is let cool and the MR is measured again. If the MR ratio is, e.g., ≥0.8, the mixture is considered well-performing (Solaimanian et al. 2003)

Wheel testers

For example, Hamburg wheel tracking device: Two cylindrical gyratory compacted samples are immersed in water at 50°C. The device then applies rolling steel wheel passes on the cylinders until 20 mm of deformation is reached or a maximum of 20,000 passes. The test results are illustrated in Fig. 4-40. Research suggests that the stripping inflection point is higher than 10,000 for pavements that are moisture resistant (Solaimanian et al. 2003)

TABLE 4-21

(Continued)

FIGURE 4-39 An example of rolling bottle test results (courtesy of Per Redelius, AB Nynäs Petroleum).

179

180

Chapter Four

FIGURE 4-40 Example of Hamburg wheel tracking test results [Solaimanian et al. 2003 (Figure 7, p. 96); reproduced with permission of TRB].

Tests for compacted asphalt-aggregate mixtures comprise of preparation of two sets of samples. One set is conditioned dry to the testing temperature. The other set undergoes either immersion or more severe handling, such as vacuum saturation and freeze-thaw cycling. The Superpave mix design (a mix design method adapted recently by most of the states in the United States and also by other road agencies; see Chap. 7 for more information) evaluates the moisture sensitivity of compacted mixtures using indirect tensile strength test at 25°C (AASHTO T283). Use of Marshall stability ratio has been used in the past and is again regaining approval (Mostafa et al. 2006). Other methods that test the moisture susceptibility of compacted asphalt mixtures include different types of wheel tracking devices, where conditioned samples are exposed to a rolling wheel.

Mixture Performance Tests Pavement performance tests include tests for resistance against plastic deformation and fatigue cracking. These tests are described in Table 4-22 (see Chap. 7 for more information on performance testing). Performance tests that predict the pavement behavior in cold climates include tests for low-temperature cracking resistance and tests for resistance against wear by studded tires. Monismith et al. (1965) suggested that in order to predict the low-temperature cracking resistance of pavement mixtures, the thermally induced stress, strength, and temperature at failure could be measured in a laboratory test that simulates the conditions to which a pavement slab was subjected in the field. The basic requirement for the test system is that it maintains the test specimen at constant length during cooling. The thermal stress restrained specimen test (TSRST) specified, for example, by the AASHTO TP10-93 is the most recent version of this system (Jung and Vinson, 1994). The TSRST is shown in Fig. 4-41. A beam or cylindrical specimen is mounted in the load frame that is enclosed by the cooling cabinet. The chamber and specimen are cooled with vaporized liquid nitrogen. As the specimen contracts, linear variable differential transducers (LVDTs) sense the movement and a signal is sent to the computer that in turn causes the screw jack to stretch the specimen back to its original length. This closed-loop process continues

Test

Description

Application

Dynamic modulus ASTM D3497, variations are being developed

Cylindrical specimens are loaded with a uniaxial haversine stress pattern. Resulted strains are measured with LVDTs attached to the sides of the sample to calculate the dynamic complex modulus |E*| and the phase angle, f

Fatigue, plastic deformation, mix design, MEPD software

Wheel track1

A wheel moves back and forth (or one way) in an environmental chamber

Fatigue, rutting, research

Road simulator

The road simulators consist of a circular track loaded by a rotating wheel(s) with actual vehicle tire and tire pressure. The systems are enclosed in an environmental room with temperature and moisture control systems

181

TABLE 4-22

Illustration

Pavement Performance Tests (Continued)

182 Test

1

Illustration

Description

Application

Beam fatigue

A piston rod applies upward haversine load cycles with a rest period to an HMA beam. Downward load, approximately 10% of the upward load, is applied to force the beam back to its horizontal position and stay there during the rest period. The dynamic deflection of the beam at a midspan is measured with an LVDT. A range of stresses is used to establish the fatigue relationship at various test temperatures

Fatigue, research

IDT creep compliance and strength1

A static load is applied along the diametral axis of a specimen for a fixed duration of time. The vertical and horizontal deformations are measured near the center of the specimen and used to calculate a tensile compliance at a particular duration of time. The strength test is conducted immediately after the creep compliance test. Without releasing the creep load, a constant rate of vertical deformation is applied to the specimen until it fails

Thermal cracking in MEPD software (see Chap. 7)

Photo courtesy of U.S. Department of Transportation, Federal Highway Administration.

TABLE 4-22

(Continued)

I n v e s t i g a t i o n a n d Te s t i n g

FIGURE 4-41 (a) Schematic of TSRST system and (b) TSRST specimen after testing [after Jung and Vinson 1994 (Figure 1, p. 13); reproduced with permission of Transportation Research Board].

as the specimen is cooled and ultimately fails. Measurements of elapsed time, temperature, deformation, and tensile load are recorded with a data acquisition system. The thermally induced stress gradually increases as temperature decreases until the specimen fractures (see Fig. 2-24). At the break point, the stress reaches its maximum value, which is called the cracking strength, with a corresponding cracking temperature (see Fig. 2-24). Two laboratory tests exist to evaluate compacted asphalt-aggregate mixtures’ suitability for road use under-studded tire traffic: the Prall test and the PWR test. In the Prall test (EN 12697-16 Method A, CEN 2006), a cylindrical specimen (Fig. 3-12) having a diameter of 100 mm and a height of 30 mm is conditioned at 5°C and then hammered for 15 min with forty bouncing steel spheres. The steel spheres are bounced using a rotating counter force at 950 rpm (see Fig. 4-42). Water is circulated continuously at 5°C, which rinses the worn pavement particles out of the testing chamber. The loss of volume in cm3 is the Prall or abrasion value. It is defined as the ratio of the mass difference

FIGURE 4-42 Schematic of Prall device.

183

184

Chapter Four

FIGURE 4-43 PWR testing equipment. (Photograph courtesy of SR Consulting Ltd.)

of surface-dry water-stored specimens weighed in the air before and after the test to the bulk density of the specimen. In the PWR method (EN 12697-16 Method B, CEN 2006), three miniature-studded tires are rotating around a wet 100-mm Marshall mix design sample at 5°C for 2 h (see Fig. 4-43). The test result, an abrasion value, is the volume of lost mixture in cm3 during the test.

4-3-2

Soils and Unbound Materials

From the mechanistic point of view, the only property of soils and unbound pavement materials that matters is stiffness. Indeed, most mechanistic-empirical pavement design and pavement analysis models are based on elastic properties (resilient modulus and Poisson ratio) of soil and pavement materials. Although they can be improved using different types of treatments, soils are generally considered as a “given” in pavement engineering. Pavement materials are selected and modified to meet specific requirements. In addition to stiffness required to ensure proper load distribution, other properties are generally sought to ensure seasonal stability and long-term durability of mechanical

Property

Test (Standard) Purpose

Stiffness Direct measurement or estimation from simple test (CBR or stabilometer)

Resilient modulus and poison ratio (AASHTO T307) • Determination of the resilient modulus of soils and unbound granular materials based on cyclic triaxial testing for different stress states California Bearing Ratio (CBR) (AASHTO T193; ASTM D1883) • Determination of a material stiffness/strength index based on the load required to force penetration at constant speed of a piston in the sample Stabilometer or R-Value (AASHTO T190; ASTM D2844) • Determination of a material stiffness index based on the resistance of a confined compacted sample to induced lateral deformation using a pressurized fluid

TABLE 4-23

Tests for Soils and Unbound Materials (Continued)

185

186 Properties contributing to stiffness

Property

Test (Standard) Purpose Particle size distribution

Particle-Size Distribution (AASHTO T88, T27; ASTM D422, C136) • Determination of proportion of soil mass in each particle-size class. Soil sample is washed through a set of standard sieves and dry mass retained in each sieve is recorded

Density

Laboratory Compaction Characteristics (AASHTO T99, T180; ASTM D698, D1557) • Determination of the relationship between water content and dry density of compacted soil or material. Compaction tests are conducted for various water content using standard compaction energy and protocol. Dry unit weight is recorded as a function of water content

Particle shape Particle surface texture

Amount of flat and elongated particles • Determination of the massic proportion of particles meeting the flat and/or elongated particle criteria for the specified particle-size class Amount of crushed particles • Determination of the massic proportion of particles meeting the fractured face criterion for the specified particle-size class

Durability (Long term) Hardness

Los Angeles abrasion test (AASHTO T96; ASTM C131) • Determination of the massic proportion of the sample crunched to a specified particle-size class after rotation in a steel drum with a mixing blade in presence of steel balls

Micro-Deval abrasion test (AASHTO T327) • Determination of the massic proportion of the sample reduced by attrition to a specified particle-size class after rotation in a smooth steel drum in presence of steel balls and water

Durability

Soundness test (AASHTO T104; ASTM C88) • Determination of the massic proportion of the sample reduced by weathering to a specified particle-size class after cycles of exposure to magnesium sulfate crystallization and drying in an oven

Stability (Seasonal) Permeability and water retention

Constant head permeability test (AASHTO T215; ASTM D2434, D1557) • Determination of the permeability of pavement materials based on the measurement of a water flow induced through a soil sample (Permeability can be estimated from particle-size distribution) Suction test • Determination of water retention characteristics of soils and pavement materials

Frost susceptibility (segregation potential)

187

TABLE 4-23

(Continued)

Frost heave test • Measurement of frost heave rate or total frost heave resulting from a thermal gradient induced in a soil sample placed in a freezing cell (Frost susceptibility can be qualified based on particle-size distribution)

188

Chapter Four properties of these materials. Table 4-23 summarizes soil and material testing generally required for pavement engineering purposes.

Assessment of Soil and Material Stiffness Stiffness is a fundamental property of soils and unbound pavement materials for mechanistic design and analysis. Resilient modulus is now widely accepted as the best available parameter to characterize mechanical properties of unbound pavement materials and subgrade soils. It is one of the main parameters required to compute mechanical response (stresses, strains, and displacements) of pavements subjected to loading. Resilient modulus is essentially an elastic modulus measured in conditions representative of stress state and history experienced by unbound pavement materials and soils in a pavement system. It can be measured directly on an intact or reconstituted sample subjected to cyclic loading in a triaxial cell. It can also be estimated from the CBR test or from physical properties of soils and aggregates. By improving the quality of grain contacts, high densities increase the stiffness of unbound materials and the load distribution effectiveness. Particle-size distribution can be used to maximize the density of a given material while the density/water-content relationship obtained through the laboratory compaction test provides a reference for achievable field density. Particle shape and surface texture have a significant influence on the level of density of pavement materials and on the quality of grain contact. Flat and elongated particles tend to reduce density and stiffness of granular materials. They are also more prone to fragmentation, and therefore, their content should be limited. Crushed particles tend to increase internal friction in granular material. As a consequence, stiffness and strength tend to increase with increasing crushed particle content. However, they also resist compaction and, for equal compaction energy, density tends to be reduced with increasing crushed particle content.

Factors Contributing to Stiffness

Laboratory Measurement of Resilient Modulus The resilient modulus test method is designed to accurately represent the loading conditions of soils and materials while remaining simple and manageable. Resilient modulus testing is done in a triaxial cell using computer-controlled cyclic loading. As illustrated in Fig. 4-44, the compacted sample is placed between two loading platens and wrapped by a latex membrane. Water can flow freely to the sample base and head through porous stones attached to the platens. Confinement pressure is applied in the triaxial chamber using a pressurized fluid (water, oil, or air). Cyclic axial stress is applied to the sample using a loading piston attached to the top platen. The cyclic load pulse transmitted to the sample has a haversine shape with a 0.1-s loading period and a 0.9-s rest period (Fig. 4-45). After 500 to 1000 cycles of sample conditioning at the stress level specified in Table 4-24, the sample is submitted to a series of 100 load cycles for each stress level specified in Table 4-24. Axial strain is recorded in the center portion of the sample using two or three displacement transducers attached to the wall of the sample. The average of the five recovered strains, er is used to compute the resilient modulus MR according to Eq. (4-16): MR =

σd εr

(4-16)

where sd is the deviator stress = s1 − s3 (see Fig. 4-44). Resilient modulus is a mechanical property of soils and unbound pavement materials which strongly depends on the level of stress applied to the specimen. Resilient

I n v e s t i g a t i o n a n d Te s t i n g

FIGURE 4-44 Schematic illustration of the triaxial cyclic loading apparatus used to determine the resilient modulus of soils and unbound pavement materials.

FIGURE 4-45 (a) Loading conditions and (b) parameters used for the determination of the resilient modulus.

modulus test results are thus, usually reported as a function of stress state. The generalized constitutive equation proposed as part of the Mechanistic-Empirical Pavement Design Guide (M-E PDG; NCHRP 1-37A, ARA 2004) is as follows: k

 θ  2 τ  MR = k1 pa    oct + 1   pa   pa

k3

(4-17)

where MR = resilient modulus, MPa, q = bulk stress (s1 + 2s3), toct = octahedral shear stress = 31 (σ 1 − σ 2 )2 + (σ 1 − σ 3 )2 + (σ 2 − σ 3 )2 , pa = normalizing stress (atmospheric pressure), k1, k2, k3 = regression constants obtained by fitting resilient modulus test data to Eq. (4-17).

189

190

Chapter Four Granular materials Stresses, kPa

Soils Stresses, kPa Step

s3

sd

s3

sd

Conditioning

101.4

27.6

103.4

103.4

1

41.4

13.8

20.7

20.7

2

41.4

27.6

20.7

41.4

3

41.4

41.4

20.7

62.1

4

41.4

55.2

34.5

34.5

5

41.4

68.9

34.5

68.9

6

27.6

13.8

34.5

103.4

7

27.6

27.6

68.9

68.9

8

27.6

41.4

68.9

137.9

9

27.6

55.2

68.9

206.8

10

27.6

68.9

103.4

68.9

11

13.8

13.8

103.4

103.4

12

13.8

27.6

103.4

206.8

13

13.8

41.4

137.9

103.4

14

13.8

55.2

137.9

137.9

15

13.8

68.9

137.9

275.8

These loading conditions are specified in AASHTO T307 procedure for the determination of resilient modulus of soils and unbound pavement materials.

TABLE 4-24

Loading Conditions Specified in AASHTO T307 (from Tables 1 and 2 from T307 in Standard Specifications for Transportation Materials and Methods of Sampling and Testing, 2003, by the American Association of State Highway and Transportation Officials, Washington, D.C. Used by permission.)

More simple constitutive relationships are also used to report resilient modulus testing results. Among others, the simple k-θ model is expressed as follows (Huang 2004): MR = k1θ k2

(4-18)

As described in Sec. 2-3 (Chap. 2), in addition to overburden stress and transient load stress, moisture content is also likely to induce internal negative pressure or matric suction, which also has an important influence on the resilient modulus. Figure 4-46 illustrates the effect of the degree of saturation on the resilient modulus of granular base materials as measured on some 20 materials sampled and tested at three levels of saturation as part of the Canadian Strategic Highway Research Program (C-SHRP) (Doucet and Doré 2004). Level of saturation is one of the important parameters related to seasonal change in material conditions. Resilient modulus should, thus, be seasonally adjusted by testing with various water contents going from saturation (spring conditions) to low saturation

I n v e s t i g a t i o n a n d Te s t i n g

800

200 Initial Saturated Drained

100

600 400 200 0 0

Initial Drained

150

M R – M R sat (MPa)

Resilient modulus, M R (MPa)

1000

50 0

–50 200

400 600 Bulk stress, (kPa)

800

–100

0

20

(a)

40 60 80 Degree of saturation (%)

100

(b)

FIGURE 4-46 (a) Influence of the level of saturation on the resilient modulus of a granular base material sample and (b) relationship between the level of saturation and the change in resilient modulus (from saturated conditions) for C-SHRP samples (Doucet and Doré, 2004).

levels (summer conditions). It is important to consider that saturation level will not fully explain seasonal changes in material behavior, which can also be affected by changes in density through freeze-thaw cycles as explained in Sec. 3-8-3 (Chap. 3).

Prediction of Resilient Modulus Resilient modulus can also be estimated from simple tests or physical properties of soil and unbound materials. The most commonly used approach is to estimate MR from CBR test results. Several correlations are proposed to link the CBR to MR. The following model is proposed for level 2 MR assessment in the M-E PDG (ARA 2004): MR (MPa) = 17 . 6 × CBR 0. 64

(4-19)

CBR testing is relatively simple and, being based on the measurement of a stress-strain relationship, it is somewhat related to resilient modulus. CBR value is the load required to force penetration of a 50-mm-diameter piston at a constant speed in a compacted sample expressed as a proportion (percent) of a reference load. The highest value between the load measured at 2.5 mm and the load measured at 5.0 mm is recorded at the CBR value. Despite several imperfections, CBR has the merit of being a measurement made specifically on the soil or material considered for construction. Resilient modulus can also be estimated based on stabilometer R-value based on the following model: MR(MPa) = 8.0 + 3.8·R

(4-20)

The stabilometer test is also a relatively simple test in which a relationship between applied stress and sample deformation is measured. The force is applied by increasing lateral confinement between 35 and 700 kPa for a sample subjected to a vertical confinement of 1120 kPa. The quantity of fluid required to increase the confinement pressure is used to measure deformation of the sample. Resilient modulus can also be estimated from soil or unbound material properties. Among several models proposed in the literature, the following models were proposed by Rahim and George (2005):

191

192

Chapter Four For fine-grained soils  σd  MR = k1 Pa 1 + 1 + σ c  

k2

(4-21)

with  LL k1 = 1 . 12(γ dr )1 .996   ω

0 . 639

and  LL  k2 = − 0 . 27(γ dr )1. 04 (ω cr )1. 46    P80 µm 

0 . 47

and, for coarse grained soils  θ  MR = k1 Pa 1 + 1 + σ d  

k2

(4-22)

with k2 = 0.12 + 0.90(gdr) – 0.53(wcr) – 0.017 (P80mm) + 0.314(logcu) and k2 = 0 . 226(γ dr ⋅ ω cr )

1 . 2385

 P80 µm   log(c )  u 

0 . 1 24

where Pa = atmospheric pressure, sd = deviator stress, sc = confining stress, q = bulk stress, gdr = gd/gdopt (ratio of dry density to maximum proctor dry density), LL = liquid limit, w = water content, wcr = w/wopt (ratio of water content to optimum proctor water content), P80mm = passing 0.080 mm and cu = uniformity coefficient.

Assessment of Material Durability Long-term pavement performance requires that the integrity of granular particles remain over extended periods of time. Construction operations (handling, placement, and compaction) tend to provoke fragmentation of particles. Figure 4-47a illustrates an example of the evolution of the grain-size distribution during construction of a pavement with a schistose aggregate. From the aggregate production to final compaction in the pavement, a 40 percent increase in the passing 5 mm and a 100 percent increase in the passing 80 mm were observed. Repeated loading of the pavement structure by construction vehicles and heavy traffic during the life of the pavement induces wear of the aggregates. The Los Angeles abrasion test is a good indicator of the resistance of mineral particles to fragmentation, while the Micro-Deval abrasion test is an indicator of long-term resistance of particles to wear.

I n v e s t i g a t i o n a n d Te s t i n g

FIGURE 4-47 Evolution of the grain-size distribution of schistose granular materials exposed to (a) construction operations and to (b) weathering.

Long-term exposure to weathering cycles such as wetting-drying, freezing-thawing, and warming-cooling is also likely to induce stresses in aggregates and ultimately to break them. Figure 4-47b illustrates the effect of exposure to these cyclic events for a schistose embankment material. Embankment materials were sampled near a surface exposed to weathering for 20 years, from another surface exposed recently and from an unexposed layer of the embankment. It can be observed that the evolution of coarse particle size occurs very rapidly during the first months of exposure. The proportion passing 5 mm has doubled in the first 4 months and do not appear to evolve significantly thereafter. Small particles tend to evolve more progressively as shown by the 50 percent increase in passing 80 mm during the first 4 months and an additional 50 percent increase afterward. Despite some problems with the reliability of the soundness test, it can provide some valuable insight on the durability of pavement aggregates exposed to weathering cycles.

Assessment of the Seasonal Stability of Soil and Material As previously discussed in Chaps. 2, 3, and 4, the mechanical behavior of soils and unbound pavement materials is highly dependent on the material density and level of saturation. Seasonal stability of the mechanical behavior of these materials is essentially a function of the materials’ ability to remain at a low level of saturation and at a high level of density. Seasonal stability is thus a function of three properties: • Permeability • Water retention characteristics • Frost susceptibility (segregation potential) Permeability and water retention characteristics will control the ability of the material to effectively drain excess pore water and to rapidly reach a low and stable saturation level. Permeability can be measured using the constant head permeability test (AASHTO T215; ASTM D2434, D1557) or estimated from grain-size distribution. Water retention characteristics can be assessed using the pressure-plate test (Fig. 4-48). In this test, a soil sample is subjected to increasing pressure forcing water to drain out of the sample. The relationship between applied pressure and water content can thus be

193

Chapter Four

FIGURE 4-48 The pressure-plate test for measurement of water retention characteristics of a soil.

established to construct a water retention characteristic curve such as the one illustrated in Fig. 2-25. Matric suction in unbound pavement material can have an important effect on the stiffness of these materials. Figure 4-49 illustrates the results of suction measurements and resilient modulus tests on some 20 samples of granular base materials sampled on the Canadian Strategic Highway Research Program (C-SHRP) test sites. The study shows that negative pressures exceeding 20 kPa are generated in samples with low saturation levels and the resulting increase in resilient modulus from saturated conditions can exceed 100 MPa. This type of variation can be found in cracked or unpaved pavements between spring-thaw conditions and dry summer conditions. Segregation potential can be measured directly on an intact or reconstituted sample subjected to step freezing in a freezing cell. It can also be estimated from simple tests or from physical properties of soils. 200

100 80 60 40

Initial Saturated Drained

0.1 kPa; 88 %

100

– 2.4 kPa; 52 % – 10 kPa; 28 %

20 0 – 25

Initial Drained

150

M R – M R sat (MPa)

Degree of saturation (%)

194

50 0

– 50 – 20

–5 – 15 – 10 Matric suction (kPa) (a)

0

5

– 100 – 25

– 20

–5 – 15 – 10 Matric suction (kPa) (b)

FIGURE 4-49 Effect of saturation level on (a) matric suction and (b) on resilient modulus of granular base materials (Doucet and Doré 2004).

0

I n v e s t i g a t i o n a n d Te s t i n g Laboratory Determination of the Segregation Potential In the laboratory, the segregation potential of pavement subgrade materials is usually measured by freezing tests using step-freezing conditions, which simulate closely the freezing conditions of pavement subgrade soils. Under these conditions, frost penetrates at the selected rate and tends to stabilize at a certain level in the soil. Depending on the soil type, the soil is placed into the freezing cell using either the modified proctor procedure or soil consolidation. Undisturbed samples can also be tested with appropriate equipment. Several thermistors are inserted into a freezing cell wall and are in contact with the soil sample inside the cell. They are used to determine the temperatures regime in the sample throughout the test. The cell is fixed to a base plate in which thermal liquid can flow. A top plate is used to cover the specimen and thermal liquid can also flow in this plate. Porous stone and filter paper are placed at each end of the sample. Thermal liquid circulates through thermal baths in order to control the temperature at the base and the top of the sample. Temperatures of −4°C at the top of the sample and 2°C at the base of the sample are common for subgrade soils. A linear variable displacement transducer is fixed to the top plate shaft and installed on a fixed reference. To measure the segregation potential of pavement subgrade soils, an overburden pressure of approximately 20 kPa can be applied onto the sample in order to simulate the stress applied by a pavement structure of approximately 1 m. The sample is saturated using low hydraulic gradients to prevent fines migration in the sample. To perform the test, this burette is equipped with a mariotte to allow free flow of the water in the sample. Throughout the test, the temperatures at different depths within the sample and the frost heave are recorded at fixed intervals. A schematic illustration of frost cell is presented in Fig. 4-50. Typical results of a segregation potential test for St-Alban silty clay are presented in Fig. 4-51. To determine the segregation potential, the relationship between the frost front penetration in the sample and time must be plotted and the time to obtain steady-state

FIGURE 4-50 Frost heave cell.

195

196

Chapter Four

FIGURE 4-51 Typical results of a frost heave test on the St-Alban silty clay; (a) frost front depth versus time, (b) frost heaving curve, and (c) temperature gradients.

conditions must be determined as shown in Fig. 4-51a. Then, the heaving rate dh/dt must be determined at the time steady-state conditions are reached as shown in Fig. 4-51b. This is done by drawing the tangent line to the plot at that time. To determine the temperature gradient, the temperatures in the sample are plotted for different times. The temperature gradient is determined by drawing a tangent line at 0°C at the time of the steady-state conditions. The segregation potential of this sample from St-Alban is 141 mm2/°C·day.

Estimation of the Segregation Potential from Simple Tests or Soil Physical Properties Several authors have developed methods for estimation of the segregation potential from simple tests or physical soil properties. The following sections describe the methods developed by Rieke et al. (1983), Kujala (1991), Doré et al. (2004), and Konrad (2005). Rieke et al. (1983) performed a study on various combinations of sand, silt, and different types of clay in order to develop an empirical parameter that includes variables related to the soil fines fraction to estimate the segregation potential of soils. The fines percentages tested were 5, 10, and 20 percent and these fines were blends of silt, montmorillonite, and two types of kaolinite. Strong correlations were observed between the segregation potential and the specific surface area of the fines. In addition, a good correlation was also observed between the liquid limit of the fines and the specific surface area of the fines. The observations suggested that the segregation potential is dependent on the clay mineralogy of the soil. The segregation potential increases as the percentage of fines increases and decreases as the activity (defined as the liquid limit of the fines fraction divided by the % clay sizes in fine fraction) increases. The fines factor parameter Rf was suggested to estimate the segregation potential. Rf is defined by the

I n v e s t i g a t i o n a n d Te s t i n g

FIGURE 4-52 Estimation of the segregation potential with Rf.

equation given in Fig. 4-52, where % fines is the particles percentage with particle diameter < 75 mm and LLff is the liquid limit of the fine fraction. This relationship is presented in Fig. 4-52, where the segregation potential is plotted versus the fines factor, Rf . Kujala (1991) developed a predictive model for the segregation potential using two independent variables. Those are the volumetric water content wvol and the unfrozen water content a(T=−2.5°C). The model is presented in Fig. 4-53. Using these two independent variables, Kujala obtained regression coefficients ranging from 0.76 to 0.80. It was found that the unfrozen water content is the most influent variable on the segregation potential, since water flows through the partly frozen layer on the colder side of the zero isotherm. Doré et al. (2004) developed a method to estimate the segregation potential of various soils using simple tests including unfrozen water content measurements and the methylene blue test. The effect of unfrozen water content on the frost susceptibility is widely described in the literature (Kujala 1991). The unfrozen water is composed from adsorbed water to soil grains and capillary water. Thus, the segregation potential is related to adsorbed water content, which is related to soils specific surface, and to capillary unfrozen water. To perform an unfrozen water content test, Doré et al. (2004) use a 200-mm-high cylindrical sample (diameter = 101.4 mm). This sample is placed in a plastic mould using either dynamic modified proctor compaction or consolidation depending

197

198

Chapter Four

FIGURE 4-53 Prediction of segregation potential based on unfrozen water content (Kujala 1991).

on the soil type. The compacted soils are saturated from bottom up afterward. A porous plate and a filter paper are placed under the sample to ensure good water distribution and to prevent fines migration. The mould is entirely surrounded with circulating conduits connected to a programmable liquid bath and a thermistor is inserted into the sample, as shown in Fig. 4-54. A temperature conditioning of 1°C is applied to the sample until it has reached thermal stability. The sample is then cooled down at a rate of 0.01667°C/min. The unfrozen water content (UWC) is measured using ThetaProbes. UWC and temperature were recorded every 5 h. ThetaProbes measure the dielectric constant, which can be converted to the volumetric unfrozen water content of the sample

I n v e s t i g a t i o n a n d Te s t i n g

14

Unfrozen water content (Vallée-Jonction) DC +1

Dielectric constant

12 10 8 6 4 2 0 –6

–5

–4

–3 –2 –1 0 Temperature (°C) (b) Typical test results

(a) experimental setup

1

2

40

45

200

2

R = 0.7102

(

BV 0.6

–2.8273

(

160

SP (mm2 /°C*d)

SP = 56 894

x DC+ 1

120 80 40 0 0

5

10

15

20

25

30

35

0.6

* DC +1 /BV (c) Segregation potential

FIGURE 4-54 (a) Experimental setup for the unfrozen water content test, (b) typical results, and (c) correlation of test parameters with the segregation potential.

using appropriate calibration. Dielectric constant was, however, used directly in the test in order to avoid inducing errors through the calibration function. Typical results of dielectric constant as a function of temperature are presented in Fig. 4-54. Using unfrozen water content tests and methylene blue tests, Doré et al. (2004) developed the relationship given in Fig. 4-54 to estimate the segregation potential. In the relationship, SP(mm2/°C·day) is the segregation potential, g is the ratio of the dielectric constant measured at −2°C divided by the dielectric constant measured at 1°C, DC+1 is the dielectric constant measured at 1°C, and BV(cm3/g) is the blue value measured with the methylene blue test. All factors selected to build the relationship represent parameters that are physically linked to the segregation freezing process. DC+1 represents the total water volume available for freezing in the sample pores, while g represents the unfrozen water proportion in the frozen sample. Those two parameters are proportional to SP, since the available unfrozen water is the main path for water flowing in freezing soil. On the other hand, the BV value is inversely proportional to the segregation potential since it is related to the soil specific surface. A higher BV leads to a decreasing water flow channels, since water adsorbed by particles is fixed and cannot contribute significantly to water flowing. The standard error of the estimate of the relationship is 24.5 mm2/°C·day. This relationship is based on 21 measurements of unfrozen water content, and methylene blue tests on several subgrade soils sampled in various Quebec geologic conditions and ranging from silty clay to sand and gravel.

199

200

Chapter Four Konrad (2005) developed a methodology to estimate the segregation potential using the frost heave response of two reference soils. This methodology is based on Konrad’s (1999) demonstration that the segregation potential can be assessed from soil index properties that considers the grain-size distribution and the fines content, the clay mineralogy, the soil fabric, and the overburden pressure. According to his work, the segregation potential with no surcharge SP0 is related to the fines fraction (< 75mm), d50 of the fine fraction (FF), the specific surface of the fines fraction SS and the ratio of the material’s water content to its liquid limit, w/wL. These properties are related to water movements in capillary channels. It is suggested to use two reference soils used in the study performed by Rieke et al. (1983), which are sand-silt-kaolinite mixture with fines content of 20 percent. Using these soils, it can be observed that both SS and d50(FF) increase with increasing clay mineral content. Using the two reference soils, the following relationships suggest the reference characteristics: (1) For d50(FF) 1mm, SS− ref = 25 . 95 − 11 . 78 × log(d 50 (FF))

(4-25)

SP0− ref = 489 − 232 × log(d50 (FF))

(4-26)

in which d50(FF) is expressed in mm. Good relationships were observed for w/wL ratios of 0.7 and >0.8. The following equations are proposed to characterize the frost heave response of fine-grained soils: (1) For SS/SS-ref 0.8)

(4-28)

(if w/wL = 0.7 ± 0.1)

(4-29)

(2) For SS/SS-ref >1, SP0/SP0 ref = (SS/SS ref )− 0. 85

SP0/SP0 ref = 1 . 5(SS/SS ref )−0 . 55

(w/wL > 0.8)

(4-30)

As a reference, Table 4-25 provides typical values of SP as a function of frost susceptibility classification.

I n v e s t i g a t i o n a n d Te s t i n g Frost Susceptibility Class

Segregation Potential, mm2/°C·day

Negligible

192

Source: Saarelainen 1996.

TABLE 4-25

Segregation Potential of Soils

Review Questions 4-1. (a) What is the apparent resistivity of the soil, if during a resistivity measurement 0.4 A circulates between electrodes (A and B, Fig. 4-8) at 10 m apart and a potential difference of 2 V is measured between inner electrodes (M and N) 2 m apart? What type of soil could it be? (b) The person performing the test misread at first the distance AB and used 100 m instead of 10 m for a previous test on the same soil. If the resistivity measured was 1900 Ω, what could one conclude? 4-2. A horizontal resistivity profile needs to be done. The current intensity used is 15 A and the inner electrodes are 6 m apart. For the first four tests, distanced of 15 m, the potential difference measured is 1, 4, 1.3, and 3.8 V, respectively. What are the corresponding electrical resistivities and what could one conclude about the soil type and configuration?

4-3. A seismic refraction survey has given the following results: Geophone Number

Distance from Impact Point, m

Time Required for the Waves to Reach the Geophones, s

1

20

0.150

2

40

0.300

3

60

0.322

4

80

0.345

5

100

0.366

6

120

0.385

What is the stratigraphy of the soil investigated and what types of soil are present?

4-4. A seismic refraction survey has given the following results: Geophone Number

Distance from Impact Point, m

Time Required for the Waves to Reach the Geophones, s

1

20

0.133

2

40

0.266

3

60

0.287

4

80

0.307

5

100

0.310

6

120

0.313

201

202

Chapter Four What is the stratigraphy of the soil investigated and what types of soil are present? Would the same results have been obtained if the topmost layer had been frozen?

4-5. A vane test has been conducted to estimate the bearing capacity of a soft soil. If a blade set of 120 mm of length and 60 mm of diameter is used and the minimum torque observed during the test was 12 N·m, what is the maximum thickness of the road embankment that can be supported by the soil? Consider the embankment material’s unit weight to be 22 kN/m3. 4-6. A light weight deflectometer was used to estimate the elastic modulus of a homogenous soil. The following results were obtained with a rigid plate of 75 mm of radius. If Poisson’s coefficient is 0.2, what is the elastic modulus of the soil?

so, kPa

d, mm

13

0.021

27

0.041

53

0.086

68

0.103

96

0.160

4-7. A falling weight deflectometer has been used to evaluate the pavement response to dynamic loading. The following data were acquired. Evaluate the surface curvature index, the base curvature index, the radius of curvature of the center of the basin, and the tensile strain if the asphalt-bound layer is 120-mm thick and the loading plate has a radius of 150 mm.

Distance, mm

0

150

300

450

600

750

Deflection, mm

225

212

201

185

153

120

Distance, mm

900

1050

1200

1350

1500

93

75

63

55

49

Deflection, mm

4-8. Estimate the resilient modulus of a subgrade soil, using data from Question 4-6, if the pavement total thickness is 600 mm. 4-9. An oven-dry sample of coarse aggregate has a mass of 2.4 kg. If the apparent specific gravity is 1.72 and the bulk specific gravity of the dry sample is 1.68, what is the bulk specific gravity of the saturated-surface-dry sample?

4-10. A pycnometer was used to evaluate specific gravities. The apparent specific gravity of the sample is 1.85. The weight of the sample and vessel full of water is 12.8 N. The pycnometer has a mass of 500 g and can contain 500 mL of water. What is the apparent volume of the sample? 4-11. The following results obtained from a resilient modulus test done in a triaxial cell.

I n v e s t i g a t i o n a n d Te s t i n g s1, kPa

s3, kPa

er

58

43

8.571E-05

73

43

1.604E-04

120

43

3.348E-04

40

25

1.351E-04

69

25

3.077E-04

135

25

5.116E-04

25

12

1.711E-04

52

12

3.960E-04

80

12

5.271E-04

Find the coefficients k1, k2 of the kq model [Eq. (4-18)] for the soil tested.

4-12. Considering the following grain-size distribution of a coarse-grained soil and s1 and s3 data from Question 4-11, find the resilient modulus of the soil, using Eq. (4-21). Use gd/gopt = 90 percent and w/wopt = 75 percent. Sieve Opening, mm

Passing, %

31.5

100

20

96

14

83

10

69.6

5

50

2.5

39.2

1.25

30.4

0.63

21

0.315

13.8

0.16

9.8

0.08

5

4-13. The 50 percent passing diameter of the fine fraction of a soil is 5 mm. Considering a specific surface of 0.015 km2/kg and a ratio w/wL to be 0.85, in what frost susceptibility class is the soil? What would be the frost heave rate if the thermal gradient was 1°C/m? If the ratio was 0.8 instead of 0.85, would it be in the same class?

4-14. The segregation potential of a soil has been estimated to be 12 mm2/°C·day. If the ratio of the material’s water content to its liquid limit is 0.82 and the 50 percent passing diameter of the fine fraction is 0.8 mm, what is the specific surface of the fine fraction?

203

204

Chapter Four

References AASHTO (2003). June 2003 Edition of AASHTO Provisional Standards, American Association of State Highway and Transportation Officials, Washington, D.C. AKDOT&PF (2005). Alaska Test Methods Manual. Alaska Department of Transportation and Public Facilities. Juneau, Alaska. Allard, M., Lévesque, R., Séguin, M. -K., and Pilon, J. A. (1991). “Les caractéristiques du pergélisol et les études préliminaires aux travaux de génie au Québec nordique (Permafrost characteristics and preliminary studies for engineering work in northern Quebec),” Report for Quebec Ministry of Transportation, Centre d’études nordiques, Université Laval, p. 94 (in French). American Geological Institute (1976). Dictionary of Geological Terms, Anchor Books, New York, New York, p. 471. ARA Inc. (2004). “Guide for the Mechanistic-Empirical Design of New and Rehabilitated Pavement Structures,” Final report, NCHRP 1-37A. Transportation Research Board of the National Academies, Washington, D.C. Atkins, H. N. (2003). Highway Materials, Soils, and Concretes, Prentice Hall, United States, p. 383. Avery, T. E. (1977). Interpretation of Aerial Photographs, 3d ed., Burgess Publishing Company, Minneapolis, Minn., p. 392. Boutet, M. (2007). “Élaboration de modèles mathématiques pour l’interprétation des données obtenues avec le pénétromètre à cône dynamique (Development of mathematical models for the interpretation of dynamic cone penetrometer data),” Master thesis, Civil Engineering Department, Laval University, Quebec City, Canada (in French). Brown, R. J. E. (1974). “Some Aspects of Airphoto Interpretation of Permafrost in Canada,” Technical paper no. 409 of the Division of Building Research, National Research Council of Canada, Ottawa, Canada. Budhu, M. (2000). Soil Mechanics and Foundations, John Wiley & Sons, Hoboken, New Jersey, p. 586. CEN (2006). Standards and Drafts, European Committee for Standardization, http:// www.cenorm.be/cenorm/index.htm (December 14, 2006). Conseil National de Recherche Canada. (1988). “La terminologie du pergélisol et notions connexes (Permafrost terminology and related notions),” Note de service No. 142, Ottawa, Canada. Doré, G., Pierre, P., Abdelwahab, A. I., Juneau, S., and Bilodeau, J. P. (2004). “Développement d’un essai simple et rapide pour l’estimation du potentiel de segregation (Development of a simple and rapid test for the estimation of the segregation potential),” Research report GCT-2004-14, Civil Engineering Department, Laval University, Quebec City (in French). Doucet, F., and Doré, G. (2004). “Module réversible et coeffcient de poisson réversible des matériaux granulaires C-LTPP (Resilient modulus and resilient poison coefficient of the C-LTPP granular materials),” Proceedings of the Annual Conference of the Canadian Geotechnical Society, Canadian Geotechniocal Society, Quebec City, Canada (in French). Fradette, N., Doré, G., Pierre, P., and Hébert, S. (2005). “Evolution of the Pavement Winter Roughness,” Transportation Research Record: Journal of the Transportation Research Board, No. 1913, Transportation Research Board of the National Academies, Washington, D.C., pp. 137–147.

I n v e s t i g a t i o n a n d Te s t i n g Fredlund, D. G., and Rahardjo, H. (1993). Soil Mechanics for Unsaturated Soils, Wiley InterSciences. Hoboken, New Jersey. Gagnon, H. (1974). La photo aérienne (Aerial photography), Les éditions HRW, Montréal (in French). Grenier, S. (2007). “Analyse dynamique du déflectomètre à masse tombante (Dynamic analysis of the falling weight deflectometer),” Ph.D. thesis, Laval University, Civil Engineering Department. Quebec City, Canada (in French). Haas, R. (1997). Pavement Design and Management Guide, Transportation Association of Canada, Ottawa, Canada, p. 389. Hoekstra and McNeill (1973). “Electromagnetic Probing of Permafrost,” Permafrost: North American Contribution to the Second International Conference, National Academy of Sciences, Washington, D.C., pp. 517–527. Horak, E. (1987). “The Use of Surface Deflection Basin Measurements in the Mechanistic Analysis of Flexible Pavements,” Proceedings of the Sixth International Conference on the Structural Design of Asphalt Pavements, vol. 1, International Society for Asphalt Pavements, White Bear Lake, Minn., pp. 990–1001. Huang, Y. H. (2004). Pavement Analysis and Design. 2d ed. Pearson Prentice Hall, Upper Saddle River, New Jersey. Janoo, V. C., and Berg, R. L. (1990). “Thaw Weakening of Pavement Structures in Seasonal Frost Areas,” Transportation Research Record: Journal of the Transportation Research Board, No. 1286, Transportation Research Board of the National Academies, Washington, D.C., pp. 217–233. Jung, D., and Vinson, T. (1994). “Thermal Stress Restrained Specimen Test to Evaluate Low-Temperature Cracking of Asphalt-Aggregate Mixtures,” Transportation Research Record: Journal of the Transportation Research Board, No. 1417, Transportation Research Board of the National Academies, Washington, D.C., p. 13. Jung, F. W. (1988). “Direct Calculation of Maximum Curvature and Strain in Asphalt Concrete Layers of Pavements from Load Deflection Basin Measurements,” Transportation Research Record: Journal of the Transportation Research Board, No. 1196, Transportation Research Board of the National Academies, Washington, D.C., pp. 125–132. Knutsson, S., Domaschuk, L., and Chankler, N. (1985). Analysis of large scale laboratory and in situ frost heave tests, Fourth International Symposium on Ground Freezing, Sapporo, Japan, Kinosita, S. and Fukuda, M. (eds.), pp. 65–70. Konrad, J. -M., and Morgenstern, N. R. (1982). Prediction of frost heave in the laboratory during transient freezing, Canadian Geotechnical Journal, vol. 19, no. 3, pp. 250–259. Konrad, J. -M., and Morgenstern, N. R. (1983). Frost susceptibility of soils in terms of their segregation potential Prediction of frost heave in the laboratory during transient freezing, Proceedings of the Fourth International Conference on Permafrost, National Academies Press, Washington, D.C. Konrad, J. -M. (1999). “Frost Susceptibility Related to Soil Index Properties,” Canadian Geotechnical Journal, vol. 36, pp. 403–417. Konrad, J. -M. (2005). “Estimation of the Segregation Potential of Fine-Grained Soils Using the Frost Heave Response of Two Reference Soils,” Canadian Geotechnical Journal, vol. 42, no. 1, pp. 38–50. Kujala, K. (1991). “Factors Affecting Frost Susceptibility and Heaving Pressure in Soils,” Acta Universitatis Ouluensis, Series C 58, Oulu, Finland.

205

206

Chapter Four Ladanyi, B. (1996). “La conception et la réhabilitation des infrastructures de transport en régions nordiques (Design and rehabilitation of transportation infrastructure in northern regions),” Études et recherches en transports, RTQ-94-07, Gouvernement du Québec, Ministère des Transports, p. 123 (in French). Lambert, J. P., Fleming, P. R., and Frost, M. W. (2006). “Laboratory Assessment of Coarse Granular Road Foundation Materials,” TRB 2006 Annual Meeting CD-ROM. Transportation Research Board of the National Academies, Washington, D.C. LCPC (1982). “Reconnaissance géologique et géotechnique des tracés de routes et autoroutes (Geological and geotechnical survey of highway alignments),” Note d’information technique, Ministère de l’urbanisme et du logement—Ministère des Transports, Paris, France, p. 111 (in French). Lo, C. P. (1976). Geographical Application of Aerial Photography, Crane, Russak & Company, N.Y., David & Charles, London, p. 330. Loudon, A. A., and Partners (1995). Cold Deep in Place Recycling: Technical Recommendation and Application Specifications, South Africa. McCarthy, D. F. (1998). Essentials of Soil Mechanics and Foundations Basic Geotechnics, Fifth edition, Prentice Hall, New Jersey. Mollard, J. D., and Janes, J. R. (1985). “Airphoto Interpretation and the Canadian Landscape,” Dept. of Energy Mines and Resources, Canada, p. 415. Monismith, C., Secor, G., and Secor, K. (1965). “Temperature Induced Stresses and Deformations in Asphalt Concrete,” Journal of the Association of Asphalt Pavement Technologists, vol. 34., White Bear Lake, Minn. Morin, P. (1994). Manuel canadien d’ingénierie des foundations (Canadian Foundation Engineering Manual), Seconde édition, Société canadienne de géotechnique, Richmond, Canada, p. 558 (in French; also available in English). Mostafa, A., Abd El Halim, A. O., Easa, S., and Niazi, Y. (2006). “Suitable Test Method for Predicting Effect of Stripping on Mechanical Properties of Canadian Pavements” Proceedings of the Tenth International Conference on Asphalt Pavements, International Society for Asphalt Pavements, White Bear Lake, Minn., vol. 2, pp. 562–571. OECD (1984). “Surface Characteristics of Pavement Surfaces, Their Interaction and Their Optimization,” Research in Pavements and Transportation, Organisation for Economic Co-operation and Development, Paris. Paré, J. J., Lavallée, J. G., and Rosenberg, P. (1978). Frost penetration studies in glacial till on the James Bay hydroelectric complex, Canadian Geotechnical Journal, vol. 15, no. 4, November, pp. 473–493. Penner, E., and Ueda, T. (1977). The dependence of frost heaving on load application, Proceedings of Frost action in soils 1, University of Lulea, Sweden. Phukan, A. (1985). Frozen Ground Engineering, Prentice-Hall International Series on Civil Engineering and Engineering Mechanics, Upper Saddle River, New Jersey, p. 336. Rahim A. M., and George K. P. (2005). Models to estimate subgrade resilient modulus for pavement design, The International Journal of Pavement Engineering, Taylor & Francis, Oxfordshire, U.K., vol. 6, no. 2, pp. 89–96. Riddle, C. H., and Hardcastle, P. K. (1991). “Drilling & Sampling of Permafrost for Site Investigation Purposes: A review,” AIME International Arctic Technology Conference, Anchorage, Alaska, 29-31 May, Journal of Society of Petroleum Engineers, pp. 611–620. Rieke, R., Vinson, T., and et Mageau, D. (1983). “The Role of Specific Surface Area and Related Index Properties in the Frost Heave Susceptibility of Soils,” Proceedings of the Fourth International Conference on Permafrost, National Academies Press, Washington, D.C., pp. 1066–1071.

I n v e s t i g a t i o n a n d Te s t i n g Rhode, G. T. (1994). Determining a Pavement Structural Number from FWD Testing, TRB 73rd Annual Meeting, Preprint no. 940351, Transportation Research Board, Washington, D.C. Saarelainen, S. (1996). “Pavement Design Applying Allowable Frost Heave,” Proceedings of the Eighth International Conference on Cold Regions Engineering, ASCE Press, Reston, Va. Sayers, M. W. (1995). “On the Calculation of IRI from Longitudinal Road Profile,” Preprint TRB 74th Annual Meeting, Washington, D.C., January 1995. Sayers, M. W., and Karamihas, M. (1998). The Little Book of Profiling, University of Michigan, Transportation Research Institute, http://www.umtri.umich.edu/ content/LittleBook98R.pdf (July 25, 2007). Schmalzer, P. N. (2006). LTPP Manual for Falling Weight Deflectometer Measurements, V 4.1, FHWA-HRT-06-132, Federal Highway Administration, Office of Infrastructure Research and Development, McLean, Va. Solaimanian, M., Harvey, J., Tahmoressi, M., and Tandon, V. (2003). “Test Methods to Predict Moisture Sensitivity of Hot-Mix Asphalt Pavements,” Proceedings of Moisture Sensitivity of Asphalt Pavements, A National Seminar, San Diego, California, TRB, February 4–6, p. 96. St-Laurent, D. (1995). “Évaluation structurale de chausses souples dans un contexte climatique nordique (Structural evaluation of flexible pavements in cold climate),” Rapport GCS-85-05, Civil Engineering Department, Laval University, Quebec City, Canada (in French). Sylwester, R. E., and Dugan, B. (2002). “Evaluation of Geophysical Methods, Field Program,” Report No, FHWA-AK-RD-02-07, Alaska Department of Transportation and Public Facilities, Juneau, Alaska. Todd, D. K. (1980). Ground Water Hydrology, John Wiley & Sons, Inc., Hoboken, New Jersey. Van Deusen, D. A. (1996). Selection of Flexible Backcalculation Software for the Minnesota Road Research Project, Final report, Minnesota Department of Transportation, St Paul, Minn. Von Quintus, H. L., and Simpson, A. L. (2002). “Back Calculation of Layer Parameters for LTPP Test Sections, Volume II: Layered Elastic Analysis for Flexible and Rigid Pavements,” Report FHWA-RD-01-113. Federal Highway Administration, Washington, D.C.

207

CHAPTER

5

Calculation of Engineering Parameters

D

esigning a structure that extends over large distances across several geologic and climatic environments is a major challenge for pavement engineers. Prior to pavement construction, natural soils are in balance with their environment. The construction of the pavement structure will unavoidably disrupt this balance by modifying temperatures, moisture, and stress regimes in the natural soil. The following principles are proposed as general guidelines when undertaking a pavement project in cold climates: 1. Minimize disruptions of natural soil conditions. 2. Always strive for more stable conditions. 3. Minimize spatial and temporal variations in soil and material properties. Important environmental effects on pavement systems have been described in the previous chapters. It is important to take into consideration all of those factors while preparing a pavement project in cold regions. However, not all of the factors need to be implicitly taken into account in engineering calculations. This section proposes a list of basic parameters that need to be obtained or calculated in the design process. They include: • Air temperature, freezing and thawing indices (Sec. 5-1) • Surface temperature, freezing and thawing indices (Sec. 5-2) • Representative temperature in hot mix asphalt (HMA) layer (Sec. 5-3) • Thermal properties of soils and pavement materials (Sec. 5-4) • Freezing and thawing indices within the pavement structure (Sec. 5-5) • Frost and thaw depth (Sec. 5-6) • Frost heave (Sec. 5-7) • Thaw settlement for pavements in permafrost areas (Sec. 5-8) • Stresses and strains at critical pavement interfaces (Sec. 5-9) With modern computer capacity, most of these parameters can be calculated using sophisticated computerized methods. This section proposes simple calculations methods that can be used manually for estimation and verification purposes.

209

210

Chapter Five When available, computer programs allowing for detailed calculation of the parameters will be identified.

5-1 Air Temperature and Air Freezing and Thawing Indices Most pavement engineering applications require the surface temperature as a boundary condition parameter. However, information on surface temperature is rarely available and the parameter has to be estimated from a widely available climatic parameter: air temperature. As described in Sec. 2-1, excluding changes due to alternating climatic systems, air temperature follows daily and seasonal temperature cycles. Figure 5-1 illustrates temperature data taken in Quebec City between February 1st and 16th, 2004. The figure illustrates hourly temperature data, daily averages, and medium term trend. A nice daily pattern can be observed on February 5th: the minimum temperature was −16.6°C at 8:00 hours and the maximum temperature of −8.9°C was reached at 15:00 hours. The average daily temperature during that day was −12.9°C. When to use each of these parameters? The relevant air temperature parameters to be considered for each application are given in Table 5-1. Mean air temperature: Mean air temperature for a given period can be obtained by averaging mean daily temperatures over the period considered: MATt =

n

1 MDAT n∑

(5-1)

t=1

where MDAT is the mean daily air temperature and n is the number of days in the period considered (see Example 5-1).

FIGURE 5-1

Air temperature data for 15 days in February 2004 in Quebec City (Canada).

Calculation of Engineering Parameters Parameter

Application

Mean air temperature (MAT) Mean annual air temperature (MAAT)

• Estimation of the yearly temperature variation • Estimation of the boundary condition at the bottom of the pavement system (Tb ≈ MAAT) • Indication of presence of permafrost (if MAAT 0°C)

Maximum air temperature

• Selection of asphalt cement grade • Prediction of asphalt concrete rutting tendency

Minimum air temperature

• Selection of asphalt cement grade • Prediction of asphalt concrete cracking tendency

Maximum daily cooling rate

• Prediction of asphalt concrete thermal cracking performance

Freezing index

• Prediction of frost depth and frost heave

Thawing index

• Prediction of thaw depth and thaw consolidation in permafrost conditions • Prediction of thaw depth in seasonal frost conditions

TABLE 5-1

Air Temperature Parameters Used in Pavement Engineering

Example 5-1 Considering the data shown in Fig. 5-1, the mean air temperature for the 15 day period is MAT15 =

1 15 MDAT 15 ∑ 1

MAT15 =

− 147 . 92 °C = − 9 . 86 °C 15

Mean annual air temperature: It is a climatic parameter which can usually be found in climatic databases. For many applications, the MAAT should be averaged over a number of years in order to obtain a representative value. When several years of data are used to assess MAAT, a probabilistic approach using probability functions and standard deviation of MAAT can be used to assess the probability of extreme values of MAAT over a given analysis period. By definition, MAAT is an historical value that can lead to errors when estimating future climatic conditions in a context of climate change. The use of historical temperature data for design without consideration for local warming trend can lead to unconservative designs. A probabilistic assessment of the evolution of climatic parameters over the design life of transportation facilities is now a must in cold region pavement engineering. Cooling rate: The air cooling rate is an important factor to take into consideration when analyzing thermal contraction and related stress development in asphalt concrete. The cooling rate can be readily obtained from hourly air temperature data by subtracting each hourly temperature by the previous hourly temperature. The result can be plotted as shown in the top part of Fig. 5-1. It can be seen from the plot that the maximum cooling rate was obtained at the end of the day on February 4th and exceeded 3°C/h. It can also be noted that a less intense, but more persistent cooling occurred on the 14th leading to the lowest temperature during the 15 day period. Air freezing and thawing indices from daily temperature data: Air freezing index (FIa) and air thawing index (TIa) are two widely used climatic parameters for the quantification

211

212

Chapter Five

FIGURE 5-2 Schematic illustration of the freezing and thawing indices.

of the “severity” of a winter or of a summer with respect to freezing or thawing effects on pavements. The indices can be defined as the area between the MDAT curve and the 0°C line over a given period of time (usually one year). Figure 5-2 is a schematic illustration of the freezing and the thawing indices. The freezing and thawing indices can be defined mathematically as follows: t

FI a = ∫ −T− dt 0

(5-2)

or t

FI a = ∑ − MDAT− 0 t

TI a = ∫ T+ dt 0

(2-3)

or t

TI a = ∑ MDAT+ 0

where T− and T+ are temperatures below and above 0°C, respectively, MDAT and MDAT are mean daily air temperatures below and above 0°C respectively, and t is the period of time considered (see Example 5-2). Example 5-2

Considering the data shown in Fig. 5-1, the freezing index for the 15 day period is t

FI a = ∑ −MDAT− = 148 °C ⋅ days 0

The freezing index corresponds to the area between the 0°C line and the MDAT line in Fig. 5-1. It also roughly corresponds to the area between the trend line and the 0°C line.

Calculation of Engineering Parameters

FIGURE 5-3 Sinusoidal representation of the temperature relationship with time.

Air freezing and thawing indices from annual summary temperature data: When daily temperature statistics are not available for a specific site, air freezing and thawing indices can be estimated using the assumption that air temperature is following a sinusoidal relationship with time. Two parameters are used to estimate the sinusoidal curve: the mean annual air temperature (MAAT) and the maximum amplitude of the sinusoid A0 (see Fig. 5-3). Based on the assumption, the air temperature “Ta” at any point “t” in time can be obtained from (Zarling and Braley 1988):  2π t 2πφ  Ta = MAAT − A0 cos  − p   p

(5-4)

where p is the period of time considered (365 days) and f is the phase lag (note that the cosine and sine arguments in Eqs. (5-4) to (5-8) are expressed in radians). The beginning of the thawing, t1, and the freezing, t2, seasons can readily be obtained from Eq. (5-4) by setting Ta = Tf (freezing temperature) in the equation (Zarling and Braley 1988):  MAAT − T f  p cos −1   +φ 2π A0  

(5-5)

 MAAT − T f   p  2π − cos −1   + φ 2π  A0    

(5-6)

t1 =

t2 =

It should be noted that the phase lag factor, f, can be used to adjust the timing of t1 and t2 as well as the occurrence of the maximum and minimum temperature during the

213

214

Chapter Five year. Having estimated the air temperature function, freezing, and thawing indices can be obtained from (Zarling and Braley 1988): FI a =

365+ t1

∫

t2

  2π t 2πφ   T f − MAAT + A0 cos  p − p   dt   

or

(5-7)

pA0 FI a = (T f − MAAT)( p + t1 − t2 ) + 2π

  2π ( p + t1 − φ )  2π (t2 − φ )  − sin  sin     p p     

and t

2    2π t 2πφ  TI a = ∫ MAAT − A0 cos  −  − T f  d t p p   t1 

or

(5-8)

TI a = (MAAT − T f )(t2 − t1 ) −

pA0 2π

  2π (t2 − φ )  2π (t1 − φ )  − sin  sin     p p     

Example 5-3 illustrates the determination of freezing and thawing indices using Eqs. (5-7) and (5-8). Example 5-3 Considering a mean annual air temperature of −2°C, an amplitude of the seasonal air temperature variation of 25°C and a phase lag of 30 days, compute the air temperature function and the freezing and the thawing indices. Solution Step 1: The air temperature function can be computed using Eq. (5-4) in a spreadsheet. The resulting function is given in Fig. 5-4: Step 2: Compute season change times (t1 and t2) from Eqs. (5-5) and (5-6): t1 =

365 days  − 2 − 0 cos − 1  + 30 = 125 . 9 days  25  2π

t2 =

365 days   − 2 − 0  2π − cos −1    + 30 = 2 9 9 . 1 days 2π 25   

Step 3: Compute air freezing and air thawing indices in °C·day from Eqs. (5-7) and (5-8):

FI a = (0 − (− 2))(365 + 125 . 9 − 299 . 1) +

 2π (299 . 1 − 30)  365 × 25   2 π (365 + 125 . 9 − 30)  sin    − sin     2π   365 365

FI a = 3278 . 9 TI a = (− 2 − 0)(299 . 1 − 125 . 9) − TI a = 2548 . 9

 2 π (125 . 9 − 30)  365 × 25   2 π (299 . 1 − 30)  sin   − sin   2 π   365 365   

Calculation of Engineering Parameters

FIGURE 5-4

Air temperature function computed for Example 5-3.

Creating a virtual weather station for site specific evaluation: When using climatic data for pavement evaluation purpose, it is recommended to make best possible use of information from weather stations located in the vicinity of the considered site. A “virtual” weather station can be created by combining the information from the nearby weather stations using a 1/r2 weighting scheme (Wu et al. 2000). Virtual weather data can be interpolated using the following equation: k

V=

V

∑ r 2i i =1 i k

1 ∑ r2 i =1 i

(5-9)

where V = climatic value to be estimated, Vi = climatic value measured at weather station i, ri = distance between weather station i and the pavement site, and k = number of stations considered for the estimation (see Example 5-4). Weather stations selected for the development of the virtual weather station should be within a 30 km radius of the considered pavement site and the difference of elevation should not exceed 500 m (Wu et al. 2000). Example 5-4 Considering the situation illustrated in Fig. 5-5, compute the freezing index for a virtual weather station located at the indicated pavement site. All sites are within a 500 m difference of elevation. Solution 1210 1250 1275 + + 2 22 2 302 = 22 . 9 0 6 = 1218 . 4 FI = 8 0 . 0188 1 1 1 + + 82 22 2 302

215

216

Chapter Five

FIGURE 5-5 Construction of a virtual weather station (Example 5-4).

Dealing with multiple freeze-thaw cycles: It is common to have periods of freezing temperatures alternating with periods of warm air temperatures. Though these alternating sequences can occur any time during the year, they are more commonly observed at the beginning of winter and at the beginning of spring. Corté et al. (1995) and Dysli et al. (1997) have proposed the concept of “significant freezing index.” The calculation of the significant freezing index is based on the following three logical rules: 1. (|FI(i)| > 25°C·day) and (TI(i+1) < 15°C·day) 2. |FI(i)| > TI(i+1) 3. TI(i+1) < |FI(i+2)| The freezing and thawing indices for n successive periods are accumulated if the three conditions are verified. Example 5-5 clarifies this concept.

FIGURE 5-6

Accumulation of freezing and thawing indices (Example 5-5).

Calculation of Engineering Parameters Example 5-5 Based on the data provided in Fig. 5-6, compute the significant freezing index. Solution First freezing event: Rule 1. (|FI(1)| = 20°C·day < 25°C·day) ⇒ rule not verified. The first event will thus, not be accumulated in the significant freezing index. Second freezing event: Rule 1. (|FI(3)| = 30°C·day > 25°C·day) and (TI(4) = 12°C·day < 15°C·day) ⇒ rule verified; Rule 2. |FI(3)| > TI(4) ⇒ rule verified; Rule 3. TI(4) < |FI(5)| = 300°C·day ⇒ rule verified. The second freezing event and the following thawing event will be accumulated in the significant freezing index. Thus FI = 30 – 12 + 300 = 318°C·day

5-2

Surface Temperature and Surface Freezing and Thawing Indices Surface temperature is the most important temperature data in pavement engineering as it represents the boundary condition at the surface of the pavement system. As explained in Sec. 2-1, surface temperature and air temperature are in complex interaction. A complete surface energy balance analysis [Eq. (2-8)] should ideally be done to obtain the true surface temperature. Such an approach is, however, impractical due to the difficulty to obtain site specific information for the calculation of the radiation balance, the convective heat exchange coefficients and other factors necessary to compute the surface energy balance (Shur and Slavin-Borowskly 1993). The current state of the practice in pavement engineering is, thus, to use empirical approaches to convert air freezing and thawing indices into surface freezing and thawing indices. Two approaches are proposed in the literature for the conversion: The first one uses a coefficient referred to as the “n-factor” which is a multiplier used to correct the air freezing and thawing indices. The second method consists of adding a correction term named the “radiation index” directly to the freezing index.

5-2-1 The n-Factor Approach The n-factor has widely been used for soil and pavement temperature analysis since the 1950s. Amongst others, it has been documented by Carlson (1952), Brown (1963), Lunardini (1978), Shur and Slavin-Borowskly (1993), and Andersland and Ladanyi (2004). The n-factor can be defined as follows: nf =

FI s FI a

(5-10)

TI nt = s TI a where nf , nt = the freezing and the thawing n-factors, FIa, FIs = air and surface freezing indices, and TIa, TIs = air and surface thawing indices. Notwithstanding the fact that the n-factor approach is highly practical, the variability of the factor makes it an unreliable tool. The n-factor is known to vary considerably with • Surface characteristics (albedo, latent heat of fusion/evaporation, thermal conductivity, and thermal capacity of soil) • Radiation balance on the pavement (latitude, season, cloud cover, slope, and direction of the slope, presence of shading obstacles)

217

218

Chapter Five nf

nt

Material

Range

Suggested Practical Range

Range

Asphalt concrete

0.25–2.50

0.8–0.95

1.60–3.00

Gravel

0.60–1.50

0.9–1.0

1.10–2.00

Trees and bushes, moss, and peat soil

0.25–0.50

0.30–0.35

0.37–0.80

Snow

1.00

Source: compiled from Dysli et al. (1997), Zarling and Braley (1988), Lunardini (1978), Ladanyi (1996), Andersland and Ladanyi (2004).

TABLE 5-2

Typical Values of n-Factors for Pavement Engineering Purposes

• Convective heat transfer (difference in temperature between air and surface, wind speed) • Damping effects by large water bodies Typical values for n-factor reported in the literature for pavement conditions are listed in Table 5-2. As a general rule, higher values of the practical range should be used where surfaces are exposed to intense sun radiation, while lower values of the range should be used in conditions where surfaces are protected from the sun radiation (clouds, shading obstacles, or low sun angle) or exposed to high winds. Note that typical values are difficult to find in the literature for nt as it varies considerably with site condition. Additional information can be found in the literature. For example, Shur and SlavinBorovskly (1993) propose a map of nt values for Russia (values for bare silty soils) showing variation of nt-factor with distance from a large water body, latitude, and period during the year. The U.S. Department of the Army (1966), cited by Andersland and Ladanyi 2004, also proposes a chart relating nt to wind speed for different pavement surface types. Lunardini (1978) proposes a rationale method to obtain n-factor values for specific site conditions. The method, however, requires detailed information on surface thermal characteristics and surface heat transfer characteristics and is, thus, difficult to apply in practice. More information in Lunardini’s approach can be found in Dysli et al. (1997) and Lunardini and Ibrahim (1990). The n-factor is sometimes used to directly convert air temperatures into surface temperatures. This approach assumes that the days with an average temperature below (or above) 0°C is the same for the air and surface temperatures. Prediction of an asphalt concrete surface temperature is given in Sec. 5-3.

5-2-2

Radiation Index Approach

Dysli (1991) and Dysli et al. (1997) have proposed an interesting alternative to the n-factor approach. The proposed approach is based on the premises that solar radiation is the main contributor to the energy balance at the surface of pavements and that the energy balance is generally obtained by summing the effect of various contributing factors including solar radiation. According to Dysli, the freezing and thawing indices at

Calculation of Engineering Parameters

FIGURE 5-7 The radiation index as a function of quantity of sunshine or total radiation, GH [redrawn from Dysli et al. 1997 (Fig. 4.5, p. 22) with permission from Taylor & Francis Books UK].

the surface of the pavement can be obtained by subtracting a correction factor termed the radiation index (RI) from the air freezing index as follows: FI s = FI a − RI

(5-11)

The radiation index can be obtained from Fig. 5-7 based on total radiation data from weather stations, on the average daily number of hours of sunshine during the considered period (also available from weather stations) or on a simple qualitative assessment of exposure to sunlight. The radiation index relationship proposed in Fig. 5-7 was derived from data gathered in Switzerland and is considered valid between the 40th and the 55th degrees of latitude north. Validation of the relationship should be done before applying it to other contexts. Example 5-6 demonstrates the calculation of surface freezing indices using the n-factor and RI approaches. Example 5-6 Considering a pavement asphalt concrete surface exposed to an average of 3 h of daily exposure to sunshine in moderately windy conditions, compute the surface freezing index using the n-factor and the RI approaches. The air freezing index obtained from a nearby weather station has been established as 1200°C·day. Solution The n-factor, based on the given site conditions; a n-value of 0.9 is selected from Table 5-2. FIs = 0.9 × 1200°C·day = 1080°C·day RI, from Fig. 5-7, RI can be estimated to be about 105°C·day FIs = 1200°C·day – 105°C·day = 1095°C·day

219

220

Chapter Five

5-3 Temperature in Asphalt Concrete Representative temperature in asphalt concrete is needed for pavement deflection analysis, selection of appropriate asphalt cement, and low-temperature cracking and rutting prediction testing. One single temperature analysis is not always adequate, as the deflection analysis needs the actual pavement temperature at the moment of the deflection measurement, whereas the selection of the asphalt cement and performance prediction testing need design maximum and minimum pavement temperatures. Determination of the maximum and minimum design pavement temperatures are discussed further in Chap. 7. This section describes the determination of representative pavement temperature for deflection or another type of analysis, when real time pavement temperature is needed. When the surface temperature is measured as a part of the analysis, such as deflection analysis, the representative pavement temperature can be estimated using the air temperature, measured pavement surface temperature, depth, and seasonal and diurnal temperature variations. The estimation could be conducted using mathematical models, such as finite difference approximation or finite element method (e.g., Hermansson 2002). Other methods are based on regression analysis where measured pavement temperatures have been fitted with sinusoidal seasonal and diurnal temperature variations. A model by Lukanen et al. (2000) is based on the long-term pavement performance (LTPP) data from 41 sites in North America including several sites in cold regions. The models given in Eqs. (5-12) and (5-13) apply for HMA layer thickness from 46 to 305 mm and surface temperature range from 0 to 40°C. Equation (5-12) is used with deflection measurements according to LTPP testing protocol, where the surface temperature is measured from a pavement that has been in shade for 6 min. Equation (5-13) used with deflection measurements according to routine testing methods, where the surface temperature is measured from a pavement that has been in shade for 15 to 30 s (Lukanen et al. 2000).   2π (h1 − 15 . 5) Td = 2 . 78 + 0 . 912 ⋅ IR + (log(d) − 1 . 25)  − 0 . 428 ⋅ IR + 0 . 5 5 3 (1 − day) + 2 . 63 ⋅ sin      18  2π ( h2 − 13 . 5) + 0 . 027 ⋅ IR ⋅ sin    18

(5-12)

  2π ( h1 − 15 . 5) Td = 0 . 95 + 0 . 892 ⋅ IR + (log(d) − 1 . 25)  − 0 . 448 ⋅ IR + 0 . 6 2 1 ⋅ (1 − day) + 1 . 83 ⋅ sin      18  2π ( h2 − 13 . 5) + 0 . 042 ⋅ IR ⋅ sin    18

(5-13)

where Td = pavement temperature at depth d(°C), IR = infrared surface temperature (°C), d = depth at which pavement temperature is to be predicted (mm), 1 − day = average air temperature the day before testing (°C), sin = sine function on an 18-h clock system (2p rad equal to one 18-h cycle), h1, h2 = time of day in 24-h clock system, but calculated using an 18-h asphalt concrete temperature rise- and fall-time cycle (rules given in Table 5-3). The hours are used as decimals, for example, 13:15 hours = 13.25. The 18-h sine function is assumed to have a flat −1.0 segment between 05:00 and 11:00 hours for the

Calculation of Engineering Parameters Time of the Day, h

h1

h2 Actual decimal hour + 24.00

0–3:00

Actual decimal hour + 24.00

3:00–5:00 5:00–9:00

9.00 11.00

9:00–11:00 11:00–24:00

Actual decimal hour Actual decimal hour

Source: based on Lukanen et al. 2000.

TABLE 5-3 Values for the Variables h1 and h2 in Eqs. (5-12) and (5-13)

FIGURE 5-8

Eighteen-hour cycle sine functions (adapted from Lukanen et al. 2000).

(h − 15.5) term, and between 3:00 and 9:00 hours for the (h − 13.5) term as shown in Fig. 5-8. Example 5-7 clarifies the determination of the h variables. Example 5-7 Pavement surface temperatures and deflections were measured from 8:00 to 16:00 hours using an infrared temperature gauge and a falling weight deflectometer. Some of the test results and testing times are given in Table 5-4. The surface was shaded for about 20 s during the routine type testing, and the average air temperature the day before was 14.0°C. Determine (a) values for variables h1 and h2 for the given testing times and (b) the pavement temperature for the measurement taken at 9:15 hours at the depth of 25 mm. Solution

(a) Determination of the values for the variables h1 and h2 using Table 5-3:

Time of the Day, h

h1

h2

4:12

4 + 12/60 + 24 = 28.2

9.0

8:00

11.0

9.0

9:15

11.0

9 + 15/60 = 9.25

13:21

13 + 21/60 = 13.35

13 + 21/60 = 13.35

221

222

Chapter Five Time of the Day, h

Measured Surface Temperature, °C

4:12 8:00

15.2

9:15

17.8

13:21

26.3

TABLE 5-4 Testing Times and Surface Temperatures

(b) Because the testing area was shaded for only 20 s, estimate the pavement temperature using Eq. (5-13):

T25mm

  − 0 . 448 × 1 7 . 8 + 0 . 621 × 14 +     . + . × . + (log( ) − . ) 0 95 0 892 17 8 25 1 25    2π (11 . 0 − 15 . 5)     1 . 83 ⋅ sin     ° C = 15 . 9 ° C = 18      2π (9 . 25 − 13 . 5) + 0 . 042 × 17 . 8 ⋅ sin     18  

For pavement temperature predictions, were the surface temperature cannot be measured, models developed with pavement temperature data from Temmes, Finland, could be used (Ryynänen 2000, Savolainen et al. 2001). Temmes’ latitude is approximately 65°C and the instrumented test sections were located in both sunny and shady areas. The maximum and minimum recorded air temperatures were +31.1 and −36.6°C, respectively. The pavement surface temperature for warm and cold season can be estimated with Eqs. (5-14) and (5-15) using the seasonal and diurnal temperature variation and average air temperature (Ryynänen 2000). 2π ( h − 10)  2π (day − 97 ) Tsw = 1 . 981 sin   + 6 . 655 sin   + 0 . 702Tair + 2 . 49 24 365

(5-14)

 2π (day − 97 ) Tsc = 6 . 655 sin   + 0 . 702Tair + 2 . 4 9 365

(5-15)

where Tsw, Tsc = surface temperatures for warm and cold season, respectively, h = hour of the day (from 0 to 24), day = sequel number of the days in a year from 0 to 365 and Tair = air temperature. The warm season is defined as time between March 31 and October 15, and the cold season is defined as time between October 16 and March 30. The average pavement temperature for the top 100 mm of asphalt concrete can then be estimated using Eq. (5-16) (Savolainen et al. 2001): 2π (4 . 681 − h)  2π (108 . 6 − day) Tave = 0 . 470 sin   − 1 . 212 sin   + 0 . 917Tsurface + 0 . 308 365 24 (5-16) where Tsurface is Tsw [Eq. (5-14)] for warm season and Tsc [Eq. (5-15)] for cold season. Example 5-8 illustrates the calculation of pavement surface and average temperatures using air temperature data.

Calculation of Engineering Parameters Example 5-8 Determine the average pavement temperature for the top 100 mm for a pavement with surface temperature of 17.2°C taken at 9:15 hours on July 3rd. Use an air temperature of 14.0°C. Solution July 3rd is the184th day of the year, so day = 184, and h = 9 + 15/60 = 9.25. Use Eqs. (5-14) and (5-16) to calculate the pavement temperature:

Tsw

  2π (9 . 25 − 10)   2π (1 84 − 97 )   + 6 . 655 sin  1 . 981 sin       °C = 18 . 57 °C 24 365 =   + 0 . 702 × 14 . 0 + 2 . 49 

Tave

  2 π ( 4 . 681 − 9 . 25)   2 π (108 . 6 − 184)   − 1 . 212 sin  0 . 470 sin       °C = 18 . 1 °C 24 365 =   + 0 . 917 × 1 8 . 57 + 0 . 308 

5-4 Thermal Properties of Soils and Pavement Materials The estimation of thermal conditions within the pavement system requires the use of basic thermal properties of soils and pavement materials. These properties include: • Thermal conductivity (k) • Heat capacity (c) • Latent heat of fusion (L) • Thermal diffusivity (a) Thermal conductivity: For most practical applications, it is estimated based on relevant physical properties of soils and pavement materials. The most commonly used reference for thermal conductivity of soils is characterization work done by Kersten in 1949. Fine-grained soils and coarse-grained soils were characterized for different dry densities and water content and results were reported in charts and equations. The results of Kersten’s work are summarized in the following four equations (converted to metric by Farouki 1981): Clay and silt: Unfrozen: k u = 0 . 1442(0 . 9 log ω + 0 . 2) × 100 . 6243 ρd

(5-17)

k f = 0 . 001442(10)1 . 373 ρd + 0 . 01226ω(10)0 . 4994 ρd

(5-18)

k u = 0 . 1442(0 . 7 log ω + 0 . 4) × 10 0 . 6243 ρd

(5-19)

Frozen:

Sand and gravel: Unfrozen:

223

224

Chapter Five Frozen: k f = 0 . 01096(10)0 . 8116 ρd + 0 . 00461ω (10)0 . 9115 ρd

(5-20)

where ku and kf are unfrozen and frozen thermal conductivities (W/m·°C) and w is gravimetric moisture content (%) and rd is dry density (Mg/m3) Heat capacity of soils is as well a function of density and moisture content of the soil. It can be estimated using the following empirical relationship (Ladanyi 1996): ρ  c vf = 4 . 187  d  (0 . 17 + ω u + 0 . 5ω f )  ρω 

(5-21)

where cvf = volumetric heat capacity of frozen soils (MJ /m3·°C), rd = dry density of soil (kg/m3), rw = density of water (1000 kg/m3), wu = unfrozen gravimetric water content (decimal), and wf = frozen gravimetric water content (decimal). For unfrozen soils, Eq. (5-21) becomes ρ  c vu = 4 . 187  d  (0 . 17 + ω)  ρω 

(5-22)

The corresponding mass heat capacities are obtained from relationship cm =

cv cv = ρ ρ d (1 + ω )

(5-23)

where r is the wet density of the soil. Latent heat of fusion of soil is the energy released by freezing water or absorbed by melting ice present in the pores of 1 m3 of soil. Latent heat of fusion of water being 0.334 MJ/kg, latent heat of fusion of soil (Ls) can be estimated using the following formula: Ls = ρd ⋅ ω ⋅ 334

(5-24)

where Ls = latent heat of fusion of the soil volume (kJ/m3), rd = dry density of soil (kg/m3), and w = water content (decimal). Thermal diffusivity of soil is defined as being the ratio of thermal conductivity on heat capacity:

α=

k cv

(5-25)

where a = thermal diffusivity (m2/s), k = thermal conductivity (W/m°C = J/s·m·°C), and cv = volumetric heat capacity (J/m3·°C). Table 5-5 gives typical values for different types of soils and common pavement materials.

Calculation of Engineering Parameters

Soil or Material

Thermal Conductivity, k (W/m°C)

Volumetric Heat Capacity, cv (MJ/m3°C)

Fresh snow

0.06–0.10

0.21

Compacted snow

0.3–0.6

0.42–1.05

Asphalt concrete

1.50

2.0–2.5

Granular material

1.3–1.7

2.0

Polystyrene insulation

0.03–0.06

0.04–0.06

Peat

0.6

3.0

Sand-gravel

1.2–3.0

2.4–3.0

Silt

1.2–2.4

2.5–3.1

Clay

0.9–1.8

2.6–3.4

Sources: Ladanyi 1996; Pufahl 1996; Andersland and Ladanyi 2004.

TABLE 5-5

Thermal Properties of Various Soils and Pavement Materials

It should be noted that for most thermodynamic calculations, a multilayer system can be replaced by an equivalent homogeneous volume of soil having thermal properties equal to the weighted average of the thermal properties of all layers included in the volume according to the following equation: n

Pv =

∑ Pi × Di i=1

(5-26)

n

∑ Di i=1

where Pv = equivalent value of the thermal property for a volume of thickness, Di = thicknesses of layer i included in the volume, and Pi = value of the property for layer i having a thickness Di.

5-5

Freezing and Thawing Indices within the Pavement Structure Without the use of numerical analysis or similar computing techniques, it is impractical to calculate actual temperatures within the pavement structure. Instead, the severity of the temperature variations within the pavement system is estimated using freezing and thawing indices. The indices can be predicted based on the thermal diffusivity “a ” of soils and pavement materials. Assuming that the annual temperature variation follows a sinusoidal trend, the temperature at any depth in the pavement system will follow a similar sinusoidal trend with a reduced amplitude Ax and a time lag ∆tx. Figure 5-9 illustrates the relationship between surface temperature and temperature at depth x in the pavement system. Ax and ∆tx can be estimated using  

Ax = A0 e − x

π

 p α 

(5-27)

225

226

Chapter Five

FIGURE 5-9

Temperature variations at a depth x in the pavement system.

and ∆tx =

x p πα 2

(5-28)

where A0 = amplitude of the sinusoidal temperature wave at the pavement surface (°C), x = depth in the pavement system (m), p = period considered (365 days), and a = computed using Eq. (5-25). The temperature at any point in time, the length of the freezing period, and the freezing index transmitted can thus, be estimated at depth x in the system by replacing A0 by Ax and f by (f + ∆tx) in Eqs. (5-4) to (5-8) (see Example 5-9). Example 5-9 Given the characteristics of the pavement provided in Table 5-6, estimate the freezing period at the surface of the subgrade soil and the freezing index transmitted at that level. A sinusoidal variation of surface temperatures, with an amplitude of 15.5°C, f = 30 days and a mean annual surface temperature of 5°C is assumed for the pavement. The resulting surface freezing index computed from Eq. (5-7) is 982°C·day. Solution Step 1: Compute the equivalent thermal properties for the pavement structure using Eqs. (5-26) and (5-25): k=

  (1 . 5 × 0 . 15) + (1 . 3 × 0 . 65) J = 1 . 34( W m°C) = 1 . 34    s ⋅ m ⋅ °C  0 . 15 + 0 . 65

cv =

 MJ  (2 . 2 × 0 . 15) + (2 . 4 × 0 . 65) = 2 . 36  3   m ⋅ °C  0 . 15 + 0 . 65

α=

1 . 34 × 10− 6 = 0 . 57(m 2 s × 10− 6 ) = 0 . 049(m 2 day) 2 . 36

Calculation of Engineering Parameters Layer

Thickness, m

k, W/m°C

cv, MJ/m3

Asphalt concrete

0.15

1.5

2.2

Granular base

0.65

1.3

2.4

Subgrade: Clayey silt: rd = 1600 kg/m3, w = 0.20, segregation potential (SP) = 100 mm2/°C·day

TABLE 5-6 Pavement Properties for Examples 5-9 to 5-12

Step 2: Compute the modified parameters of the sinusoidal curve at depth x using Eqs. (5-27) and (5-28):  

Ax = 15 . 5e − 0. 8 ∆ tx =

π

 365⋅0 . 049

= 11 . 08(°C)

0 . 8 365 = 19 . 48(day) π × 0 . 049 2

Step 3: Compute freezing period and freezing index at surface of subgrade soil using Ax and f + ∆tx in Eqs. (5-5), (5-6), and (5-7):

t1 =

 5−0  365 cos −1  y)  + 49 . 48 = 113 . 53(day  11 . 08  2π

t2 =

 5 − 0  365  2 π − cos−1  + 49 . 4 8 = 350 . 43(day)  11 . 08  2π 

The duration of the freezing period in the subgrade soil is, thus t fs = 113 . 53 + (365 − 350 . 43) = 128 . 10(day) And the transmitted freezing index is FIt = (0 − 5)(365 + 113 . 53 − 350 . 43) +

 2π (350 . 43 − 49 . 48)  365 × 11 . 08   2π (365 + 113 . 53 − 49 . 48) sin   − sin    2π 365 365  

FIt = 508 . 2(°C ⋅ day)

5-6

Frost and Thaw Depth Frost penetration in soils and pavement system is a result of the heat extraction process described in Sec. 2-1. As illustrated in Fig. 5-10, when surface temperature (Ts) is below the freezing temperature, the resulting thermal gradient will induce a heat flow toward the pavement surface. If the resulting heat flow is larger than the geothermal heat flow,

227

228

Chapter Five Ts

Tb

0°C T G1 ∆X

Frost depth

TG2

Heat flux

FIGURE 5-10

Heat extraction and progression of the frost front in the pavement system.

the system is unbalanced and attempts to regain balance by releasing heat. A sustained cold temperature at the surface of the system will consume the heat stored in the pavement system, and will eventually fall below the freezing temperature. The freezing front will initially progress rapidly in the pavement because the temperature gradient is steeper at shallow depth, TG1, and because the pavement materials in the top portion of the pavement are drier and have thus less heat accumulated (heat capacity and latent heat of fusion). When reaching lower layers of the pavement system (frost penetration ∆X) and eventually the subgrade soil, the frost front progresses more slowly because of the gentler thermal gradient, TG2, and the larger quantity of moisture available in the subgrade soil. It is, thus, easy to understand the role of an insulation layer that will impede heat flow and consequently reduce pavement cooling. Frost penetration is therefore a function of the thermal properties of soils and pavement materials and is a square root function of the “quantity of below-freezing temperature” (or freezing index) to which the system is exposed. A simple method is proposed to estimate the frost (thaw) depth in pavement systems. The method is a two step approach based on (1) the estimation of the freezing (thawing) index transmitted at the base of the pavement structure and (2) the calculation of frost (thaw) depth in the subgrade soil, assumed to be homogeneous. The method is based on the Stefan’s equation extended to include the effect of the segregation potential in soils. The modified Berggren equation, also known as the Aldrich-Berggren method (Aldrich 1956; Ladanyi 1996) can also be used to perform the frost depth calculation, but is not described in this book. Detailed explanation on the use of the modified Berggren equation to compute frost depth in a multilayer system can be found in Andersland and Ladanyi (2004). Frost and thaw depth can alternatively be calculated using free or commercial software.

5-6-1 Transmitted Freezing Index Method The transmitted freezing index method is based on the aforementioned estimation of the freezing index transmitted to the frost susceptible subgrade soil. An empirical alternative method for the estimation of the transmitted freezing index through the pavement

Calculation of Engineering Parameters structure is proposed by Corté et al. (1995). The transmitted freezing index can be estimated based on the following empirical relationship:  FI − b h s e FIt =    1 + ah 

2

(5-29)

where FIt = freezing index transmitted at the surface of the subgrade soil, FIs = freezing index at the surface of the pavement, h = total pavement thickness (cm), a, be = coefficients depending on the nature of pavement materials. If the pavement structure is constituted of a single material, then be is equal to b in Table 5-7. If the pavement structure is constituted of several layers, then the effective coefficient be can be obtained as follows: n

be =

∑ bi hi

(5-30)

i =1 n

∑ hi i =1

where a and bi are coefficients for materials in layer i obtained from Table 5-7 and hi is the thickness of layer i (cm). The method proposed by Corté et al. (1995) is simple and appealing. It is, however, based on the conditions in France and tends to give too low FIt values in cold climates. The method should be calibrated to local conditions before being used in frost depth predictions. The frost penetration into the subgrade soil can then be determined using the Stefan equation modified to include the effect of segregation on freezing through the incorporation of the segregation potential of the freezing soil: Xss =

2(k f − (SP × L)) Ls

× FIt

(5-31)

where Xss = depth reached by the frost front in the subgrade soil (m), k = thermal conductivity of the frozen soil (W/m·°C or J/s·m·°C), L = latent heat of fusion of water (334 MJ/m3), Ls = latent heat of fusion of the freezing soil and can be obtained from Eq. (5-24), and SP = segregation potential of the freezing soil, m 2/s·°C (see Example 5-10).

Material

a, cm−1

b, (°C·day)0.5/cm

Asphalt concrete Asphalt stabilized base

0.008

0.06

Granular material

0.008

0.10

Source: Corté et al. 1995.

TABLE 5-7 Experience.

Coefficients for Calculation of Transmitted Freezing Index Based on the French

229

230

Chapter Five Example 5-10 Given the characteristics of the pavement structure and the climatic conditions described in Example 5-9, estimate frost depth using the transmitted freezing index method. Solution The transmitted freezing index is equal to 508 . 2(°C ⋅ day) = 43 . 91 × 106 (°C ⋅ s) Step 1: Compute thermal characteristics of subgrade soil:   J k f = 0 . 001442 × 101. 373×1. 6 + 0 . 01226 × 20 × 100. 4994×1 . 6 = 1 . 77    s ⋅ m ⋅ °C  Ls = 1600 × 0 . 2 × 334 = 106, 8 80(kJ m 3 ) = 106 . 88 × 106 ( J m 3 ) SP = 100(mm 2 °C ⋅ day) = 100 × 10− 6 (m 2 °C ⋅ day) = 116 × 10− 11 (m 2 °C ⋅ s) Step 2: Compute frost depth. Based on the information provided and on soil thermal characteristics, frost depth can be obtained from Xss =

2(1 . 77 − (116 × 10−11 × 334 × 106 )) × 4 3 . 91 × 106 = 1 . 07(m) 106 . 88 × 106

The total frost depth from the pavement surface can be readily obtained by adding the total thickness of the pavement to Xss: X = (1 . 07 + 0 . 80)m = 1 . 93 m

5-7

Frost Heave Frost heave is the most important consequence of frost penetration in the pavement system. Frost heave causes stresses and distortions at the pavement surface. Water accumulated by the frost heave process is one of the important factors of bearing capacity loss during spring thaw. Several models are available to estimate frost heave in soils based on different physical theories. Methods based on the segregation potential theory (Konrad and Morgenstern 1980) remain the only practical methods for the prediction of frost heave based on relatively simple test methods. Two simple methods are described to estimate frost heave from the segregation potential of the freezing soil.

5-7-1

Konrad’s Method for Frost Heave Prediction

The method proposed by Konrad (2001) requires the following information: • The segregation potential of the subgrade soil • The average temperature gradient in the freezing soil • The duration of the freezing period in the subgrade soil The segregation potential can be obtained using one of the procedures described in Chap. 4. The average temperature gradient can be estimated using the simple procedure illustrated in Fig. 5-11. The temperature gradient, TG, is thus equal to TG =

Txf X−D

(5-32)

Calculation of Engineering Parameters

FIGURE 5-11

Determination of average thermal gradient [see Eq. (5-32)].

where T xf is the average temperature at the surface of the subgrade soil during the freezing period that can be obtained by dividing the freezing index transmitted to the subgrade soil by the duration of the freezing period at the subgrade soil level, X is the maximum frost depth, and D is the total thickness of the pavement structure. The duration of the freezing period in the subgrade soil, tfs, can be estimated based on Eqs. (5.5) and (5.6) using the procedure described in Example 5-5. The total amount of the frost heave, h, can then be estimated using the following equation: h = 1 . 09 ⋅ SP ⋅ TG ⋅ t fs

(5-33)

where SP = segregation potential, TG = thermal gradient is the estimated thermal gradient in the subgrade soil, and tfs = duration of the freezing period in the subgrade soil. Example 5-11 Given the information provided in Examples 5-9 and 5-10, estimate the frost heave in the pavement system. Solution Step 1: Compute relevant parameters: SP = 100(mm 2 °C ⋅ day) = 100 × 10− 6 (m 2 °C ⋅ day) Txf =

FIt 508 . 2 = = 3 . 97(°C) 128 . 1 t fs

TG =

3 . 97 = 3 . 7 (°C m) 1 . 87 − 0 . 80

Step 2: Compute frost heave: h = 1 . 09 × 100 × 10− 6 × 3 . 7 × 128 . 1 = 0 . 052(m)

231

232

Chapter Five

5-7-2

Saarelainen’s Method for Frost Heave Prediction

Saarelainen (1996) has observed that the ratio of frost heave on the thickness of the frozen soil is constant for a given soil and is proportional to the segregation potential of the soil following the relationship: h=

2 × SP(X − D)  X     FI s 

(5-34)

2

where h = average frost heave (m), SP = segregation potential (m2/°C·day), X = maximum frost depth (m), D = thickness of the pavement structure (m), and FIs = surface freezing index (°C·day) (see Example 5-12). Example 5-12 Given the information provided in Examples 5-9 and 5-10, estimate the frost heave using the Saarelainen method. Solution h=

2 × 100 × 10− 6 (1 . 87 − 0 . 80)  1 . 93     982 

2

= 0 . 056 m

5-8 Thaw Settlement Pavement engineers are confronted to two important questions in relation to pavement behavior when subjected to thawing. For seasonally frozen pavements, it is generally assumed that heaved pavements will settle back to their original elevation. The important question in those conditions is the duration of the consolidation period or time required for full recovery of pavement strength. In permafrost conditions, time is less of a concern but the amount of settlement that will occur as a result of thawing permafrost is a major concern for engineers. Ice-rich permafrost may become unstable as a result of changes in thermal regime caused by the construction of a pavement structure or caused by climatic changes. As a result, the pavement system will be subjected to progressive settlement which will take place over a period of several years. As illustrated in Fig. 5-12, a modification in the thermal balance at the surface of the pavement will cause thaw to penetrate deeper into the permafrost (light-grey zones in the thaw penetration bars). The resulting settlement of ice-rich permafrost will reduce the heat capacity and the latent heat of fusion of the newly thawed layer of soil increasing the thaw penetration in the subsequent year. The process will continue until a new equilibrium is reached. In conditions where surface conditions keep changing with time, thaw consolidation will keep evolving with time until the climatic trend changes or until a remedial action is applied to the pavement. Two methods can be used to estimate thaw settlement in pavements constructed on unstable permafrost. The first one uses the result of laboratory thaw-consolidation tests and the second one is based on measurements of soil density (Ladanyi 1996).

Calculation of Engineering Parameters

FIGURE 5-12 Change in thaw depth resulting from (a) a sudden warming of surface temperature and (b) a progressive warming of surface temperature.

The most accurate method to predict thaw settlement of ice-rich permafrost is to use results of thaw-consolidation tests done on undisturbed samples of frozen soil immediately beneath the active layer. As shown on Fig. 5-13a, most of the consolidation occurs during thawing and subsequent consolidation will occur as the effective stress increases in the thawed sample. The total settlement (sx) that will be experienced by the sample at

FIGURE 5-13 Typical results of a thaw-consolidation test of an ice-rich soil sample in (a) a void ratio-stress space and in (b) a settlement-stress space.

233

234

Chapter Five a given effective stress is the sum of the settlement caused by thawing (st) and of the subsequent consolidation (sc). A typical result of a thaw-consolidation test is shown on Fig. 5-13b. The relative settlement caused by soil thawing is A0 =

e f − et

(5-35)

1+ ef

where ef and et are frozen and thawed void ratios. The settlement related to thawing can be obtained by st = A0D f

(5-36)

And the settlement related to subsequent consolidation by sc = mvσ 'x D f

(5-37)

where mv is the coefficient of volume compressibility obtained from the consolidation test and Df is the thickness of the soil layer subjected to thawing on which an effective stress s¢x is acting. A quick evaluation of the potential thaw settlement of all soils within the thaw depth without t he need for thaw-consolidation tests can be obtained using soil densities (Ladanyi 1996):  ρdf  s = 1 − D f  ρd ,th 

(5-38)

where s is the thaw settlement, rdf and rd,th are the frozen and thawed dry densities of the soil, respectively (see Example 5-13). Example 5-13 Estimate the thaw settlement using Eq. (5-38) for a site with a thaw depth of 5.0 m in frozen silty clay with the specific gravity of soil solids of 2.70. The total water content is 58 percent with ice inclusions forming about one-third of the frozen volume and the unforzen water content is estimated to be 15 percent. Soil between the ice lenses had a saturated water content of 30 percent (adapted from Andersland and Ladanyi 2004). Solution Step 1: Compute the thawed dry density of soil between ice inclusions:

ρ d ,th =

ρw 1 + wsat Gs

=

1000 1 + 0 . 30 2.7

 kg  = 1492  3  m 

Step 2: Compute the frozen dry density:

ρ d , fr =

 kg  ρw 1000 = = 1011  3  m  1 1 + 1 . 09(w − wu ) + wu + 1 . 09(0 . 588 − 0 . 15) + 0 . 15 2.7 Gs

Step 3: Compute the thaw settlement using Eq. (5-38):  1011 s = 1 −  5 . 0 = 1 . 6(m)  1492 

Calculation of Engineering Parameters

5-9

Stresses and Strains in Pavements A simple way to calculate stresses, strains, and deflections in a soil mass is to use a set of equations developed in 1885 by French mathematician, Boussinesq. The equations, initially valid for a point load, were later refined and adapted for loads applied on flexible circular plates by Foster and Ahlvin (1954) and Ahlvin and Ulery (1962). Boussinesq’s solution is based on the following assumptions: • The soil volume is a semi-infinite space, it is infinite in the horizontal plane and in depth from the surface • The soil volume is constituted of elastic material characterized by an elastic modulus E (Young’s modulus) and Poisson’s coefficient m • The material properties are homogeneous and isotropic Based on these assumptions, vertical and radial stresses induced at any depth in the soil mass underneath the center of the circular loading plate can be obtained by: z3   σ z = σ 0 1 − 2 2 1.5   (a + z ) 

σr =

σ0  z3  2(1 + µ )z + 2 1 + 2µ − 2  2 0 5 . 2  (a + z ) ( a + z 2 )1. 5 

(5-39)

(5-40)

where s0 = stress uniformly applied on a plate of radius a at the surface of the soil mass, z = depth considered, m = Poisson’s coefficient that typically varies from 0.15 to 0.45. Table 5-8 gives typical values for the coefficient for several soils and pavement materials. Vertical and radial strains can be determined using Eqs. (5-41) and (5-42):

εz =

 z3 (1 + µ )q  2µ z − 2 1 − 2µ + 2  2 0 . 5 2 1 . 5 E  (a + z ) (a + z ) 

(5-41)

εr =

 z3 (1 + µ )q  2(1 − µ )z + 1 − 2µ − 2  2E  (a + z 2 )0 . 5 (a 2 + z 2 )1 . 5 

(5-42)

where E is the Young’s modulus and q is the vertical load on the ground surface. And finally, the vertical deflection can be determined as d=

 (1 + µ )qa  1 − 2µ 2 a + [(a + z 2 )0 . 5 − z]  2 2 0.5 E a ( a + z )  

(5-43)

More detailed information on the use of these equations, including graphical solutions, can be found in Ullidtz (1987). Solutions at any points in the soil mass have also been developed by Foster and Ahlvin (1954) and Ahlvin and Ulery (1962). These solutions and detailed information on stress and strain analysis in pavements can be found in Huang (2004).

235

236

Chapter Five Soil or Material

Range

Typical Value

Asphalt concrete

0.30–0.40

0.35

Portland cement concrete

0.15–0.20

0.15

Dense graded aggregates

0.30–0.40

0.35

Dense sand

0.30–0.45

0.35

Loose sand

0.20–0.40

0.30

Fine grained soils

0.30–0.50

0.40

Saturated clays

0.40–0.50

0.45

Source: Huang 2004 (reprinted by permission of Pearson Education).

TABLE 5-8

Typical Values of Poisson’s Coefficient for Different Soils and Pavement Materials

The main limitation of the Boussinesq model is that it is limited to a homogeneous soil mass infinite in depth and in the horizontal plane. Thus, it limits considerably the possibility to use the model in a multilayer system such as pavements. Burmister (1943a and b) was the first to propose a complete analytical solution for the calculation of stresses and displacement in a multilayer elastic system. The Burmister solution is, however, complex and can only be resolved through a series of calculation charts or using a computer. Most of the software codes currently available are based on the Burmister solution. In parallel to Burmister’s work a Swedish mathematician, Odemark, has developed a simple method to convert a layer of elastic material into an equivalent thickness of another material. The method is based on the principle that a material of low rigidity has less load distribution capacity than a more rigid material. As illustrated in Fig. 5-14, more thickness of material 1 will be required to have the same load distribution capacity than the material 2. The Odemark method, also known as the equivalent thickness method, is much easier to use than the Burmister solution. It is, however, more restrictive and based on additional assumptions: • The layer must have decreasing rigidity with depth with a minimum modulus ratio between two adjacent layers equal to two • The layers must have at least a thickness equal to the radius of the loading plate. A correction factor is proposed for cases where the thickness of the surfacing layer is smaller than the plate radius [see Eq. (5-44)] • All layers above the subgrade soil are considered to have a pure flexural behaviour and to have perfect friction at interfaces The method proposed by Odemark is the following: 1. When the analysis is conducted at a depth less than the depth of the first interface, stresses and strains can be computed with Eqs. (5-39) to (5-42) using the characteristics of the first layer.

Calculation of Engineering Parameters

FIGURE 5-14 Schematic illustration of the Odemark principle.

2. When the analysis is conducted at the first interface or between the first and the second interface, the first layer is transformed in an equivalent thickness, he, of the material in the second layer:

( (

E 1 − µ22 he = f h1  1 ×  E2 1 − µ12

) )

1/3

(5-44)

where h1 is the original thickness of layer 1 having a modulus E1 and Poisson’s coefficient m1, E2, and m2 are the elastic properties of layer 2, f is a correction factor: f = 1.0 for the first interface f = 1.1 for the first interface if the radius of the plate a > h1 f = 0.8 for all other cases 1. Stresses, strains, and deflections can be computer at the new depth “z” considering the equivalent thickness using the Boussinesq equations. 2. The same principle can be applied to other underlying layers. When two adjacent layers have the same Poisson’s coefficient, Eq. (5-44) can be written using the following simplified form: E  he = f h1  1   E2 

1/3

(5-45)

The Odemark/Boussinesq method is simple and accessible, yet it is accurate enough for most pavement engineering applications requiring an estimation of stresses and strains under a circular load. When the conditions of the method are respected, estimations done using the Odemark/Boussinesq method are generally in close agreement

237

238

Chapter Five with the results obtained using sophisticated analytical or numerical models (Ullidtz 2002). Moreover, the method can easily be programmed in a spreadsheet. Example 5-14 illustrates the use of the method. Example 5-14 Given the characteristics of the pavement described below, compute the horizontal strain at the bottom of the asphalt layer and the vertical strain at the top of the subgrade soil under a wheel applying a 560-kPa pressure on a circular area of 0.15 m radius.

E = 5,000 MPa, m = 0.35, D = 0.15 m E = 250 MPa, m = 0.35, D = 0.50 m E = 60 MPa, m = 0.35, D = ∞

Solution Step 1: Transformation of layer 1. Horizontal strain needs to be calculated at interface 1. According to Odemark’s method, the first layer needs to be converted in an equivalent thickness of material 2 using Eq. (5-45) (since m1 = m2). Being at the first interface, a correction factor of 1.0 is selected. 1/3

 5000 he = 1 . 0 × 0.15    250 

m = 0.407 m

Step 2: Computation of horizontal strain. A depth z of 0.407 m is used in the computation of strain using Eq. (5-42).

εr =

 2(1 − 0 . 35)0 . 407 (1 + 0 . 35)0 . 56  0 . 407 3 + 1 − 0 . 70 −  mm/mm 2 2 0 . 5 2 2(250) (0 . 15 + 0 . 407 ) (0 . 15 + 0 . 407 2 )1 . 5  

ε r = − 141 . 7 × 10− 6 mm/mm The negative sign indicates a tensile strain. Step 3: Transformation of the pavement in equivalent thickness of subgrade soil. The calculation of vertical strain at the second interface requires transforming the pavement layers (already transformed into equivalent thicknesses of material 2) into an equivalent thickness of material 3 using Eq. (5-45). Since the original pavement system is a multilayer system, a correction factor of 0.8 is used. 1/3

 250 he = 0 . 8(0 . 50 + 0 . 407 )    60 

m = 1 . 167 m

Step 4: Compute the vertical strain at the depth corresponding to the equivalent thickness of the transformed pavement structure using Eq. (5-41):

εz =

 2 × 0 . 35 × 1 . 167 (1 + 0 . 35)0 . 56  1 . 167 3 mm − 1 − 0 . 7 +  mm/m 2 2 0.5 2 2 1.5 60 (0 . 15 + 1 . 167 ) (0 . 15 + 1 . 167 )  

ε z = 234 . 2 × 10− 6 mm/mm

Calculation of Engineering Parameters

Review Questions 5-1. Determine the maximum cooling rate of air temperature for the following hourly temperature data set. Month

Day

Hour

Temperature, °C

3

15

0

−12.2

3

15

1

−13.3

3

15

2

−14.4

3

15

3

−15.5

3

15

4

−16.4

3

15

5

−17.3

3

15

6

−18.2

3

15

7

−18.3

3

15

8

−18.5

3

15

9

−18.6

3

15

10

−17.6

3

15

11

−16.6

3

15

12

−15.6

3

15

13

−14.9

3

15

14

−14.2

3

15

15

−13.5

3

15

16

−12.4

3

15

17

−11.4

3

15

18

−10.3

3

15

19

−10.0

3

15

20

−9.6

3

15

21

−9.3

3

15

22

−10.2

3

15

23

−11.1

3

15

24

−12.0

239

Chapter Five 5-2. Determine the air freezing index and the air thawing index of the following mean daily air temperature records.

Day 0

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 0 0

–1 –3

Temperature (°C)

–5

–4 –5

–6

–7

–10

–8

–9 –10 –12 –13 –13 –15

–10 –11 –12 –14

–15

–11 –13 –15

–14

–17

–20

–20 –22 –22 –23

–21 –25

–25–26

–30 (a)

20 17 15 Temperature (°C)

240

10

10 8 6 5

6

5

7

6

5

4

3 2

2

1 2 0

0 –2

–2 –5

7

–4

4 2 2

2

1 1

2

1

–3 –3

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 Day (b)

Calculation of Engineering Parameters 5-3. The air temperature can be estimated from a sinusoidal curve as shown in the figure below, for which parameters in Eq. (5-4) can be found. Using Eqs. (5-7) and (5-8), estimate the freezing and the thawing indices for the illustrated curve.

25 20

Temperature (°C)

15 10 5 0 –5 –10 –15 0

50

100

150

200 Day

250

300

350

400

5-4. Climatic measurements are made at points 1, 2, 3, and 4 (elevations are denoted by z for each point). Determine the freezing index at point 0 with the following mean daily temperatures:

z4 = 2020 m z1 = 1503 m .4

km

.7

16

37

km zo = 2030 m 22.

29 .3

km

8k

z2 = 1980 m

m

z3 = 2000 m

241

Chapter Five Temperature, °C

Days

Point 1

Point 2

Point 3

Point 4

1

−5.0

−8.0

−12.0

−10.0

2

−4.8

−7.3

−11.6

−9.5

3

−4.5

−7.0

−11.5

−8.8

4

−3.8

−6.9

−10.6

−8.6

5

−3.7

−6.1

−10.5

−8.1

6

−3.3

−5.5

−9.8

−7.2

7

−3.1

−5.1

−9.0

−6.2

8

−2.7

−5.1

−8.6

−5.8

9

−2.3

−4.3

−8.2

−5.7

10

−1.4

−3.8

−7.7

−5.5

11

−1.2

−3.3

−7.0

−5.5

12

−1.0

−3.1

−6.5

−4.8

13

−1.0

−3.1

−6.1

−4.0

14

−0.5

−2.6

−6.0

−3.3

15

0.0

−2.5

−5.2

−2.3

5-5. Determine the significant freezing index of the following data:

10 8

8

7 7 6

6

5

5

4

4 Temperature (°C)

242

2

2

2

2

2

0 –1

–2 –4 –6 –8

–2

–2 –3

–3 –3 –4

–4 –5

–5 –6 –7 –8

–3 –4

–5 –5 –6

–7 –8

–10 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 Day

Calculation of Engineering Parameters 5-6. A site characterized by an air freezing index of 1000°C·day is exposed to sunshine during an average of 2.5 h/day. Considering a pavement with an asphalt concrete surface, evaluate the surface freezing index at the surface of the pavement using the radiation index and the n-factor approaches.

5-7. Pavement surface temperatures and deflections were measured from 8:00 to 14:00 hours using an infrared temperature gauge and a falling weight deflectometer. The pavement (asphalt concrete layer thickness = 69 mm) has been in shade for 6 min during the whole test. The average air temperature the day before testing was 8°C. Determine the pavement temperature at 8:00 hours, at a depth of 48 mm.

Time, h

Measured Surface Temperature, °C

8:00

6.2

10:00

8.2

12:00

14.6

14:00

12.4

16:00

10.3

5-8. Estimate the average temperature of the top 100 mm of asphalt concrete on October 26th at noon. The air temperature is 5.2°C.

5-9. Evaluate the thermal conductivity of the following multilayer system: Compacted snow 0.18

0.05 0.10

Asphalt concrete Frozen layer

1.05

Granular material

r = 1.65 Mg/m w = 10%

Sand

r = 2.65 Mg/m w = 40%

0.78

Clay

r = 2.10 Mg/m w = 45%

0.57

5-10. A pavement has as the following structure: • 0.15 m of asphalt concrete • 1.00 m of granular base • 2.00 m of silt Will the top of the silt layer be below freezing ( −10

16

22

28

PG 64 34

40

>

>

>

>

−16

−22

−28

−34

−40

> −46

> −16

10

230

Viscosity, T316:b Max. 3 Pa, test temp (°C)

135

Dynamic shear, T315:c G∗/sinδ, min.

46

> −22

> −28

>

−34

> −40

>

−10

52

58

64

52

58

64

1.00 kPa, test temp @ 10 rad/s, (°C) Rolling thin-film oven test residue (T240) Mass changee, max., %

1.0

Dynamic shear, T315: G∗/sinδd, min.

46

2.20 kPa, test temp @ 10 rad/s, (°C)

22

28

34

40

> −16

>

>

−22

−28

> −34

−40

Pressure aging vessel residue (R 28) PAV aging temp., °Cf

90

Dynamic shear, T315:

10

7

4

25

22

19

16

13

10

7

25

22

19

16

13

31

28

25

22

19

16

−24

−30

−36

0

−6

−12

−18

−24

−30

−36

−6

−12

−18

−24

−30

0

−6

−12

−18

−24

−30

90

100

100

G*sinδd, max. 5000 kPa, test temp @ 10 rad/s, (°C) Critical low cracking temp., PP 42:g Determine critical cracking temp. as described in PP42, test temp, (°C)

PG 70 Performance grade

10

Average 7-day maximum pavement design temperature, °Ca

−10

16

230

Viscosity, T316:b Maximum 3 Pa·s, test temp, °C

135

TABLE 7-4

28

34

40

10

16

22

PG 82 28

34

−16

Original binder Flash point temp, T48, °C

22

PG 76

Example of Performance-Based Specifications

> −22

> −28

> −34

> −40

> −10

10

16

22

28

34

> −16

> −22

> −28

> −34

< 82 > −16

> −22

> −28

> −34

> −10

273

274 PG 70 Performance grade

10

Dynamic shear, T315:c G∗/sinδ, minimum 1.00 kPa, test temp @ 10 rad/s, °C

70

16

22

28

PG 76 34

40

10

16

22

PG 82 28

34

10

76

82

76

82

16

22

28

34

Rolling thin-film oven test residue (T240) Mass changee, maximum, %

1.0

Dynamic shear, T315: G∗/sinδd, minimum 2.20 kPa, test temp @ 10 rad/s, °C

70

Pressure aging vessel residue (R 28) PAV aging temperature, °Cf

100 (110)

Dynamic shear, T315: G∗sinδd, maximum 5000 kPa, test temp @ 10 rad/s, °C

34

31

28

25

22

19

37

34

31

28

25

Critical low cracking temperature, PP 42:g Determine critical cracking temp as described in PP42, test temp, °C

0

–6

–12

–18

–24

–30

0

–6

–12

–18

–24

100 (110)

100 (110) 40

0

37

34

31

28

–6

–12

–18

–24

a Pavement temperatures are estimated from air temperature using an algorithm contained in the LTPP Bind program, may be provided by the specifying agency, or by following the procedures as outlined in MP 2 and PP 28. b This requirement may be waived at the discretion of the specifying agency if the supplier warrants that the asphalt binder can be adequately pumped and mixed at temperatures that meet all applicable safety standards. c For quality control of unmodified asphalt binder production, measurement of the viscosity of the original asphalt cement may be used to supplement dynamic shear measurements of G∗/sinδ at test temperatures where the asphalt is a newtonian fluid. d ∗ G /sinδ = high temperature stiffness and G∗sinδ = intermediate temperature stiffness. e The mass change shall be less than 1.0 percent for either a positive (mass gain) or a negative (mass loss) mass change. f The PAV aging temperature is based on simulated climatic conditions and is one of three temperatures 90, 100, or 110°C. Normally the PAV aging temperature is 100°C for PG 58-xx and above. However, in desert climates, the PAV aging temperature for PG 70-xx and above may be specified as 110°C. g For verification of grade, perform T 313 at the test temperature and the test temperature minus 6°C and T 314 at the test temperature. Compare the failure stress from T 314 to the calculated induced thermal stress as per PP 42. If the failure stress exceeds the induced thermal stress, the asphalt binder is deemed a “PASS” at the specification temperature. Source: Table 1 from MP 1a in AASHTO 2003a, by the American Association of State Highway and Transportation Officials, Washington, D.C. Used by permission.

TABLE 7-4

Example of Performance-Based Specifications (Continued)

Mix Design Grade Test

PMB65

PMB75

PMB85

Penetration at 25°C, 1/10 mm

70–150

70–150

50–100

Softening point, °C

≥65.0

≥70.0

≥75.0

Flash point, °C

≥235

≥220

≥220

Elastic recovery at 10°C

≥75

≥75

≥75

Softening point difference top-bottom, °C

≤5.0∗

≤5.0∗

≤5.0∗

∗If the softening point difference is greater than 5°C, the binder supplier must provide recommendations on the handling necessary to prevent separation. Source: PANK 2000. Reprinted by permission.

TABLE 7-5

Example of Polymer-Modified Binder Specifications

Grade Test

B250/330

B330/430

B500/650

B650/900

Penetration at 15°C, 1/10 mm

70–130

90–170

140–260

180–360

Viscosity, 60°C, Pas

≥18

≥12

≥7.0

≥4.5

Viscosity, 60°C, mm2/s

V1500

V3000

1,000–2,000

2,000–4,000

Viscosity, 135°C, mm2/s

≥100

≥85

≥65

≥50

Fraass Breaking Point, °C

≤−16

≤−18

≤−20

≤−20

Flash Point, °C

≥180

≥180

≥180

≥180

≥160

≥160

Solubility in toluene, mass-%

≥99.0

≥99.0

≥99.0

≥99.0

≥99.0

≥99.0

Tests on residue from:

Rolling Thin Film Oven Test

Thin Film Oven Test

Mass loss, %

≤1.0

≤1.0

≤1.5

≤1.5

≤2.0

≤1.7

Viscosity ratio, 60°C

≤4.0

≤4.0

≤4.0

≤4.0

≤3.0

≤3.0

Source: PANK 2000. Reprinted by permission.

TABLE 7-6 Example of Specifications for Soft Asphalt Binders

its standard deviation is calculated for each station including data for all the years in operation. In the same way, a 1-day minimum air temperature of each year was identified, and the mean and standard deviations were calculated (Asphalt Institute 2001). The air temperatures are converted to pavement temperatures using Eqs. (7-1) and (7-2). Equation (7-1) is based on theoretical analyses of actual conditions performed with models for net heat flow and energy balance, and assuming typical values for solar absorption (0.9), radiation transmission through air (0.81),

275

276

Chapter Seven

FIGURE 7-4 Bitumen test data chart comparing two penetration grade asphalt cements and polymer-modified binder.

atmospheric radiation (0.7), and wind speed (4.5 m/s). It gives the high 7-day pavement design temperature at a depth of 20 mm below the pavement surface, T20mm (Asphalt Institute 2003): T20mm = (Tair – 0.00618 Lat2 + 0.2289 Lat + 42.2)0.9545°C – 17.78°C

(7-1)

where Tair is 7-day average high air temperature, °C and Lat is the geographical latitude of the project. Several methods have been proposed to convert the minimum air temperature to minimum pavement design temperature (Raad et al. 1997; Asphalt Institute 2003). For cold regions, it is generally agreed that the pavement temperature is warmer than the air temperature during a cooling trend when the low-temperature cracking occurs. Equation (7-2) developed by Canadian Strategic Highway Research Program (SHRP) researchers is used as a recommended conversion model for the minimum pavement design temperature, Tmin (Asphalt Institute 2001): Tmin = 0.859 Tair + 1.7°C

(7-2)

where Tair is minimum air temperature in average year, °C. The Superpave system allows the engineers to use reliability concepts to determine the degree of risk to the high and low pavement temperatures. The reliability is the percent probability in a single year that the actual pavement temperature will not exceed the design temperatures. This concept is illustrated in Example 7-1. Example 7-1 The mean high 7-day pavement design temperature is calculated to be 40°C with a standard deviation of 2°C. The mean low 1-day pavement design temperature is calculated to be −27°C with a standard deviation of 3°C. Select the performance grade using (a) 50 percent reliability and (b) 98 percent reliability.

Mix Design Solution (a) Select the appropriate grade from Table 7-4. The average temperatures of 40°C and −27°C represent the 50 percent reliability. However, these specific values are not listed in Table 7-4. In this case, the grade will be given using the principle of 6°C increments between the grades. Select PG 46-28. (b) Assume normal temperature distribution. The 98 percent confidence limits can then be found at the mean temperature plus 2 times the standard deviation: high temperature: (40 + 2⋅2)°C = 44°C; low temperature: (−27 − 2⋅3)°C = −33°C. Select PG 46-34.

The PG grades estimated by temperature considerations apply for typical highway loading conditions. If the traffic is standing, slow, or has extremely high volume, the high temperature grade needs to be increased to avoid permanent deformation (AASHTO MP2, 2003b): • Standing traffic; average speed