TECHNOLOGY/ENGINEERING/CIVIL
Geotechnical Engineering in Residual Soils digs deep to help enrich the reader’s knowledge...
270 downloads
3075 Views
2MB Size
Report
This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!
Report copyright / DMCA form
TECHNOLOGY/ENGINEERING/CIVIL
Geotechnical Engineering in Residual Soils digs deep to help enrich the reader’s knowledge on the subject of soils—in particular, residual soils—as they pertain to engineering. Appearing mostly in underdeveloped parts of the United States and tropical countries, these soils are playing an increasingly important role in building designs as construction encroaches into these areas. In recognition of this fact, this guide equips geotechnical engineers with essentials for learning the concepts and principles of residual soil behavior—and serves as a starting point to assist them in pursuing innovative engineering strategies for working effectively with residual soils. Geotechnical Engineering in Residual Soils: • Introduces geotechnical engineers to those aspects of residual soil behavior that they ought to be aware of when undertaking projects in these soils • Highlights the mistaken interpretations of soil behavior that can result from the application to residual soils of traditional concepts derived from sedimentary soils • Includes numerous illustrations throughout, specifically addressing the unique properties of residual soils • Includes coverage of special topics, such as the role of negative pore pressure above the water table, the influence of weather conditions on soil behavior, the properties of volcanic soils, and compaction of residual soils • Is written by an author with more than thirty years of firsthand experience analyzing and designing for construction on residual soils Thorough and insightful, Geotechnical Engineering in Residual Soils delivers a fresh overview on understanding the structural and mechanical properties of soils from an engineering perspective—and informs readers how to solidify design approaches to set their projects on a sure footing. LAURENCE D. WESLEY worked as a practicing geotechnical engineer for more than thirty years, with experience in New Zealand, Australia, Indonesia, Malaysia, and Bahrain. He is a Lifetime Member of the American Society of Civil Engineers, and a retired senior lecturer in geotechnical engineering at the University of Auckland. Cover Art © Istockphoto.com/Alejandro Raymond | Cover Design: Holly Wittenberg
Geotechnical Engineering in Residual Soils
The pioneering guide that breaks ground on the unique engineering properties of residual soils
Wesley
Geotechnical Engineering in Residual Soils
Laurence D. Wesley
GEOTECHNICAL ENGINEERING IN RESIDUAL SOILS
GEOTECHNICAL ENGINEERING IN RESIDUAL SOILS
Laurence D. Wesley
JOHN WILEY & SONS, INC.
This book is printed on acid-free paper. Copyright © 2010 by John Wiley & Sons, Inc. All rights reserved Published by John Wiley & Sons, Inc., Hoboken, New Jersey Published simultaneously in Canada No part of this publication may be reproduced, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical, photocopying, recording, scanning, or otherwise, except as permitted under Section 107 or 108 of the 1976 United States Copyright Act, without either the prior written permission of the Publisher, or authorization through payment of the appropriate per-copy fee to the Copyright Clearance Center, 222 Rosewood Drive, Danvers, MA 01923, (978) 750-8400, fax (978) 646-8600, or on the web at www.copyright.com. Requests to the Publisher for permission should be addressed to the Permissions Department, John Wiley & Sons, Inc., 111 River Street, Hoboken, NJ 07030, (201) 748-6011, fax (201) 748-6008, or online at www.wiley.com/go/ permissions. Limit of Liability/Disclaimer of Warranty: While the publisher and the author have used their best efforts in preparing this book, they make no representations or warranties with respect to the accuracy or completeness of the contents of this book and specifically disclaim any implied warranties of merchantability or fitness for a particular purpose. No warranty may be created or extended by sales representatives or written sales materials. The advice and strategies contained herein may not be suitable for your situation. You should consult with a professional where appropriate. Neither the publisher nor the author shall be liable for any loss of profit or any other commercial damages, including but not limited to special, incidental, consequential, or other damages. For general information about our other products and services, please contact our Customer Care Department within the United States at (800) 762-2974, outside the United States at (317) 572-3993 or fax (317) 572-4002. Wiley also publishes its books in a variety of electronic formats. Some content that appears in print may not be available in electronic books. For more information about Wiley products, visit our web site at www.wiley.com. Library of Congress Cataloging-in-Publication Data: Wesley, Laurence D. Geotechnical engineering in residual soils / Laurence D. Wesley. p. cm. Summary: “Wiley has long held a pre-eminent position as a publisher of books on geotechnical engineering, with a particular strength in soil behavior and soil mechanics, at both the academic and professional level. This reference will be the first book focused entirely on the unique engineering properties of residual soil. Given the predominance of residual soils in the under-developed parts of the United States and the Southern Hemisphere, and the increasing rate of new construction in these regions, the understanding of residual soils is expected to increase in importance in the coming years. This book will be written for the practicing geotechnical engineer working to any degree with residual soils. It will describe the unique properties of residual soil and provide innovative design techniques for building on it safely. The author will draw on his 30 years of practical experience as a practicing geotechnical engineer, imbuing the work with real world examples and practice problems influenced by his work in South America and Southeast Asia”—Provided by publisher. Includes bibliographical references and index. ISBN 978-0-470-37627-0 (acid-free paper); ISBN 978-0-470-64436-2 (ebk); ISBN 978-0-470-64437-9 (ebk); ISBN 978-0-470-64438-6 (ebk) 1. Residual materials (Geology) 2. Engineering geology. I. Title. TA709.5.W475 2010 624.1 51– dc22 2010016879 Printed in the United States of America 10 9 8 7 6 5 4 3 2 1
CONTENTS
PREFACE AND ACKNOWLEDGMENTS
xiii
1 FUNDAMENTAL ASPECTS OF RESIDUAL SOIL BEHAVIOR
1
1.1 1.2
Introduction / 1 Formation Processes and Basic Difference between Residual and Sedimentary Soils / 2 1.3 Structure of Residual Soils / 5 1.4 Special Clay Minerals / 7 1.5 The Influence of Topography / 8 1.6 Geotechnical Analysis, Design, and the Role of Observation and Judgment / 9 1.7 Summary of Basic Differences between Residual and Sedimentary Soils / 11 References / 12 2 EVALUATION, CHARACTERIZATION, AND CLASSIFICATION OF RESIDUAL SOILS
2.1 2.2 2.3 2.4 2.5
13
Introduction / 13 Parent Rock and the Soil Profile / 13 Influence of Parent Rock on Geotechnical Properties / 15 The Role of Observation / 16 Standard Index Tests / 17 2.5.1 Particle Size / 18 2.5.2 Atterberg Limits / 19 v
vi
CONTENTS
2.5.3 “Compactness” Indexes / 20 Classification Systems for Residual Soils / 21 2.6.1 Introduction / 21 2.6.2 Methods Based on Pedological Groups / 22 2.6.3 Methods Intended for Specific Local Use / 24 2.6.4 A Grouping System Based on Mineralogy and Structure / 24 References / 33 2.6
3 PORE PRESSURES AND SEEPAGE CONDITIONS ABOVE AND BELOW THE WATER TABLE
3.1 3.2
35
Introduction / 35 Situation at Level Sites / 36 3.2.1 Static Case / 36 3.2.2 Seasonal Effects / 37 3.2.3 Coarse-Grained Soils / 37 3.2.4 Low-Permeability Clays / 37 3.2.5 Medium- to High-Permeability Clays / 39 3.2.6 Use of Terzaghi Consolidation Theory to Illustrate Seasonal Influence / 40 3.2.7 Field Records of Seasonal Effects / 42 3.3 Hill Slopes, Seepage, and Pore Pressures / 44 3.4 Permeability of Residual Soils / 47 3.5 Significance of the Water Table (or Phreatic Surface) / 48 3.6 Implications of the Groundwater and Seepage State above the Water Table for Practical Situations / 48 3.6.1 Errors in the Estimation of Foundation Settlement Using Conventional Methods / 48 3.6.2 Ground Settlement Resulting from Groundwater Lowering / 49 3.6.3 Ground Settlement or Swelling Due to Covering the Ground Surface / 50 3.6.4 Errors in Estimates of Slope Stability Ignoring Soil Suction Influence / 51 3.6.5 Errors in Estimates of Slope Stability because of Simplified Assumptions Regarding the Seepage Pattern in the Slope / 51 References / 51
vii
CONTENTS
4 CONSOLIDATION AND SETTLEMENT
53
4.1 4.2
Introduction / 53 Interpretation of Standard Oedometer Test Results and the “Omnipotence of Tradition” / 54 4.3 Behavior of Residual Soils / 57 4.3.1 Tropical Red Clay / 57 4.3.2 Piedmont Residual Soil / 58 4.3.3 Waitemata Residual Clay / 60 4.3.4 Volcanic Ash (Allophane) Soils / 62 4.3.5 Summary of Principal Aspects of Compression Behavior of Residual Soils / 63 4.4 Consolidation Behavior after Remolding / 65 4.5 Values of Stiffness Parameters for Residual Soils / 67 4.6 Time Rate and Estimation of the Coefficient of Consolidation / 67 4.7 Rate of Consolidation for Surface Foundations on Deep Soil Layers / 73 4.8 Examples of Settlement Estimates / 76 4.8.1 Foundations for a Multistory Building on Red Clay / 76 4.8.2 Settlement Estimate Involving Nonlinear Compressibility and Pore Pressure Influence / 82 4.8.3 Significance of Time Rate Assumption in the Previous Example / 95 4.9 Accuracy of Settlement Estimates Based on Oedometer Tests / 96 4.9.1 A Common Source of Error Arising from the Use of the Log Parameter (Cs ) / 96 4.9.2 Actual Settlement versus Predictions / 98 4.10 Allowable Differential Settlement for Surface Foundations on Residual Soil / 98 References / 99 5 SHEAR STRENGTH OF RESIDUAL SOILS
5.1 5.2 5.3
Introduction / 101 Undrained Shear Strength / 102 Effective Strength Properties / 103
101
viii
CONTENTS
5.3.1 5.3.2 5.3.3 5.3.4 5.3.5 5.3.6 References /
Influence of Discontinuities / 104 Correlation between φ Value and the Atterberg Limits / 105 Effective Strength Parameters of a Residual Soil Derived from Shale / 106 Stress–Strain Behavior in Triaxial Tests / 107 The Cohesion Intercept c / 107 Residual Strength / 112 114
6 SITE INVESTIGATIONS AND THE MEASUREMENT OF SOIL PROPERTIES
115
6.1 6.2 6.3 6.4 6.5
Introduction / 115 Approaches to Site Investigations / 116 Organizational and Administrative Arrangements / 116 Planning Site Investigations / 118 Field Work / 119 6.5.1 Hand Auger Boreholes / 119 6.5.2 Machine Boreholes / 120 6.5.3 Penetrometer Testing / 121 6.6 Block Sampling / 122 6.7 In Situ Shear Tests / 124 6.8 Laboratory Testing / 126 6.8.1 Index or Classification Tests / 126 6.8.2 Tests on Undisturbed Samples / 126 6.8.3 “Computer Errors” in Processing Laboratory Test Results / 128 6.9 Correlations with Other Properties and Parameters / 129 6.9.1 Undrained Shear Strength / 130 6.9.2 Relative Density of Sand / 132 References / 133 7 BEARING CAPACITY AND EARTH PRESSURES
7.1 7.2 7.3
135
Introduction / 135 Bearing Capacity and Foundation Design / 136 Earth Pressure and Retaining Wall Design / 139 7.3.1 Earth Pressure to Retain Cuts in Steep Slopes / 139 7.3.2 The Use of Residual Soils for Reinforced Earth Construction / 146 References / 150
CONTENTS
8 SLOPE STABILITY AND SLOPE ENGINEERING
8.1 8.2 8.3 8.4 8.5
8.6
8.7 8.8 8.9 8.10
8.11
ix
151
Introduction / 151 Failure Modes / 152 The Place of Analytical and Nonanalytical Methods for Assessing the Stability of Natural Slopes / 152 Application and Limitations of Analytical Methods / 154 Uncertainties in Material Properties / 154 8.5.1 Slopes Consisting of Uniform, Homogeneous Materials / 154 8.5.2 Slopes Containing Distinct, Continuous Planes of Weakness / 155 8.5.3 Slopes of Heterogeneous Material, but without Distinct Planes of Weakness / 156 Uncertainties in the Seepage and Pore Pressure State / 156 8.6.1 Influence of Climate and Weather / 156 8.6.2 Response of Seepage State and Pore Pressure to Rainfall / 157 8.6.3 Comparison with Sedimentary Soils / 159 The Worst-Case Assumption Regarding the Water Table / 159 Transient Analysis of Rainfall Influence on the Stability of a Homogeneous Clay Slope / 164 Modeling Stability Changes Resulting from Varying Rainfall Intensities / 169 The Hong Kong Situation / 172 8.10.1 Measurements of Pore Pressure Response / 173 8.10.2 The Wetting Front Method for Estimating Water Table Rise / 176 8.10.3 Importance of Antecedent Rainfall / 177 8.10.4 Results of Stability Analysis and Assumptions Regarding the Pore Pressure State / 177 8.10.5 Recommended Safety Factors for Hong Kong Slopes / 178 8.10.6 Triaxial Tests and Back-Analysis of Landslides / 178 8.10.7 Concluding Remarks on the Hong Kong Situation / 179 Back-Analysis Methods to Determine Soil Parameters / 180 8.11.1 Back-Analysis of a Single Slip or a Single Intact Slope / 180
x
CONTENTS
8.11.2
Analysis of a Number of Slips in the Same Material / 181 8.11.3 Analysis of a Large Number of Intact Slopes (No Previous Slips) / 183 8.12 Slope Design / 184 8.12.1 Selection of the Profile for a New Cut Slope / 184 8.12.2 To Bench or Not to Bench a Slope? / 186 8.12.3 A Note on Vegetation Cover on Slopes / 187 References / 187 9 VOLCANIC SOILS
9.1 9.2
189
Introduction and General Observations / 189 Allophane Clays / 189 9.2.1 Performance of Natural Hill and Mountain Slopes / 190 9.2.2 Formation of Allophane Clays / 190 9.2.3 Structure of Allophane Clays / 192 9.2.4 Particle Size / 193 9.2.5 Natural Water Content, Void Ratio, and Atterberg Limits / 193 9.2.6 Influence of Drying / 194 9.2.7 Degree of Saturation, Liquidity Index, and Sensitivity / 195 9.2.8 Identification of Allophane Clays / 197 9.2.9 Compressibility and Consolidation Characteristics / 197 9.2.10 Strength Characteristics / 200 9.2.11 Compaction Behavior / 203 9.2.12 Engineering Projects Involving Allophane Clays / 205 9.3 Volcanic Ash Clays Derived from Rhyolitic Parent Material / 205 9.4 Other Unusual Clays of Volcanic Origin / 210 9.5 Pumiceous Materials / 213 9.5.1 Pumice Sands / 213 9.5.2 Pumiceous Silts and Gravels / 217 References / 221
CONTENTS
10 RESIDUAL SOILS NOT DERIVED FROM VOLCANIC MATERIAL
xi
223
10.1 10.2 10.3
Introduction / 223 Weathered Granite (Group 1 in Figure 10.1) / 223 Weathered Sedimentary Rocks / 226 10.3.1 Soft Rocks—Sandstones, Mudstones, and Shale (Group 3 in Figure 10.1) / 226 10.3.2 Hard Sedimentary Rocks (Group 4 in Figure 10.1) / 227 10.4 Laterites and Tropical Red Clays (Group 5 in Figure 10.1) / 228 10.5 Black or Black Cotton Clays / 230 References / 232 11 COMPACTION OF RESIDUAL SOILS
233
11.1 11.2
Introduction / 233 Some Reflections on Compaction Behavior of Soils and Quality Control Methods / 235 11.3 Optimum Compactive Effort as well as Optimum Water Content / 236 11.4 Alternative Compaction Control Based on Undrained Shear Strength and Air Voids / 237 11.5 The Use of Shear Strength to Overcome Difficulties in Compacting Residual Soils / 241 11.5.1 Soils That Contain Wide and Random Variations in Properties / 241 11.5.2 Nonsensitive Soils Considerably Wetter than Optimum Water Content / 241 11.5.3 Sensitive, Highly Structured Soils / 242 11.6 Hard, Partially Weathered, Residual Soils / 243 References / 243
INDEX
245
PREFACE AND ACKNOWLEDGMENTS
This book is, in many ways, an addendum, or an appendix to my earlier book Fundamentals of Soil Mechanics for Sedimentary and Residual Soils. There were a number of things I wanted to say about geotechnical engineering in residual soils that would have been out of place in that book, so a separate book seemed the most appropriate place for saying them. This book should perhaps also have had a subtitle and been called: Geotechnical Engineering in Residual Soils—A Personal View, because it does tend to be that. However, I have attempted to produce a stand-alone book and to make it reasonably balanced in terms of its coverage of all residual soil types. This means there is considerable overlap with the content of my earlier book. The book is intended primarily for those who are already familiar with the subject of soil mechanics and have some experience of practical geotechnical engineering, but not a lot of knowledge of residual soils. At the same time I hope it will also be of value to those already familiar with residual soils. The story behind Fundamentals of Soil Mechanics for Sedimentary and Residual Soils and this book is a long one, and (at the risk of boring the reader), I will recount it briefly. Along the way I will acknowledge my indebtedness to several specific organizations, institutions, and individuals who have had an important association with the history of the two books. The story starts with my participation in a program known as the Volunteer Graduate Scheme, which was organized by the New Zealand University Students’ Association in the late 1950s, with the support of both the New Zealand and Indonesian governments. That scheme was a copy of the Australian Volunteer Graduate Scheme, which started in the early 1950s, and both were forerunners to such schemes as the United States’ Peace Corps, and the United Kingdom’s Volunteer Service Overseas. Under the auspices xiii
xiv
PREFACE AND ACKNOWLEDGMENTS
of that scheme I went to Indonesia in 1960 to work for the Indonesian government in its Institute for Soil and Highway Investigations in Bandung, West Java. That was where I first encountered residual soils, especially those of volcanic origin. I am indebted to the individuals who put together the Volunteer Graduate Scheme, and, while I cannot mention them all by name, I especially want to acknowledge Hugh Templeton and Ted Woodfield, who were key figures in its establishment. The eight years (two terms of four years) I spent at the Bandung Institute, now the Highway Research Centre, were very enjoyable and, I hope, fruitful ones. I owe a debt of gratitude to my colleagues there, including the directors, engineers, and technicians, who made me welcome and from whom I learned a great deal. A number of them are still good friends to this day. This book, and my earlier book, would not have come into existence had it not been for my time in Indonesia. In fact, the book Fundamentals of Soil Mechanics for Sedimentary and Residual Soils started life as a rewrite of a basic soil mechanics textbook that I wrote (in Indonesian) in Bandung in 1972, near the end of my second term there. In this connection I wish to record my indebtedness to Ir Luthfie, an Indonesian structural engineer and a long time friend. He urged me to put the teaching material I was using at the time for geotechnical courses into a form that would be suitable for publication as a textbook. With some reluctance I did this and left my draft of the book with him when I left the country. He arranged to have the book published, and it was a best seller in the 1970s and 1980s, being the first soil mechanics textbook to appear in the Indonesian language. I have been asked at various times during subsequent visits to Indonesia when a new edition of the book would be coming out, and put it high on my list of things to do when I retired. By the time I started writing the new edition, I had the feeling that the book could well serve a useful need in the wider world and that an English edition of the book would not be out or place. In due course, my proposals for Fundamentals of Soil Mechanics and the current book were accepted for publication by John Wiley & Sons. I would like it to be known also that the idea of writing the present book actually arose out of a postgraduate course that I have been teaching in recent years as an invited lecturer at the University of Chile in Santiago. The title of this book (Geotechnical Engineering in Residual Soils) is the same as that course, and there is much in common between the material in this book and that course. However, just as preparing material for the lecture course forced me to organize my thoughts and understanding of residual soils and present them in an orderly fashion, so writing this book has motivated me to do some further rethinking, reorganizing, and rewriting of the material. Former students will therefore recognize many similarities between the book and courses I have given in the past, but they may also find some differences and additions. I am extremely grateful to Ramon Verdugo and the geotechnical group in the Civil Engineering Department
PREFACE AND ACKNOWLEDGMENTS
xv
of the University of Chile for inviting me to Santiago, not just once but, to date, five times. My time at their department has been very enjoyable and stimulating, and, as indicated above, this book would not have come into existence were it not for my time there. Writing a textbook on residual soils is an ambitious and, indeed, a rather pretentious undertaking, because the variety of soil types and the diversity of behavior found within them are such that no book can do justice to the subject. As stated in the opening chapter of this book, residual soils are “raw material”—an unkempt, unprocessed, and unsorted group. Many of them are neither rock nor soil—they hover somewhere in between. There is no clear conceptual or theoretical framework that we can use to sort them into tidy groups. For these reasons, I have largely avoided getting into minute detail about the properties of particular soil types; instead, I have tried to highlight those aspects of residual soil behavior that are common to the whole group and distinctly different from sedimentary soils. If this all sounds rather like an apology or an excuse for the shortcomings of the book, then I will not mind if readers see it that way. In addition to the people and organizations mentioned above, I am indebted to many others. I have expressed my thanks to them in my previous book, and invite the reader to refer to the acknowledgments section of that book. However, I want to especially mention Yolanda Thorp of URS Consultants in Auckland, and Mike Dobie of Tensar International in Jakarta, who supplied me with much useful information on reinforced earth walls using clay fill. I must also repeat my thanks to my wife, Barbara, and my children, especially Kay, for their extended tolerance and support during the time I have spent writing this second book. Finally, I want to sincerely thank John Wiley & Sons for publishing these books and to say it has been a pleasure to work with their staff, especially Jim Harper and Dan Magers.
GEOTECHNICAL ENGINEERING IN RESIDUAL SOILS
CHAPTER 1
FUNDAMENTAL ASPECTS OF RESIDUAL SOIL BEHAVIOR
1.1
INTRODUCTION
My main objective in writing this book is to provide geotechnical engineers with some basic guidelines that may be helpful to them when working in residual soils. I hope it will be especially useful to those encountering residual soils for the first time. The book should also be of value to students wishing to further their basic understanding of residual soil behavior. It is not my intention to give detailed descriptions of the many types of residual soils found on the planet, and thus provide a sort of handbook that engineers could refer to when encountering any particular residual soil type. Rather I will try to identify and explain the basic aspects of soil behavior that are specific to residual soils, and that geotechnical engineers working in residual soils should be aware of. There is one exception to the omission of detailed descriptions of specific residual soil types, and that is volcanic soils. These have the most distinctive and highly unusual properties, and are a soil group in which the author has particular experience. For these reasons, one whole chapter (Chapter 9) is devoted to these soils. The extent to which residual soils differ from sedimentary soils is a matter of some debate. On the one hand, it can be argued that the most basic principles of soil mechanics are equally applicable to both residual and sedimentary soils. In particular, the principle of effective stress and the Mohr-Coulomb failure criteria are of universal applicability. On the other hand, it can be argued that fundamental differences in the way the soils
1
2
FUNDAMENTAL ASPECTS OF RESIDUAL SOIL BEHAVIOR
are formed mean that residual soils are a “class apart” and must be treated differently. For example, Vaughan (1985) states: The development of the “classical” concepts of soil mechanics (in which soil properties are related and classified according to index properties, plasticity, stress history and the like) has been based almost exclusively on the investigation of sedimentary deposits of un-weathered soil. These concepts have been found almost universally inapplicable to the behaviour of residual soils, and misleading if inadvertently applied.
There is considerable truth in both of the above positions, as should become apparent from the material presented in this book. The first and most important point to appreciate is that there is indeed a fundamental difference in the way residual and sedimentary soils are formed and, at the risk of tedious repetition of material that most readers may already be familiar with, I will consider formation processes in some detail in the next section. I should mention at this point that my experience in geotechnical engineering in residual soils is largely limited to the wet tropics of Southeast Asia, in particular, Malaysia and Indonesia, and the temperate climate of New Zealand. The contents of this book are thus strongly influenced by the soil conditions in these countries, which are predominately moderate to high plasticity clays. This fact, along with the wet climate, means that the soils are generally fully saturated, or sufficiently close to full saturation that for practical engineering purposes they can be assumed to be so. In this environment, only a shallow zone at the surface is likely to experience partial saturation, caused by evaporation during dry weather, and not because water drains out of the soil under the influence of gravity. This book, therefore, is primarily about fully saturated clays of moderate to high plasticity. 1.2 FORMATION PROCESSES AND BASIC DIFFERENCE BETWEEN RESIDUAL AND SEDIMENTARY SOILS
The basic processes by which soils are formed are illustrated in simplified form in Figure 1.1. Residual soils are formed directly by the physical and chemical weathering of the rock underlying them. Sedimentary soils undergo additional processes; the residual soil is eroded by rainfall and then transported by streams and rivers to be deposited in lakes or the sea as indicated in Figure 1.1. The soil thickness steadily grows as deposition continues; at the same time, it undergoes consolidation from its self-weight. With time, the soil may experience uplift as a result of tectonic movement and end up on dry land, where the erosion cycle will start all over again.
RESIDUAL AND SEDIMENTARY SOILS
3
Residual Soil Produced by physical and chemical weathering of underlying rock. Erosion by rainfall and run-off Soil
Transport by stream and river
Redeposition in layers in lakes or the ocean
Delta deposits Rock
Sea or lake level
Sedimentary Soil Later tectonic movement may raise this above sea level.
Figure 1.1 Formation of sedimentary and residual soils.
Figure 1.2 is a further attempt to illustrate the fundamental difference in the way the two soil groups are formed, and to help identify the factors that govern their properties. The formation process tends to influence their properties in an opposite manner. With residual soils, the weathering process normally converts solid rock into small particles and clay minerals, inevitably making the material less dense and weaker. With sedimentary soils, the compression of the soil from the weight of material above it, together with aging effects (to be described in the next section) makes it denser and harder. Two significant differences between residual and sedimentary soils become apparent from the above account of their formation: 1. Sedimentary soils undergo a systematic sorting process during erosion, transportation, and deposition. Finer particles are separated from coarse particles and are deposited in different locations or layers. Sedimentary soils, therefore, tend to be reasonably homogeneous. Residual soils do not undergo these processes, and are likely to be much more heterogeneous than sedimentary soils. 2. The concepts of stress history, normal consolidation, and overconsolidation have no relevance to residual soils. There is no such thing as the virgin consolidation line of a residual soil, a fact that is not always appreciated by those investigating their properties. The “virginal state” of a residual soil is the parent rock from which it is formed, not a soft sediment at the bottom of the sea or a lake (as is the case with sedimentary soils).
4
FUNDAMENTAL ASPECTS OF RESIDUAL SOIL BEHAVIOR
Voids Voids Solids Physical and chemical weathering Solids changes the solid matter and greatly increases the void space. Residual soil
Parent rock (little or no voids)
(substantial void space) (a) Residual Soil
Pressure
Void ratio (=
Vol. voids ) Vol.solids
A
At deposition, the stress on the soil is negligible; it is thus very soft with a high voids content.
Voids Solids
Condition at deposition (Point A)
Continuing deposition increases the pressure on the soil, causing it to compress.This reduces the voids content and increases the strength. C
Point B
B Uplift and erosion may reduce pressure on the soil, allowing it to swell slightly. Aging and hardening may make the soil stronger.
Voids Solids
Point C
Eventual condition (Point B and C)
(b) Sedimentary Soil
Figure 1.2 Another portrayal of the formation of residual and sedimentary soils.
The above factors mean there is a degree of homogeneity and predictability with sedimentary soils that is absent from residual soils. The convenient behavioral framework whereby the properties of sedimentary soils are related to stress history and divided into normally consolidated and overconsolidated soils cannot be applied to residual soils. Readers may find it helpful to think of residual soils as a raw, unkempt, and unpredictable group, which (though not generally ill behaved) lack the tidy, refined behavior of sedimentary soils, which have been through a proper “finishing” school. These differences between the two soil groups are elementary and accepted by the geotechnical fraternity, at least to the extent that residual soils are regarded as a separate group and given special treatment in the form of conferences, symposia, and books devoted specifically to them. However, despite this recognition as being different from sedimentary soils, residual soils still tend to be investigated and evaluated as though they are sedimentary soils. In this respect there is a good deal of truth in the statement of Vaughan quoted above. The most striking example of
STRUCTURE OF RESIDUAL SOILS
5
this is the continuing interpretation of standard oedometer tests using a framework developed from sedimentary soils. Graphs are plotted using the e-log p format and are routinely interpreted wrongly in terms of a preconsolidation pressure separating a virgin consolidation line from an unloading– reloading line. This issue is described in some detail in the author’s earlier book (Wesley 2009) and is discussed in further detail in Chapter 4 of the present book. 1.3
STRUCTURE OF RESIDUAL SOILS
The term structure is widely used in soil mechanics, although not always with the same meaning. In the early days of the subject it appears to have been used mainly to describe those features of the soil that are clearly visible to the naked eye, such as bedding planes, joints, fault discontinuities, and root holes. These features are best termed macrostructure. In more recent times structure has been used more specifically to designate the way in which the particles are arranged to form the soil skeleton itself. A highly structured soil is one in which the particles are arranged or even bonded together in such a way that the soil skeleton has characteristics quite different from those of a simple collection of individual particles. This kind of structure cannot be seen with the naked eye, and is termed microstructure. The term will be used in this book primarily to designate microstructure. The existence of microstructure in soils has long been recognized, in both sedimentary and residual soils; indeed, it is evident from the fact that nearly all natural soils have some sensitivity. At the same time, its importance in influencing soil behavior seems to have been rather lost sight of, or displaced, in favor of the stress history model of soil behavior. In recent years, however, the influence of structure has been increasingly recognized, in both sedimentary and residual soils. It is recognized, for example, that very few soft sedimentary clays that are “normally consolidated” geologically actually behave as normally consolidated soil. They behave as lightly overconsolidated clays due to a steady increase in strength with time after their deposition. The terms aging or hardening are being increasingly used to describe this effect. It is recognized also that structure may play a significant role even in the behavior of stiff sedimentary clays. For example, Gasparre et al. (2007) give an account of the influence of structure on London clay. Many residual soils, but certainly not all, are highly microstructured, and various conceptual pictures of their structure have been put forward. Several examples are shown in Figure 1.3. The first diagram, Figure 1.3a, for a “normal” undisturbed clay, shows an array of plate-like clay particles occupying void space between the coarser silt or fine sand particles. Soils with such an arrangement of particles may be relatively insensitive, indicating
6
FUNDAMENTAL ASPECTS OF RESIDUAL SOIL BEHAVIOR
Plate-like clay particles
Silt or fine sand particles
(a) “Normal” Clay
Bonding at contacts
Silt or fine sand particles
(b) Cemented Structure
Weak skeleton
Void space
(c) Honeycomb Structure
Figure 1.3 Conceptual pictures of soil microstructure.
that the influence of structure is not great, or they may be highly sensitive, as in the case of “quick” clays, indicating that they are highly structured. Figure 1.3b shows a common concept of the structure of cemented or bonded soils, particularly residual soils. This is a useful concept, and artificially bonded soils have been created and used in laboratory studies of soil behavior, in the belief that this is a reasonable representation of residual soil behavior. Examples are Maccarini (1987) and Toll et al. (2006). These are valuable studies, but we should note their limitations, especially with respect to residual soils. The weathering process, at least in igneous rocks and other hard rocks, is normally one that weakens the rock by breaking it up and converting rock minerals into clay minerals, not one that cements together existing hard particles. Thus, the first concept above, Figure 1.3a, may be just as valid for some residual soils as Figure 1.3b. Some particular weathering processes may still produce the cemented structure of Figure 1.3b. Finally, Figure 1.3c shows a honeycomb structure, which consists of a skeleton of relatively weak material with very large void space. The honeycomb material may be a single material or may be concentrations or aggregations of particles. This honeycomb structure appears to be valid for many sensitive or highly sensitive volcanic soils. We shall see later that the weathering of siltstones or mudstones (shales) may be different again from the above concepts, and may involve the solution of bonding material and the release of preexisting clay minerals, rather than the creation of new clay minerals or bonds between particles. The concepts illustrated in Figure 1.3 are likely to be gross oversimplifications of the true situation. They are included here to give the reader an indication of how soil structure can be visualized. The most essential point to be appreciated with respect to soil structure is that compression of the soil does not just involve pressing the particles into a tighter arrangement.
SPECIAL CLAY MINERALS
7
It also involves destroying the natural structure of the soil, and in effect producing a new material. Compression of structured soils is thus also a form of remolding the soil. The point at which the structure begins to collapse may indicate a yield pressure in the soil but this is not necessarily the case. The influence of soil structure on the compression behavior of residual soils is discussed further in Chapter 4. For an interesting and detailed description of the structure of one particular soil see Zhang et al. (2007). The term destructured is being increasingly used these days to denote soils that in their natural state are clearly influenced by structure of some sort, but that have been treated or manipulated in such a way that bonds between particles or any other structural effects have been eliminated. Its meaning is essentially the same as remolded, but it is intended to indicate that bonds or other forms of attachment between particles have been removed but the particles themselves are still intact. Remolding, on the other hand, means simply that the soil has been thoroughly reworked, and in the case of residual soils may mean that the particles themselves have been destroyed along with the structure. In this context we should note that some residual soils are not strictly particulate, that is, they do not consist of discrete individual particles. To the naked eye they may appear to consist of individual particles, but when remolded these particles disintegrate to form a collection of much smaller particles. 1.4
SPECIAL CLAY MINERALS
Apart from structure as a distinctive feature of many residual soils, geotechnical engineers should be aware of a group of very unusual clay minerals found only in residual soils. These are the two minerals, allophane and imogolite, which are normally linked together, and a third called halloysite. The extremely unusual properties of soils containing these minerals, especially allophane, can be a source of considerable puzzlement to engineers encountering them for the first time. A good example of this is the story of the Sasamua dam built in Kenya in the 1950s. A fairly comprehensive account of the construction of the dam is given by Terzaghi (1958). The dam is built of tropical red clay containing a large proportion of the clay mineral halloysite. Investigations and laboratory testing for the project indicated that the clay did not conform to conventional behavior. In particular, it consisted of very fine-grained clay but had much higher shear strength than “normal” clays of similar particle size. Recognized authorities at the time, including Terzaghi, were called in to review the available data and provide specialist advice. As part of the review, they gathered information from several existing dams around the world believed to be built of similar soil. One of the dams was the Cipanunjang dam in West Java, Indonesia. It was built by Dutch engineers in 1927, and forms a water supply reservoir still in operation today. The author
8
FUNDAMENTAL ASPECTS OF RESIDUAL SOIL BEHAVIOR
reinvestigated this dam, purely out of curiosity, while working in Indonesia in the 1970s (Wesley 1974). Based on the performance of Cipanunjang and the other dams, Terzaghi and his team concluded that the soil was satisfactory, and the construction of the dam proceeded without difficulties. The Cipanunjang dam in Indonesia is not actually built of red clay; it is built of a yellowish brown clay in which the predominant clay mineral appears to be allophane. However, it is weathered from volcanic ash and does contain some halloysite, so it has much in common with the Sasamua clay. Had the clay at Sasamua been similar to that at Cipanunjang, it would no doubt have raised even more concern as to its suitability for dam construction, since allophane clays frequently have extremely high water content, in the range of 75–200 percent. The Dutch engineers in 1927 were presumably happily unconcerned about whether their clay conformed to expected patterns of “normal” behavior. According to published records the engineers were actually building two earth dams at the time: one was Cipanunjang and the other was built out of a “normal” sedimentary clay. The latter suffered a major slope failure during construction. Cipanunjang, with its very unusual soil, was successfully completed, and is still in use today, as already indicated. An important lesson from this story is that case records and observation of field behavior are generally more reliable guides to the geotechnical properties of a soil than a collection of field or laboratory test statistics. Examination of natural or cut slopes in soils of volcanic origin in the wet tropics (which include Kenya and Indonesia) shows them to remain stable at remarkably steep angles. This simple fact should take precedence over test data as a reliable indicator of their geotechnical properties and behavior. A more detailed account of the properties of volcanic clays is given in Chapter 9. 1.5
THE INFLUENCE OF TOPOGRAPHY
Topography has a strong and fairly consistent influence on the weathering process, and thus on the type of clay minerals formed, especially in the wet tropics. In hilly and mountainous areas, the soil is well drained and seepage flow has a strong downward component, as illustrated in Figure 1.4. This leads to the formation of low-activity clay minerals, especially kaolinite. In volcanic areas, as noted above, the minerals allophane and halloysite may be formed initially before ending up as kaolinite. Soils containing these minerals generally have good engineering properties. As Vaughan (1985) states, with some caution, “residual soils are generally quite well behaved.” In wide, flat areas, drainage of any sort is much more limited, and moisture movement occurs primarily as a result of seasonal changes. Water is lost during dry periods from evaporation and the soil takes up moisture again
GEOTECHNICAL ANALYSIS, DESIGN, AND THE ROLE OF OBSERVATION AND JUDGMENT
9
Well drained hilly and mountainous areas: Downward seepage results in deep weathering, and soils tend to have good engineering properties.
Poorly drained, flat, low lying areas: Absence of vertical drainage results in shallow weathering and soils of poor engineering properties.
Downward seepage Water
table
Figure 1.4 Influence of topography on residual soil formation.
during periods of rainfall. This environment tends to produce montmorillinite and associated high-activity clay minerals (smectites). Soils containing these minerals normally have poor or highly undesirable geotechnical properties. The term vertisol is used by soil scientists for these soils because the cyclic wetting and drying process and associated surface cracking tends to cause movement of soil as well as water in both the upward and downward, that is, vertical, direction close to the surface. The term black clays or black cotton clays is used in geotechnical literature for these soils.
1.6 GEOTECHNICAL ANALYSIS, DESIGN, AND THE ROLE OF OBSERVATION AND JUDGMENT
Some general comments are appropriate at this stage on the design process used in geotechnical engineering and the extent to which this may be influenced by residual soil properties. The term design is used here to mean the complete process by which the geotechnical engineer arrives at an answer to the question he or she is addressing. This may be the design of a foundation, the deformation of a retaining wall, or the stability of a natural hill slope. The design process can be considered (somewhat simplistically) to consist of the following steps: 1. Gathering basic information on soil conditions, that is, the geology of the site and the soil stratigraphy 2. Undertaking suitable tests, in the field, or in the laboratory, to determine soil properties, particularly the parameters needed for analysis 3. Carrying out an analysis by using the relevant parameters in an appropriate theoretical model, which could be a bearing capacity formula, a slip circle calculation, or highly sophisticated numerical modeling treating the soil as an elastic plastic nonlinear material
10
FUNDAMENTAL ASPECTS OF RESIDUAL SOIL BEHAVIOR
There is a growing tendency to think of the above steps as the complete design procedure, even to the extent that the analysis (i.e., the calculations) is thought of as the design. The advent of the computer and the increasing prevalence of design codes have accentuated this tendency. This view of design downgrades, or even leaves out entirely, the nonanalytical aspects of geotechnical design, namely observation, precedent, experience, and judgment, which are (or should be) essential components of all geotechnical design. The relative importance of the analytical component of design and the nonanalytical components varies depending on the situation. For the determination of the bearing capacity of a foundation on a homogeneous clay layer, the analysis could well be the principal component, but for the determination of the stability of a natural slope, the contribution of analysis may be quite insignificant compared with the roles of visual observation of the slope and geological appraisal. Figure 1.5 shows two very elementary soil profiles: one of a sedimentary clay, and the other of a residual soil. It is possible that the sedimentary clay is essentially homogeneous, apart possibly from faint traces of horizontal bedding layers. However, it is unlikely that the residual soil profile is at all homogeneous, although there are situations where it may be so. It is much more likely to be heterogeneous, with a gradual change in properties with depth, and possibly containing joint or fault planes. Residual soils are thus less likely to be amenable to tidy analytical procedures than sedimentary soils, and the roles of observation, experience, and judgment become even more important parts of geotechnical engineering in residual soils. Terzaghi once used the delightful phase “the omnipotence of theory” in the following statement: However, as soon as we pass from steel and concrete to earth, the omnipotence of theory ceases to exist. In the first place, the earth in its natural state is never uniform. Second, its properties are too complicated for rigorous theoretical treatment. Finally, even an approximate mathematical solution of some of the most common problems is extremely difficult (Terzaghi 1936).
Sedimentary Soil Homogeneous apart from traces of horizontal bedding layers
Residual Soil Heterogeneous, varies with depth from soil to weathered rock, contains various discontinuities in the form of joints, faults, and partially weathered rocks
Figure 1.5 Simplified soil profiles in sedimentary and residual soils.
SUMMARY OF BASIC DIFFERENCES BETWEEN RESIDUAL AND SEDIMENTARY SOILS
11
If the omnipotence of theory ceases to exist with sedimentary soils, then it inevitably declines even further with residual soils. However, we should not read more into Terzaghi’s words than he presumably intended. He is not rejecting theory, only its omnipotence, and uses the term “theory” to denote the process of collecting a set of figures, introducing them into appropriate equations, and coming up with the required answer. Given this meaning, theory is not to be confused with fundamental concepts and principles and Terzaghi’s statement should not be taken to mean that we can downgrade the latter. 1.7 SUMMARY OF BASIC DIFFERENCES BETWEEN RESIDUAL AND SEDIMENTARY SOILS
Although we have not yet covered them all, the most significant differences between residual soils and sedimentary soils can be summarized as follows: 1. Residual soils are generally more heterogeneous than sedimentary soils. 2. Because they have not been formed by a sedimentation process, stress history is an irrelevant concept and not a significant influence on residual soil behavior. 3. The theoretical framework for understanding sedimentary soils involving the e-log p plot and the division into normally consolidated and overconsolidated soils is not applicable to residual soils. 4. Some residual soils, especially those of volcanic origin, may have unusual properties due to the presence of clay minerals not found in sedimentary soils. 5. Some residual soils in their undisturbed state are not strictly particulate, that is, they do not consist of discrete particles. Such soils may appear to consist of individual particles, but when the soil is disturbed or remolded the particles disintegrate into smaller particles. 6. Empirical correlations between soil properties developed from the study of sedimentary soils may not be valid when applied to residual soils. 7. The water table in residual soils is often relatively deep, and subject to fluctuations from climatic effects. This means that much of the action of interest to geotechnical engineers takes place above the water table, and an understanding of the pore pressure regime above the water table becomes an important component to understanding residual soil behavior. 8. In evaluating the properties of residual soils it is very important to first observe carefully their behavior in the field, before looking at the results of laboratory tests.
12
FUNDAMENTAL ASPECTS OF RESIDUAL SOIL BEHAVIOR
REFERENCES Gasparre, A., M. R. Nishimurai, M. R. Coop, and R. J. Jardine. 2007. The influence of structure on the behaviour of London Clay. Geotechnique 57: 19–31. Maccarini, M. 1987. Laboratory Studies of Weakly Bonded Artificial Soil. Thesis, University of London. Terzaghi, K. 1936. Presidential address. First International Conference on Soil Mechanics and Foundation Engineering. Cambridge, Massachusetts, USA. Terzaghi, K. 1958. Design and performance of the Sasumua dam. Proceedings, the Institution of Civil Engineers. London, April, Vol. 9, 369–396. Toll, D. G., V. Malandraki, Z. Ali Rahman, and D. Gallipoli. 2006. Bonded soils: problematic or predictable? Proceedings 2nd International Conference on Problematic Soils, Malaysia. Ci-Premier Pte Ltd, 55–62. Vaughan, P. R. 1985. Mechanical and hydraulic properties of in situ residual soils. Proceedings, First International Conference on Geomechanics in Tropical, Lateritic, and Saprolitic Soils, Brazil, 1–33. Wesley, L. D. 1974. Cipanunjang (Tjipanundjang, old spelling) Dam in West Java, Indonesia. Journal of the Geotechnical Division ASCE100/GT5: 503–522. Wesley, L. D. 2009. Fundamentals of Soil Mechanics for Sedimentary and Residual Soils. New York: Wiley. Zhang, G., A. J. Whittle, J. T. Germaine, and M. A. Nikolinakau. 2007. Characterisation and engineering properties of an old alluvium in Puerto Rico. Proceedings, Characterisation and Engineering Properties of Natural Soils, Vol. 4, 2557–2588. Leiden, The Netherlands: Taylor & Francis/Balkema.
CHAPTER 2
EVALUATION, CHARACTERIZATION, AND CLASSIFICATION OF RESIDUAL SOILS
2.1
INTRODUCTION
The main objective of this chapter is to provide some general guidelines for evaluating the geotechnical properties of residual soils. In doing this we need to be mindful of the fact that it is the complete soil profile that is normally of interest to us, and not the properties of any particular soil element within that profile. This is no different to the situation with sedimentary soils. The most useful starting points for evaluating the profile as a whole are probably the topography and the parent rock with their associated weathering patterns, along with simple visual observation. The influence of topography was described in the previous chapter and will not be dealt with further here. 2.2
PARENT ROCK AND THE SOIL PROFILE
The most common representation of the profile of a residual soil is that first put forward by Little (1969), based on earlier work by Moye (1955). This is shown in Figure 2.1a. The profile is divided into six zones based on the degree of weathering, ranging from fresh rock to soil. Similar systems have been proposed by a number of other authors, sometimes on a general basis and sometimes in relation to a particular formation or locality. Pender (1971), for example, describes the use of a slightly amended version for 13
14
EVALUATION, CHARACTERIZATION, AND CLASSIFICATION OF RESIDUAL SOILS
6
Soil: clay or silt
5
Completely weathered
4
Highly weathered
3
Moderately weathered
2
Slightly weathered
1
Fresh rock
(a) Gradual weathering profile—typical of weathered granite
Soil: clay or silt
Weathered rock Fresh rock
(b) Sharp transition from rock to soil —typical of weathered basalt
Silty clay layers, almost homogeneous but distinguished by slightly different coloring
Soil: Inter-bedded clay, silty clay, silt, and dense silty sand
Fresh rock
Fresh rock: inter-bedded sandstone and clay-stone
(c) Uniform layers, degree of weathering not necessarily related to depth—typical of volcanic ash
(d) Stratified nature of parent rock reflected in soil profile—typical of weathering of soft sedimentary rock, especially sandstone
Figure 2.1 Variations in residual soil weathering profiles.
use in the classification of weathered greywacke rock in Wellington, New Zealand. Saunders and Fookes (1970) review and describe some of the other systems. The objective and limitations of these systems should be clearly recognized. First, they are not systems for describing or classifying residual soils. They are methods for describing weathered rock profiles, rather than describing the soil itself. They provide information on the in situ state of the soil, but not on the actual composition of the soil. They are presumably intended to be used in conjunction with systems that describe the composition or nature of the soil itself, such as the Unified Soil Classification System. Second, they are only relevant to the weathering of particular rock types. Little (1969) stated that the classification system he proposed was intended for the residue resulting from the weathering of igneous rocks in the humid tropics, and not for weathering profiles generally. As shown in Figure 2.1a, the profile consists of a series of thick zones of not greatly differing thicknesses. With different parent rock, however, this may not be the case at all. The boundary between soil and rock can be abrupt, with only a thin zone of transition material, as shown in Figure 2.1b. This is frequently the case with the weathered basalt soils found in the North Island of New Zealand, and is also the case with red clays derived from andesitic or basaltic rock in Java, Indonesia. Townsend (1985) indicates that the weathering process is different in acidic and basic rocks, and comments: “one major difference between the basic and acidic rocks is that most pedologists suggest that basic rocks weather rapidly into soils, providing a sharp contact zone with the weathering of minerals occurring within a layer of only a few millimeters. Conversely, the zone of alteration in acidic quartz-rich rocks appears to be quite thick.”
INFLUENCE OF PARENT ROCK ON GEOTECHNICAL PROPERTIES
15
Volcanic ash soils are different again, as illustrated in Figure 1.2c. They may also show an abrupt boundary between the soil and the underlying rock, which is likely to be basaltic or andesitic lahar, or solid rock. The sharp boundary is not because of the nature of the weathering process, but because the ash has been deposited on top of the rock and the soil is derived from fresh ash and not from the underlying rock itself. It is also the case that the deeper layers in the soil profile may by more highly weathered than those above because they are older and have been subject to the weathering process over a greater period of time. With sedimentary rocks, especially soft sandstones, mudstones, and shales, the picture is different again. The weathering process in this case may not be one in which rock minerals are broken up and converted chemically into clay minerals, but one in which cementing agents are dissolved by percolating water, thus releasing clay minerals already existing in the parent rock. In this situation, the soil profile is likely to reflect both the weathering sequence and the different layers in the parent rock. When the parent rock consists of interbedded sandstone and mudstone, for example, this may be reflected in the resulting soil, which will consist of interbedded layers of silty sand and clay, as indicated in Figure 2.1d. This is the case with the residual soils in the Auckland area of New Zealand derived from weathering of a sandstone and mudstone formation known as Waitemata series. It is a soft rock with unconfined compression strength generally in the range of 1500– 4000 kPa. The term saprolite should be noted at this point. It is used by some geologists and geotechnical engineers to describe the zone of weathering between rock and soil; this would probably include zones 3–5 in Figure 2.1a. The criteria for a soil to be described as saprolite are the following: 1. It is a soil in the geotechnical sense. 2. It exhibits clear inherited structural features that make possible the identification of the parent rock. 3. It is authentically residual, meaning it is derived directly from the weathering of the rock below it, and is an integral part of the weathered profile from the parent rock. These criteria are put forward by the Committee on Tropical Soils of the International Society for Soil Mechanics and Foundation Engineering (1985). 2.3 INFLUENCE OF PARENT ROCK ON GEOTECHNICAL PROPERTIES
Residual soils derived from the weathering of igneous or volcanic rock in areas where drainage is good generally have good engineering properties.
16
EVALUATION, CHARACTERIZATION, AND CLASSIFICATION OF RESIDUAL SOILS
Good drainage means rainwater enters the soil surface and seeps more or less continuously toward its destination at a stream or valley. This implies a reasonably strong downward component to the seepage flow. Such areas clearly include hilly and mountainous areas and highland areas from which seepage can escape to lower levels without difficulty. Igneous and volcanic rocks, not surprisingly, tend to be the dominant rock types that make up these areas. As already mentioned in Chapter 1, this environment tends to promote the formation of clay minerals with good engineering properties, especially kaolinite. Residual soils derived from the weathering of sedimentary rocks are generally less likely to have good engineering properties, regardless of the topography in which they are found. This is especially true of soft sedimentary rocks such as shale. As we saw earlier, the weathering process in these materials is quite different from the weathering of igneous or volcanic rocks. With the latter, the process is one of conversion— rock minerals are converted into clay minerals. With soft sandstones, siltstones, and shales, the weathering process is one of solution and release—cementating material is removed by solution and preexisting clay minerals are released. The properties of the soil formed in this way are very dependent on the nature of the clay minerals. In shales the clay minerals are often of a highly plastic nature and thus give the resulting soil undesirable engineering properties. Sedimentary rocks also tend to be found in lower, poorly drained areas, where the weathering process tends to produce clay minerals of the troublesome variety, especially those in the montmorillonite group. It is perhaps questionable whether soils derived from the weathering of shales should be regarded as residual soils, especially when the shale has properties that lie across the boundary of hard soil and soft rock. Their properties are more likely to reflect their former lives as sedimentary soil than the weathering process that qualifies them as residual soils. Mention was made in Chapter 1 (Section 1.6) of a type of soil called black cotton soil, made up predominantly of high-activity clay minerals of the montmorillonite group, and well known for its poor engineering properties. This soil is almost always formed in rather flat low-lying areas where drainage is poor. It is probably also the case that this soil is derived predominantly from sedimentary rocks such as shale. 2.4
THE ROLE OF OBSERVATION
The role of observation in understanding and evaluating residual soils cannot be overemphasized. It is not uncommon to find (in the geotechnical literature) papers on residual soils that provide a great number of statistical data in the form of laboratory test results, but little or no observational information on how the soils actually behave in the field. The term observational method is often used in geotechnical engineering to describe a process
STANDARD INDEX TESTS
17
whereby careful monitoring and observation is carried out during project construction to evaluate the behavior of the soil, and to make adjustments to the construction process as necessary. The term could also be used to describe the gathering of data prior to the start of construction by observing the field behavior of the soil (or soils) in the immediate vicinity of the project. This could be the simple observation of natural slopes or cuttings; the most useful information on the performance of natural slopes is direct observation of them, not values of c and φ from laboratory tests. It could also be the investigation of the performance of foundation types of existing structures in the project area. As emphasized in Chapter 1, judgment and experience play a dominant role in geotechnical engineering projects in residual soils, and careful observation of soil behavior in the field is an element of experience that is essential to the development of sound judgment. 2.5
STANDARD INDEX TESTS
The usefulness of conventional index tests and classification systems, such as the Unified Classification System, for residual soils has been questioned on a number of occasions, as, for example, by De Graft-Johnson and Bhatia (1969). It is argued that the in situ character of the soil is destroyed in preparing the soils for testing so that the results do not give an indication of the properties of the undisturbed soil. It is also argued that the results of particle size and plasticity measurements are strongly influenced by the method of sample preparation. There is some truth in these arguments, although they are also applicable to sedimentary soils, at least to some extent. It is the author’s view that conventional index tests, especially Atterberg limits, do play an important role in evaluating and characterizing residual soils. The argument that the results are influenced by drying is not an argument for rejecting the tests since there is no difficulty in avoiding drying the soil. Frost (1967), in drawing attention to this question, rightly called for correct pretesting procedures rather than rejection of the tests. When carrying out particle size or Atterberg limit tests on residual soils, it is highly desirable that the material not be air or oven dried prior to testing, especially if the soil is of volcanic origin. It should be dried only to the water content needed to carry out the test. Figure 2.2 illustrates the influence of oven drying on Atterberg limit tests carried out on samples taken at different locations in soils weathered from a series of volcanic ash layers in Tauranga, New Zealand. Some of the layers were dark brown in color, while others were a pale yellow color. Some very limited clay mineralogy tests showed the samples contained both halloysite and allophane.
18
EVALUATION, CHARACTERIZATION, AND CLASSIFICATION OF RESIDUAL SOILS
Plasticity Index
60 Natural Oven dried 40
20
0
20
40
60 Liquid Limit
80
100
120
Figure 2.2 Influence of oven drying on samples from a sequence of volcanic ashes.
We should recognize that particle size and Atterberg limit tests reflect the properties of the soil after it has been disturbed and completely remolded. They therefore reflect only the properties of the material itself, and tell us nothing about the material in its natural (undisturbed) state. These properties can be referred to as the intrinsic properties of the soil. Particle size and Atterberg limit measurements are considered in the following sections. 2.5.1
Particle Size
The situation with respect to particle size is not greatly different to that for sedimentary soils. For coarse-grained soils, particle size gives a good indication of properties, but for fine-grained soils particle size is an unreliable indicator of properties. This is a basic tenet of the Unified Soil Classification System, which does not require knowledge of the particle size distribution within the silt and clay size range, only the total percentage in this range. In addition to this general point, the following points should be noted: •
•
•
Some residual soils do not consist of discrete particles in their undisturbed state, and it is only the sample preparation process that separates the soil into discrete particles. Some residual soils, especially those made up of a high proportion of allophane or halloysite, are extremely fine-grained, but still have very good engineering properties. Conventional particle size measurements using the sedimentation process (hydrometer test) can be difficult to carry out on clays with a high proportion of allophane, because of a strong tendency of the soil to flocculate.
STANDARD INDEX TESTS
19
For these reasons, particle size measurements on soils that are clearly in the clay or silt category are not of great value as indicators of likely engineering properties. 2.5.2
Atterberg Limits
Atterberg limits are a valuable tool for evaluating and characterizing residual soils. However, we should not look at the plastic limit, the liquid limit, or the plasticity index in isolation. It is the position the soil occupies on the plasticity chart that is a useful indicator of its engineering properties, just as good if not better than with sedimentary soils. The plasticity chart, shown in Figure 2.3, should therefore be regarded as a tool for evaluating the likely properties of the soil, rather than as method for classifying it. Soils that plot well below the A-line generally have good engineering properties, while those that plot above the A-line have poor engineering properties. Tropical red clays and volcanic ash soils have remarkably good geotechnical properties, while black cotton soils have highly undesirable properties. If the plasticity chart is used for classification purposes, problems can arise with residual soils, especially clays of volcanic origin containing the clay minerals halloysite or allophane. These soils tend to plot below that A-line, especially those containing allophane. However, they do not really behave as silts; they show only faint traces of “quick” behavior or dilatant characteristics. At the same time they do not appear to be highly plastic, and deciding whether they should be called clays or silts is somewhat problematical. Silty clay is probably the best term to use for allophane soils.
150 e
Lin
A-
Plasticity Index
Clay
Silty clay
100 ils so n” ite) o t n ot illo k c or lac tm “B mon (
50 Tropical red clays (halloysite)
0
50
Silt
ils
o ash s anic Volc ophane) ll a (
100 150 Liquid Limit
200
250
Figure 2.3 The conventional plasticity chart and several tropical residual soils.
20
EVALUATION, CHARACTERIZATION, AND CLASSIFICATION OF RESIDUAL SOILS
Problems also arise when attempts are made to relate specific soil properties, or classification boundaries, to one or other of the liquid and plastic limits. For example, the British classification system (BS 5930:1981) divides soils into a number of categories based on the liquid limit. Such a division is not very relevant to residual soils. It is the position above or below the A-line that is of most significance, especially with tropical residual soils. A division into clays, silty clays, and silts, as indicated in Figure 2.3, would be a more useful division than one based on the liquid limit. It should be noted that the influence of increased mixing (or even drying) of the soil on the Atterberg limits is to move the point on the plasticity chart parallel to the A-line (Morin and Todor 1975); hence, if distance above or below the A-line is our main criterion for evaluating soil this movement is not of great significance. 2.5.3
‘‘Compactness’’ Indexes
As mentioned above, the common index tests, especially particle size and Atterberg limits, give us information on the intrinsic properties of the soil, that is, properties that arise from its composition and that are always present in the soil. They are valuable indicators of its behavior in the remolded state but on their own they provide only very limited information about its natural state in the ground. For clays, there are two additional parameters that provide us with useful information on the natural state of the soil: 1. The undrained shear strength, which is a simple measure of its undisturbed strength 2. A parameter that indicates its compactness or denseness, that is, whether its particles are tightly packed together, or have an open noncompact state In the case of clays this compactness index is the liquidity index and in the case of sands it is the relative density (or density index). The way these indexes are defined is shown in Figure 2.4. The term liquidity index is somewhat misleading. It is only an indication of liquidity if the soil is fully remolded. For undisturbed soils. the liquidity index is better regarded as a compactness index, as there is no direct connection between the liquidity index and the consistency (strength) or liquidity of the soil. A soil with a liquidity index of zero has a natural water content equal to its plastic limit and is in a dense or compact state, while a soil with a liquidity index of unity has a natural water content equal to its liquid limit and is in a very open or nondense state. This does not mean that it is almost a liquid, except after remolding. The fact that many natural clays and silts have liquidity indexes in excess of unity yet are still firm to stiff materials is simply a reflection of the influence of structure on their properties. The
CLASSIFICATION SYSTEMS FOR RESIDUAL SOILS
0
Density Index LL − PL
= Relative Density emax − en = e max − emin
emax
en
wn
emax − emin
Noncompact (loose) state
emax − en
LL
Water content = Liquidity Index w − PL = n LL − PL 0
Void ratio
Density Index wn − PL
Void ratio or water content as measures of compactness
1
21
PL
CLAY
Compact (dense) state
1
emin
SAND
Figure 2.4 Compactness indexes for clay and sand.
undrained shear strength of many, if not most, residual soils, is quite high, commonly in the range of 75–200 kPa. The liquidity index is a very useful guide to the likely behavior of residual soils. For example, a soil with a high liquidity index is likely to display the following properties: 1. A large loss of strength when it is disturbed or remolded. In other words, it is a highly sensitive soil. It will be a difficult soil in which to undertake earthworks and compaction operations. 2. A well-defined yield stress when the soil is loaded, either in one-dimensional or three-dimensional compression. Although this is likely to be the case, it is not necessarily so. Some soils of high sensitivity have a strong structure, and a good deal of energy is needed to reduce it to its fully remolded state. Such soils are less likely to display a clear yield pressure.
2.6 2.6.1
CLASSIFICATION SYSTEMS FOR RESIDUAL SOILS Introduction
Various attempts have been made over the years to devise systems for the description or classification of residual soils. However, no generally accepted methods have been established. This is not at all surprising, in
22
EVALUATION, CHARACTERIZATION, AND CLASSIFICATION OF RESIDUAL SOILS
view of the diverse nature of residual soils, and it is unlikely that a universal scheme is either desirable or a practical possibility. Methods currently in use for grouping or classifying residual soils fall into three broad types as follows: 1. Methods based on the weathering profile 2. Methods based on pedological classification 3. Methods intended for local use on specific soil types only Methods based on the weathering profile have already been considered (Section 2.2) and will not be further discussed here. 2.6.2
Methods Based on Pedological Groups
Geotechnical engineers have made use of pedological terms to designate various soil groups for many years. The term laterite or lateritic soil was one of the first such uses, and goes back at least to Bee (1948). Ranganathan (1961) made one of the earliest uses of the term black cotton clay. The author (Wesley 1973) used the terms latosol and andosol to designate two soil groups in Indonesia; this usage was taken directly from the pedological classification system being used at that time by Indonesian soil scientists (Junas Dai and Driessen 1972; Lenvain et al. 1972). Lohnes and Tuncer (1977) also used the term andosol in describing volcanic ash soils in Hawaii. Various additional terms have since been added to the geotechnical literature, and different terms are used for the above groups by different countries. The terms oxisols, andepts, and vertisols are in common use for lateritic soils (latosols), andosols, and black cotton soils, respectively. Mitchell and Sitar (1982) present a table showing the variety of names used by three pedological systems, namely the French, FAO, and U.S. Soil Taxonomy. Uehara (1982) gives a useful account of the various pedological groups, and their associated properties. The three soil types mentioned above, namely lateritic soils, andosols, and black cotton soils, however, remain the three most distinctive tropical soil types, and appear to be the types of most interest to the engineer. Table 2.1 summarizes the various names used for these groups. Table 2.1 also shows the predominant clay minerals associated with each group, and suggests that mineralogical composition is a strong influence on the properties of each group. The use of pedological names has not generally been done with the intention of establishing rigorous classification systems along the lines of those used by soil scientists. The names have simply been borrowed as a convenient way of identifying particular soil groups. Inevitably, perhaps, some confusion has been created by the rather unsystematic use of these terms. Andosols are sometimes included in the same group as red clays and vice versa. In the author’s view, andosols (volcanic ash clays) are a distinctive group and, although in the tropics they may be associated with
23
Ferralsols
Andosols
Vertisols
Volcanic ash soils Andosols
Black cotton soils Black clays Tropical black earths Grumusols
FAO
Lateritic soils Latosols Red clays
Commonly Used Names
Vertisols
Andepts
Oxisols
U.S. Soil Taxonomy
Vertisols
Eutropic brown soils of tropical regions on volcanic ash
Ferralitic soils
French
Rigorous Pedological Names
Smectite (montmorillonite)
Halloysite Kaolinite Gibbsite Geothite Allophane and minor Halloysite
Dominant Clay Minerals
Table 2.1 Distinctive Tropical/Residual Soil Groups of Interest to Geotechnical Engineers
Characterized by very high water content and irreversible changes when dried Problem soils, high shrinkage and swell, low strength
Very large group with wide variation in characteristics, properties generally good
Important Characteristics
24
EVALUATION, CHARACTERIZATION, AND CLASSIFICATION OF RESIDUAL SOILS
red clays to some extent, their composition and properties are different, and for geotechnical engineering purposes they should not be confused with red clays. Andosols occur in countries like Japan, Chile, and New Zealand, which are not tropical countries, and are not associated with red clays at all. Mention should perhaps be made of a classification system put forward by the British Geological Society (1990). This uses a complex pedological basis for its classification scheme, and does not appear to have gained significant acceptance by the geotechnical community. Apart from its complexity, it also suffers from the serious disadvantage that there is no clear link between the classification groups and their engineering properties. 2.6.3
Methods Intended for Specific Local Use
In view of the complexity of residual soils, and the almost total lack of any common features among some residual soil groups (for example, black cotton soils and weathered granite soils), it is not surprising that descriptive or classification methods have been developed for local use in particular formations. Lohnes and Tuncer (1977) for example, describe a system suggested for use with lateritic soils from Hawaii and Puerto Rico. Pender (1971, 1980) describes empirical correlations for the weathered greywacke of Wellington, New Zealand. Wirth and Zeigler (1982) describe a system specifically developed for use on the Baltimore subway project. These methods are highly desirable for dealing in a systematic way with particular formations, and it is likely that the profession will see increasing use of such systems in the future. A word of warning, however, should be exercised to those seeking to develop such systems, as there is a danger that existing systems will merely be modified in some way, or correlations valid for one group of residual soils will be assumed to provide a basis for correlations within another group. For example, correlations between strength and void ratio may be valid for some soils (Pender 1980; Lohnes and Tuncer 1977), but attempting to find such a correlation for volcanic ash soils would likely be a futile exercise. Each soil type must be evaluated on its merits. 2.6.4
A Grouping System Based on Mineralogy and Structure
The specific characteristics of residual soils that distinguish them from sedimentary soils can generally be attributed either to the presence of unusual clay minerals or particular structural affects, such as the presence of unweathered or partially weathered rock, planes of weakness, interparticle bonds, and the like. These influences can be grouped under the general headings of composition and structure. Composition includes particle size, shape, and especially mineralogical composition, while structure includes both macrostructure and microstructure, as discussed in Chapter 1.
CLASSIFICATION SYSTEMS FOR RESIDUAL SOILS
25
A useful first step in the grouping of residual soils is therefore to divide them into groups on the basis of composition alone, without reference to their undisturbed state. Such a grouping system is described by Wesley (1988) and Wesley and Irfan (1997). The following three groups are suggested: Group A: Soils without a strong mineralogical influence Group B: Soils with a strong mineralogical influence coming from conventional clay minerals commonly found in sedimentary soils Group C: Soils with a strong mineralogical influence coming from special clay minerals not found in sedimentary soils By eliminating those soils that are strongly influenced by particular clay minerals, there is some possibility of identifying a group of soils that can be expected to have similar properties. In general, soils that have a weathering profile of the type illustrated in Figure 2.1a will come within this group. Weathered granite soils are a typical example—they are generally of a fairly coarse nature, with a relatively low clay fraction. In rare instances, the top layer (the soil layer) may be sufficiently advanced in weathering to become a true clay with properties strongly influenced by distinctive clay minerals. Group A soils can be subdivided further on the basis of the extent and manner in which their behavior is influenced by structural effects. It is convenient to separate structural effects into the two broad groups mentioned earlier (Section 1.3 of Chapter 1), namely macrostructure and microstructure. On this basis group A can therefore be divided into two main subgroups:
Group A: Residual Soils without a Strong Mineralogical Influence
Subgroup (a): Soils in which macrostructure plays an important role in the engineering behavior of the soil. The lower horizons of the soil profile shown in Figure 2.1a fall into this category. Subgroup (b): Soils without macrostructure, but with a strong influence from microstructure. The most important form of microstructure is interparticle bonding or cementation, and, although this cannot be identified by visual inspection, it can be inferred from fairly basic aspects of soil behavior. Sensitivity, in particular, is a very good measure of microstructure, as high sensitivity results from the presence of a distinctive structure (involving some form of bonds) that is destroyed on remolding. Additionally, group A contains a smaller subgroup: Subgroup (c): Soils that are not greatly influenced by macro- or microstructural effects are included here as a third subgroup.
26
EVALUATION, CHARACTERIZATION, AND CLASSIFICATION OF RESIDUAL SOILS
However, this subgroup is a minor group, as few residual soils of group A fall into this category. Group B: Residual Soils Strongly Influenced by Conventional Clay Minerals This group is made up of soils that are strongly influenced by
conventional clay minerals, such as those normally found in sedimentary soils. The most significant member of this group is the black cotton soil group or vertisols, whose characteristic properties are high shrink and swell potential, high compressibility, and low strength. These characteristics are directly related to their predominant mineralogical constituent, which is montmorillonite or similar minerals of the smectite group. Information in the literature suggests that not many other residual soils belong to this group, although some soils derived from sedimentary rocks (sandstones and mudstones) have properties that are fairly strongly influenced by mineralogical composition. Some soils in the Auckland area of New Zealand derived from the weathering of Waitemata sandstone may come into this category. They can have high shrink and swell characteristics due to the presence of montmorillonite. Group C: Residual Soils Strongly Influenced by Special Clay Minerals not Found in Sedimentary Soils The two most important minerals
involved here are the silicate clay minerals halloysite and allophane, described briefly in Chapter 1, and the associated minerals known as sesquioxides. The influence of halloysite and allophane on the soil properties is fairly clear from the case studies recorded in the literature. The influence of the sesquioxides is less well documented. It is convenient, however. to subdivide this group into three subgroups: 1. Halloysite soils: The principal influence of halloysite appears to be that the engineering properties of the soil are good, despite a high clay fraction and very small particle size, and fairly high values of natural water content and Atterberg limits. The good engineering properties appear to be the direct result of their mineralogical composition, or in some cases cementation arising from the presence of the sesquioxides. Terzaghi (1958), Matyas (1969), and Wesley (1973) have given accounts of the good engineering performance of these soils. 2. Allophane soils: The properties of allophane have already been described briefly and will be covered in greater detail in Chapter 9. It should be understood, however, that the influence of allophane is both dramatic and puzzling, in that it results in soils having water contents ranging from about 80 to 250 percent, but which still perform very satisfactorily as engineering materials. They are frequently much superior to soils with water contents only a fraction of the above values.
CLASSIFICATION SYSTEMS FOR RESIDUAL SOILS
27
3. Soils influenced by the presence of sesquioxides: The principal role of the sesquioxides appears to be as cementing agents that bind the other mineral constituents into clusters or aggregations. With sufficient concentration of sesquioxides, the hard concretionary materials commonly known as laterite are formed. The silica/alumina ratio (SiO2 /A12 O3 ) and the silica/sesquioxide ratio have both been used as indicators of degree of laterization. This subgroup could perhaps be termed the lateritic group, but the term laterite is generally used very loosely, sometimes to include both halloysite and allophane clays—whose behavior is not significantly influenced by the sesquioxides. The above groups, especially the halloysite and allophane groups, can be further subdivided on the basis of structure. Allophane soils (which appear to be always associated with volcanic ash as parent material) show considerable variation in their structural influence. Allophane soils in Indonesia are generally of low to moderate sensitivity, while those in Japan are likely to be of moderate to very high sensitivity (Kuno et al. 1978), indicating a strong structural component in their undisturbed state. The halloysitic soils of Java, Indonesia, with which the author is familiar, do not appear to have significant microstructure or macrostructure. They are of very low sensitivity and their behavior in the remolded and undisturbed state is often almost identical. This is not necessarily true of other halloysite soils. Table 2.2 presents this suggested grouping system, and indicates the descriptive information needed to justify placing a soil in a particular group. Table 2.3 gives examples of soils that belong in each of these groups, and some guidelines for their identification. In general, weathered igneous rocks, such as granite, and many weathered sedimentary formations will produce soils that belong in group A. As indicated earlier, the weathered granites of Hong Kong and Malaysia are not strongly influenced by mineralogical composition, but their behavior is likely to be strongly influenced by both macro- and microstructure, so they clearly belong in group A. On the other hand, volcanic ash soils (andosols or allophane clays) are strongly influenced by the unusual clay mineral allophane, and clearly belong in group C. They do not often show evidence of macrostructure, but can be moderately to highly sensitive, indicating a strong microstructure influence. As mentioned earlier, the above grouping of residual soils is intended to provide a basis for identifying groups of residual soils that can be expected to have similar engineering properties. It appears to provide a better basis for doing this than using pedological terms. The terms red clay and lateritic soil, for example, are used to cover such a wide range of materials as to be almost useless to the engineer. The predominant clay minerals in some red clays (or lateritic soils) are kaolinite and montmorillonite; these soils plot above the A-line on the plasticity chart. The predominate clay mineral in other red clays, as already mentioned, is
28
GROUP A Soils without a strong mineralogical influence
Major Division Miscellaneous
Miscellaneous
Miscellaneous
(b) Strong microstructure influence
(c) Little or no structural influence
Common Pedological Names Used for Groups
(a) Strong macrostructure influence
Subgroup
Grouping system
Table 2.2 A Classification or Grouping System for Residual Soils
Give descriptive details of type of rock from which the soil has been derived
Parent Rock
Describe nature of structure: stratification, fractures, fissures, faults etc., presence of partially weathered rock (%?) Describe nature of microstructure and/or evidence of it: influence of remolding, sensitivity, liquidity index Indicate evidence for little or no structural effect
Information on Structure
Descriptive Information on in Situ State
29
GROUP C Soils strongly influenced by clay minerals essentially found only in residual soils
GROUP B Soils strongly influenced by normal clay minerals
(c) Sesquioxide subgroup: gibbsite, goethite, hematite
(b) Halloysite subgroup
(b) Other clay minerals? (a) Allophane subgroup
(a) Smectite (montmorillonite) group
Tropical red clays Latosols Oxisols Feralsols Lateritic soils Laterites Ferralitic soils Duricrusts
Volcanic ash soils Andosols or andisols Andepts
Black cotton soils Black soils Tropical black earths Grumusols Vertisols
Give basis for inclusion in this group Describe any structural effects, especially cementation effects or the sesquioxides
Give basis for inclusion in this group; describe any structural influences, either macrostructure or microstructure As above
30
GROUP A Soils without a strong mineralogical influence
Group
Soils formed from very homogeneous rocks
(c) Little or no structural influence
(b) Strong microstructure influence
Highly weathered rocks from acidic or intermediate igneous or sedimentary rocks Completely weathered rock, formed from igneous or sedimentary rocks
(a) Strong macrostructure influence
Subgroup
Examples
Characteristics of Residual Soil Groups
Major Group
Table 2.3
This is a very large group of soils (including the “saprolites”) where behavior (especially in slopes) is dominated by the influence of discontinuities, fissures, etc. These soils are essentially homogeneous and form a tidy group much more amenable to rigorous analysis than group (a) above. Identification of nature and role of bonding (from relic primary bonds to weak secondary bonds) important to understanding behavior. This is a relatively minor subgroup. Likely to behave similarly to moderately overconsolidated soils.
Visual inspection
Little or no sensitivity, uniform appearance
Visual inspection, and evaluation of sensitivity, liquidity index, etc.
Comments on Likely Engineering Properties and Behavior
Means of Identification
31
GROUP C Soils strongly influenced by clay minerals essentially found only in residual soils
GROUP B Soils strongly influenced by normal clay minerals
Soils weathered from volcanic ash in the wet tropics and temperate climates
Soils often derived from volcanic material, especially tropical red clays
(b) Halloysite subgroup
Black cotton soils and many similar dark colored soils formed in poorly drained conditions
(a) Allophane subgroup
(b) Other clay minerals?
(a) Smectite (montmorillonite) group
Reddish color, well-drained topography, and volcanic origin are useful indicators
Position on plasticity chart, and irreversible changes on drying
Dark color (grey to black) and high plasticity suggest soils of this group
Characterized by very high natural water contents and Atterberg limits. Engineering properties generally good, though in some cases high sensitivity may make earthworks difficult. These are generally very fine-grained soils of low to medium plasticity, and low activity. Engineering properties generally good. (Note that there is often some overlap between halloysite and allophane clays.) (continues)
These are normally problem soils, found in flat and low-lying areas, having low strength, high compressibility, and high swelling and shrinkage characteristics. Likely to be a very minor subgroup.
32
Major Group
Group
Table 2.3 (continued)
(c) Sesquioxides: gibbsite, goethite, hematite
Subgroup Laterites, or possibly some red clays referred to as lateritic clays
Examples
Nonplastic or low-plasticity materials, generally of granular or nodular appearance
Means of Identification
This is a very wide, poorly defined group, ranging from silty clay to coarse sand and gravel. Behavior ranges from low-plasticity silty clay to gravel. These materials are the end products of a very long weathering process.
Comments on Likely Engineering Properties and Behavior
REFERENCES
33
halloysite; these soils plot below the A-line and have quite distinct, and good, engineering properties. One disadvantage of using mineralogical composition as a grouping or classification basis is readily apparent, namely that geotechnical engineers seldom have ready access to the facilities needed for mineral identification. However, most countries have institutions that can undertake mineralogical studies, and cooperation between geotechnical engineers and these institutions ought not to be difficult and should be to the benefit of both parties.
REFERENCES Bee, R. J. 1948. Some notes on laterite—a soil of engineering importance in the tropics, Conf. on Civil Engineering Problems in the Colonies. London: Inst. Civil Engineers, 191. British Geological Society Engineering Group Working Party Report. Tropical Residual Soils (1990), Vol. 23, No.1, 1–101. BS5930. 1981. Code of Practice for Site Investigations. London: British Standards Institute. Committee on Tropical Soils of the International Society for Soil Mechanics and Foundation Engineering. 1985. Peculiarities of geotechnical behaviour of tropical lateritic and saprolitic soils. Progress Report 1982–1985. Brazilian Society for Soil Mechanics, 4–7. De Graft-Johnson, J. W. S., and H. S. Bhatia. 1969. General report—engineering characteristics. Proc. Specialty Session on Engineering Properties of Lateritic Soils, 7th Int. Conf. on Soil Mechanics and Foundation Engineering, Mexico, 1969, Vol. 2, 13–43. Frost, R. J. 1967. Importance of correct pre-testing preparation of some tropical soils, Proc. 1st Southeast Asian Regional Conf. on Soil Engineering, Bangkok, 43–44. Junus Dai, and P. M. Driessen. 1972. A general orientation preceeding the resumption of clay mineralogical studies at the soil research institute, Bogor, Indonesia. 2nd ASEAN Soil Conference, Jakarta, Indonesia. Kuno, G., R. Shinoki, T. Kondo, and C. Tsuchiya. 1978. On the construction methods of a motorway embankment by a sensitive volcanic clay. Proc. Conf. on Clay Fills, London, 149–156. Lenvain, J., D. Mulyadi, and A. Abdurachman. 1972. Artificial structure formation and aggregation in andosol, 2nd ASEAN Soil Conference, Jakarta, Indonesia. Little, A. L. 1969. The engineering classification of residual tropical soils. Proc. Specialty Session on Engineering Properties of Lateritic Soils, 7th Int. Conf on Soil Mechanics and Foundation Engineering. Mexico, 1969, Vol. 1: 1–10. Lohnes, R. A., and E. R. Tuncer. 1977. Engineering characteristics of andosols. Proc. 5th Southeast Asian Conf. on Soil Engineering, Bangkok, 305–312. Matyas, E. L. 1969. Some engineering properties of Sasamua clay. Proc. Specialty Session on Engineering Properties of Lateritic Soils, 7th Int. Conf. on Soil Mechanics and Foundation Engineering, Mexico, 1969, Vol. 1, 143–151.
34
EVALUATION, CHARACTERIZATION, AND CLASSIFICATION OF RESIDUAL SOILS
Mitchell, J. K., and N. Sitar. 1982. Engineering properties of tropical residual soils. ASCE Geotech. Engng. Specialty Conference on Engineering and Construction in Tropical and Residual Soils, Hawaii, 1982, 30–57. Morin, W. J., and P. C. Todor. 1975. Laterite and Lateritic Soils and Other Problem Soils of the Tropics, USAID 3682. Baltimore, MD: Lyon Associates. Moye, D. G. 1955. Engineering Geology for the Snowy Mountains Scheme. Australia: Institution of Engineers, Vol. 27, 287. Pender, M. J. 1971. Some properties of weathered greywacke. 1st Australia–New Zealand Conference on Geomechanics, Melbourne, Vol. 1, 423–429. Pender, M. J. 1980. Friction and cohesion parameters for highly and completely weathered Wellington Greywacke. 3rd Australia–New Zealand Conference on Geomechanics, Wellington, Vol. 1, 171–175. Ranganatham, B. V. 1961. Soil structure and consolidation characteristics of black cotton clay. Geotechnique 11(4), 333–338. Saunders, M. K., and P. G. Fookes. 1970. A review of the relationship of rock weathering and climate and its significance to foundation design. Engineering Geology 4, 289–325. Terzaghi, K. 1958. The design and performance of Sasamua dam. Proc. Institution of Civil Engineers 9, 369–394. Townsend, F. C. 1985. Geotechnical characteristics of residual soils, ASCE Journal of Geotechnical Engineering, 3(1), 77–94. Uehara, G. 1982. Soil science for the tropics. ASCE Geot Eng Specialty Conference on Engineering and Construction in Tropical and Residual Soils, Honolulu, Hawaii, 1982, 13–29. Wesley, L. D. 1973. Some basic engineering properties of halloysite and allophane clays in Java, Indonesia. Geotechnique 23(4), 471–494. Wesley, L. D. 1988. Engineering classification of residual soils. Proc. 2nd Int. Conference on Geomechanics in Tropical Soils, Singapore, 77–83. Wesley, L. D., and T. Y. Irfan. 1997. Classification of residual soils. In: Mechanics of Residual Soils, G. E. Blight, ed. Rotterdam, The Netherlands: Balkema, 17–29. Wirth, J. L., and E. J. Zeigler. 1982. Residual soils experience on the Baltimore subway. ASCE Geotechnical Engineering Specialty Conference on Engineering and Construction in Tropical and Residual Soils, Honolulu, Hawaii, 1982, 557–577.
CHAPTER 3
PORE PRESSURES AND SEEPAGE CONDITIONS ABOVE AND BELOW THE WATER TABLE
3.1
INTRODUCTION
The pore pressure state above the water table tends not to receive much attention in soil mechanics, possibly because it is not of great importance with sedimentary soils. However, as mentioned in Chapter 1, the pore pressure state above the water table is of considerable importance with residual soils, for two reasons. First, the water table is often deep and the zone of prime interest to geotechnical engineers is above the water table. Second, the high permeability of many residual soils means that the pore pressure state is not static. Seasonal changes in pore pressure, or those caused by isolated intense rainstorms, can be significant, and govern the behavior of the soil. For these reasons, we need to be quite clear what the water table is, and how the pore pressure and seepage state relate to it. There is a tendency among geotechnical engineers, which probably arises from the way the subject is taught and presented in textbooks, to think of the water table as a boundary line below which seepage and pore pressures exist, and above which they do not. This is not generally the case, especially in clays, as the following sections illustrate.
35
36
PORE PRESSURES AND SEEPAGE CONDITIONS ABOVE AND BELOW THE WATER TABLE
3.2
SITUATION AT LEVEL SITES
3.2.1
Static Case
The static pore pressure state and degree of saturation in relation to the water table, are illustrated for a level site in Figure 3.1. The pore pressure shown is the hydrostatic or equilibrium value, being negative above the water table and positive below it. No seepage flow occurs in this state. Above the water table the void space between the soil particles acts as fine tubes and water is drawn into this space, or retained in it, by capillarity or surface tension forces. In a fine-grained soil, made up entirely of clay-sized particles (smaller than 0.002 mm), the effective pore size will be about 20 percent of this, which is 0.0004 mm. The theoretical capillary rise in such a material would be about 75 m. Measurements in the field have shown that many clays remain fully saturated for many meters or tens of meters above the water table, and in wet or temperate climates only the top 1 or 2 m is less than fully saturated. This partial saturation occurs not because of gravity drainage downward, but because of water loss by evaporation at the surface. It is only when the material is of coarse silt or fine sand size that water will drain out of the void space from gravity forces alone. Even in fine sand, drainage will be limited, and considerable water will still remain in the void space. Only in very coarse sands and gravels will water drain almost completely from the void space.
Pore water pressure Negative Positive
Negative pore pressure
Partially saturated zone
Ground surface
u = −γwa Saturation boundary for a medium grained soil (silty clay or silt)
Fully saturated zone
a
Water table
Positive pore pressure
Saturation boundary for a fine grained soil (clay)
Saturation boundary for a coarse grained soil (silty sand, fine sand) Hydrostatic (equilibrium) pore pressure
b
u = γwb
Figure 3.1 Pore pressure (and saturation states) in relation to the water table.
SITUATION AT LEVEL SITES
3.2.2
37
Seasonal Effects
During wet weather, or during the wet season (in countries having distinct wet and dry seasons), water will seep into the ground and the pore pressures will rise, at least close to the surface. During actual rainfall, the pore pressure at the surface becomes atmospheric (or marginally above atmospheric during intense storms), so that a steep downward hydraulic gradient is created just below the surface, drawing water into the soil. During dry weather, evaporation will occur, creating high negative pore pressure at the surface and a steep upward hydraulic gradient, causing seepage flow toward the surface. If the soil is coarse-grained, water may drain out of the soil under gravity, and be replaced with air. There is normally a limit to the depth to which these weather effects penetrate below the ground surface, a fact that tends to be lost sight of within the geotechnical community. An assumption is made that the pore pressure below the water table will be hydrostatic at all times, and will rise or fall in accordance with the water table level. This is definitely not the case in clays. The depth limit of weather effects depends primarily on the properties of the soil, especially its permeability and compressibility. We will examine the way the water table and pore pressures change by considering three cases: coarse free-draining materials, very low permeability clays, and clays of medium to high permeability. The mechanics of pore pressure and water table change are quite different for coarse-grained soil and fine-grained soil. 3.2.3
Coarse-Grained Soils
This is the simplest case, and is illustrated in Figure 3.2. Sand (or gravel) acts like a reservoir, into which water flows under gravity during wet weather, causing the water level to rise, and from which water is lost by evaporation, or lateral flow, during dry weather, with an accompanying drop in water table. Below the water table the soil is fully saturated and the pore pressure state is hydrostatic. Above the water table the soil has a very low (almost zero) degree of saturation, and the pore pressure is essentially zero. Seasonal changes extend at least to the bottom of the sand layer. In many situations, the water table in the sand may be governed by external controls, such as the proximity of a river, rather than by the influence of weather at the surface. With coarse-grained soils, the properties governing the way the water table changes with seasons are the permeability and the porosity (storage capacity) of the soil. Volume change can be assumed to be negligible for most situations, so that the compressibility of the material is not an influencing factor. 3.2.4
Low-Permeability Clays
Because of their low permeability, the seasonal influence is not very deep, and may not reach the water table, as illustrated in Figure 3.3. The water
38
PORE PRESSURES AND SEEPAGE CONDITIONS ABOVE AND BELOW THE WATER TABLE
PORE PRESSURE Negative Positive
PORE PRESSURE Positive Negative
Water trickles downward towards water table
Water evaporates and travels upward towards the surface
Water table Water table variation
Water table
Hydrostatic pore pressure below the water table
DRY WEATHER STATE - evaporation at surface
WET WEATHER STATE - rainfall on ground surface
Figure 3.2 Seasonal variation in water table and pore pressures: coarse-grained soil.
table thus remains constant, and changes in pore pressure occur only within the zone of negative pore pressure above the water table. During prolonged dry periods the ground may shrink and become partially saturated close to the surface. During wet weather, the soil will gradually reabsorb water and swelling will occur. The equilibrium state shown in Figure 3.3 is a transient
Ground surface
PORE PRESSURE Negative Positive
Wet weather state
Dry weather state Limiting values of pore pressure change Depth limit of seasonal changes
Equilibrium (hydrostatic) pore pressure state
Zone of negative pore pressure
Water table (constant)
Figure 3.3 Seasonal variation in water table and pore pressures: low-permeability clay.
SITUATION AT LEVEL SITES
39
state that can be expected to occur briefly from time to time between wet and dry spells, or when the seasons are changing. Most heavily overconsolidated sedimentary clays, such as London clay, belong in this category. 3.2.5
Medium- to High-Permeability Clays
This case is illustrated in Figure 3.4. The seasonal influence now extends beyond the water table, which rises in the winter and falls in the summer. There is still a depth limit beyond which the pore pressure does not change. The equilibrium pore pressure shown in the figure is hydrostatic with respect to the average position of the water table; as with the previous case, it is a transient state that will occur from time to time, especially midway between the wet and dry seasons. At other times the pore pressure will not be hydrostatic, either below or above the water table. To summarize the situation with clays we should note the following points: 1. There is a limit to the depth influenced by seasonal effects. 2. The pore pressure is not necessarily hydrostatic below (or above) the water table, except transiently in the “average” situation between the extremes caused by the seasonal influence. 3. The mechanics governing seasonal pore pressure changes in clay are essentially the same as those governing consolidation or swelling of PORE PRESSURE Positive Negative Dry season:hydraulic gradient upward, seepage upward Winter water table
Wet season:hydraulic gradient downward, seepage downward
Water table variation
Average water table Summer water table
Limit of seasonal changes
P Hydrostatic (equilibrium) state
Figure 3.4 Seasonal variation in water table and pore pressure: medium- to high-permeability clay.
40
PORE PRESSURES AND SEEPAGE CONDITIONS ABOVE AND BELOW THE WATER TABLE
clay, which the reader is probably familiar with. The governing parameters are thus permeability and compressibility of the soil, or, in their combined form, the coefficient of consolidation. 4. In contrast to the situation with coarse-grained material, volume change is an essential component in the mechanics of pore pressure and water table change. It is not difficult to determine curves of the form shown in Figure 3.3 and 3.4 using normal Terzaghi consolidation theory and the familiar solutions provided in the form of charts. This is done in the following section. 3.2.6 Use of Terzaghi Consolidation Theory to Illustrate Seasonal Influence
Consider the situations illustrated in Figure 3.5, which shows a layer of fully saturated soil 8 m thick overlying impermeable rock. In Figure 3.5a the water table is at a depth of 6 m and in Figure 3.5b it is at the ground surface. In both cases the pore pressure is in hydrostatic equilibrium with the water table. This equilibrium situation is upset in (a) by a continuous shower of rain, and in (b) by the placing of a layer of fill at the surface. The rain reduces the pore pressure at the surface to zero, and if the rain continues it will eventually create a new equilibrium situation with the water table
Rainfall applies a pore pressure of zero at the ground surface
Fill 2 m thick applies a stress of 40 kPa Ground surface Equilibrium water table in beg tion ida sol con re en ssu wh re pre re d ssu o a p e pre re ess m h c 8 x Po 0 e e . 4 ial sur Init kPa = res ep 40 por ium libr
en
wh
qui
al e
Fin
re
ssu
pre
e sur res ep por ium libr qui e al e sur Fin res ep a por kP ess 58.8 = exc ial ead h Init ins m beg =6 ng elli sw
re
Po
Equilibrium water table
6m 8m
s
Impermeable rock (a) Rainfall influence on pore pressure state
Impermeable rock (b) Fill layer influence on pore pressure state
Figure 3.5 Pore pressure changes induced by rainfall and by a uniform surface load.
SITUATION AT LEVEL SITES
41
at the surface. The fill, on the other hand, induces a new uniform excess pore pressure throughout the layer, which will dissipate until it returns to its original position. In both cases, there is an excess pore pressure, which is the value in excess of the final equilibrium value in each case. There is also a boundary condition at the surface that initiates the process of pore pressure change. In (a) it is an instant increase from the initial negative value to zero, and in (b) it is an instant decrease from the initial positive value to zero, so that the mechanics by which the pore pressure change progresses through the layer with time are identical in each case. We can therefore use the Terzaghi charts to determine the changes in pore pressure with time. The reader will already be familiar with the normal consolidation case, and we will therefore analyze only the rainfall case in Figure 3.5a. We will start by determining the rate of rise of the water table. This is done in Table 3.1, by calculating the time for the water table to rise by 1-m intervals. For convenience, we will express the pore pressure in meters of head, rather than kPa. The initial excess pore pressure was 6 m, and when the water table rises to a depth of 5 m the excess pore pressure at this depth is now 5 m head, so the u/uo value = 5/6 = 0.83 and so on at the other depths. We will assume that the coefficient of consolidation of the soil is 1.0 m2 /day. This is a very high value for sedimentary clays but not particularly rare in residual soils (see Table 4.2 of Chapter 4). Apart from the very long rainfall duration needed for the water table to reach the ground surface, the table shows a rather unexpected trend. The time for the water table to rise from a depth of 6 to 3 m is almost 10 days but for it to rise the remaining 3 m only takes about 2 days. This seems surprising at first sight, but if we examine the form of the pore pressure contours we find that it is to be expected. Table 3.2 shows the calculation of the pore pressure profile when t = 11.2 days, the time at which the water table reaches the ground surface. Similar calculations have been done for times of 4, 8, 25, and 58 days; the results are shown in Figure 3.6. Figure 3.6a shows the pore pressure profiles at a number of time steps. This illustrates why the water table rises very rapidly once it approaches
Table 3.1 Calculation of Rate of Water Table Rise Water Table Depth (m) 6 5 4 3 2 1 0.5
u/uo
z/H
Time Factor T
Time t (Days)
1.0 5/6 = 0.83 4/6 = 0.67 3/6 = 0.50 2/6 = 0.33 1/6 = 0.17 0.5/6 = 0.085
6/8 = 0.75 5/8 = 0.63 4/8 = 0.50 3/8 = 0.375 2/8 = 0.25 1/8 = 0.125 0.5/8 = 0.063
— 0.105 0.132 0.155 0.165 0.170 0.175
— 6.7 8.4 9.9 10.6 10.9 11.2
42
PORE PRESSURES AND SEEPAGE CONDITIONS ABOVE AND BELOW THE WATER TABLE
Table 3.2 Calculation of Pore Pressure Profile at Time = 11.2 Days Depth (m) 1 2 3 4 5 6 7 8
z/H
u/uo
Excess Pore Pressure (m)
Actual Pore Pressure (m)
0.125 0.25 0.375 0.50 0.625 0.75 0.875 1.00
0.165 0.33 0.465 0.59 0.69 0.755 0.795 0.816
0.99 1.98 2.79 3.54 4.14 4.53 4.77 4.90
0.01 0.02 0.21 0.46 0.86 1.47 2.34 3.10
mid-height. The change in actual pore pressure values between 8 days and 11 days is not great but because the values are close to atmospheric, only a small change is needed to move them from negative to positive with an accompanying large rise in the water table. Figure 3.6b shows the rise in water table and also the rise in the pore pressure at the impermeable boundary at the bottom of the clay layer. The rise in the latter has hardly started when the water table reaches the ground surface. The time needed to bring the water table to the surface is surprisingly long despite the high value of the coefficient of consolidation used in the analysis. However, it should be appreciated that some residual soils, especially those of volcanic origin, can have values one or two orders of magnitude higher than this, in which case the times involved would be reduced by the same order. We can use the same method to examine the way changes occur in dry weather; however, in this case the magnitude of the negative pore pressure at the surface induced by evaporation is unknown and we would have to guess a value. This analysis is simplistic and somewhat artificial, but has been included here to illustrate the mechanism by which pore pressures and the water table in clays change during rainfall. In practice, there may not be a lower impermeable boundary, and the depth of influence of the rainfall may not reach either the water table or the impermeable boundary. However, the analysis illustrates the mechanics involved and also demonstrates that focusing only on the water table is not a reliable way to evaluate pore pressures in the ground. We need to look at the complete picture. The issue of rainfall influence on pore pressures and water table depth will be considered further in relation to the stability of slopes in Chapter 8. 3.2.7
Field Records of Seasonal Effects
Field evidence confirming the pattern described above has been obtained from measurements on a number of sites, such as for example Kenny
SITUATION AT LEVEL SITES
43
Pore pressure head (m) Negative Positive −6
0
−4
−2
0
2
4
6
8
1
2 4
Init
re
ssu te
sta
tate
5
pre
in 25 da ys 58
es
sur
Depth (m)
e
res
4
re
ep
por
11 Ti m
o al p Fin
8
ial
3
Initial water table
6
7 Hard impermeable rock 8
Pressure head (m) at base of clay
(a) Pore pressure profile changes with time. 8
4
0
10
20
Pore pressure at base of clay 30 40 50
60
70
Water table depth (m)
Time (days) 4 Water table depth 8 (b) Rise in water table and pore pressure at base of clay layer with time
Figure 3.6 Pore pressure and water table changes in fully saturated clay from continuous rainfall.
44
PORE PRESSURES AND SEEPAGE CONDITIONS ABOVE AND BELOW THE WATER TABLE
Pore pressure changes with time and depth
Changing (transient) pore pressures
Dry season
Wet season
Depth
Time
Limit of seasonal changes
Constant (steady state) pore pressures
Figure 3.7 Decline of seasonal pore pressure variation with depth (adapted from Pun and Urciuoli 2008).
and Lau (1984), Urciuoli (1998), and Pun and Urciuoli (2008). Pun and Urciuoli (2008) installed piezometers at three different depths at a site in Italy and recorded pore pressures over a 10-year period. These showed regular seasonal changes that progressively decrease with depth; they are shown conceptually in Figure 3.7. This is the same behavior as portrayed in Figure 3.3 presented in a different form. We can note that the behavior in Figure 3.7 is not related to any particular water table position. The water table could be below the limit of seasonal changes, or could be fluctuating within the zone of seasonal changes. 3.3
HILL SLOPES, SEEPAGE, AND PORE PRESSURES
In the previous section we considered only level sites and assumed no horizontal movement of water. In most natural situations this will not be the case; both the ground and the water table (phreatic surface) will be sloping. Seepage will be occurring steadily in the downhill direction, as illustrated in Figure 3.8. The slope is assumed to consist of fully saturated clay, except possibly for a shallow zone at the surface. Water can enter the slope from two possible sources. One is surface rainfall, and the other is
from e ion surfac t a r filt und f in e o on gro n o Z fall rain ce) surfa reatic h p r (o ter
Wa
e
tabl
Limit
ging e mer ge e d surfac a p See groun e at th
nges
al cha
son of sea
45
Seepage coming from adjacent catchment
HILL SLOPES, SEEPAGE, AND PORE PRESSURES
Figure 3.8 Seepage in a hillside coming from direct rainfall and from adjacent catchment.
seepage from an adjacent rainfall catchment. Water emerges again at the lower part of the slope and the valley floor, and contributes to the flow in the streams and rivers normally found in valleys. The dotted arrows indicate the likely seepage pattern. Although this is an accurate portrayal of the general shape of the seepage pattern in a hill slope, it is not strictly correct near the ground surface, where the slope will be subjected to the same seasonal influences described above for level sites. In some slopes the water table may be constant all year round, while in others it will fluctuate with seasonal influences. As with level sites, there will be a lower limit to the zone of seasonal changes, as indicated in the figure. Figure 3.9 shows two possible seepage patterns that are valid for the same water table. Both flow nets have been obtained using the computer program SEEP/W. Figure 3.9a shows the seepage pattern (flow net) normally assumed to apply for a water table in the position shown. This assumption is somewhat odd, or at least illogical, as it implies that recharge is coming only from a source distant to the slope, or, in other words, that rain falls anywhere except on the slope itself. Figure 3.9b shows a valid flow net assuming the hill is double sided, so that the only recharge into the slope is coming from rain that falls directly on the slope itself. In this case the line a–b becomes a catchment divide, which means an impermeable boundary with respect to the flow net. The valley is assumed to be symmetrical so that the line c–d is also an impermeable boundary. The adopted rainfall recharge rate on the slope is
46
PORE PRESSURES AND SEEPAGE CONDITIONS ABOVE AND BELOW THE WATER TABLE
a
ce
rfa
n
ou
Gr
u ds
ter table
red wa
Measu
d
b
c (a) Normal assumption of flow net for the given water table (implies an external recharge source)
a ted mi arge i l of ch ne re Zo nfall i ra
Zone of negative pore pressure
Zone of positive pore pressure
d
c
b (b) A valid flow net for limited recharge from rainfall on the slope
Figure 3.9 Valid seepage patterns for the same water table.
PERMEABILITY OF RESIDUAL SOILS
47
less than the maximum capacity the soil can accept. In this situation the water table will be some distance below the ground surface, and there will be a zone of negative pore pressure above the water table. The flow net in Figure 3.9b is likely to better represent the true seepage pattern than that in Figure 3.9a, although in practice the situation will probably be somewhere between the two, and similar to that illustrated in Figure 3.8.
3.4
PERMEABILITY OF RESIDUAL SOILS
Generalizations are always risky in soil mechanics, but as already indicated, it is certainly true that residual soils tend to have substantially higher permeability than sedimentary soils. This is due to microstructural features, such as the aggregation of clay particles into clusters, and the ability of bonds between particles to create a very open structure. Remolding and compacting residual soils tends to destroy this structure and generally results in a significant decrease in permeability. It should be noted also that permeability does not generally correlate well with particle size as it does in silts and sands and in some sedimentary clays. Table 3.3 gives values of the coefficient of permeability for a range of residual soils.
Table 3.3 Coefficient of Permeability (Hydraulic Conductivity) of Several Residual Soils Soil Type Parent Rock Granite Gneiss Basalt
Description
Coefficient of Permeability (m/sec) Young (Saprolitic) 4 × 10-3 to 5 × 10−9 5 × 10-6 to 1 × 10−7 3 × 10-6 to 1 × 10−9
Sandstone Gray clay Andesitic lahar/ Tropical red clay volcanic ash (halloysitic) Volcanic ash Volcanic ash clay 10-6 to 10−7 (allophane)
Mature (True Soil)
Remolded
4 × 10-6 to 5 × 10−9 5 × 10-6 to 1 × 10−6 —
—
10-9 to 10−6 1 × 10−9
10-10 to 10−7 0.3–3 × 10−10
5 × 10-7 to 10−8
10-11 to 5 × 10−10
— —
48
PORE PRESSURES AND SEEPAGE CONDITIONS ABOVE AND BELOW THE WATER TABLE
3.5 SIGNIFICANCE OF THE WATER TABLE (OR PHREATIC SURFACE)
The significance of the water table (or phreatic surface) in fine-grained soils should be clearly understood. In particular, the following points should be noted: 1. The water table is not a boundary line below which the soil is fully saturated and above which it is unsaturated or partially saturated. 2. It is not a boundary below which seepage is occurring and above which there is no seepage. 3. It is simply a line of zero (atmospheric) pore pressure, below which pore pressures are positive and above which they are negative. 4. The water table does not constitute a discontinuity in the seepage pattern. Seepage occurs in a continuous fashion above and below the water table according to the same physical laws (provided the soil remains fully saturated). 5. To obtain a reliable picture of the seepage situation in any slope, it is not sufficient to measure only the depth of the water table. Measurements of pore pressure need to be made at a number of locations over a long period of time.
3.6 IMPLICATIONS OF THE GROUNDWATER AND SEEPAGE STATE ABOVE THE WATER TABLE FOR PRACTICAL SITUATIONS
Various practical issues arise from the fact that in clays there is a negative pore pressure above the water table and it is subject to seasonal variations. The following are several important examples. 3.6.1 Errors in the Estimation of Foundation Settlement Using Conventional Methods
The normal procedure for estimating settlement resulting from construction of a building supported on surface foundations on clay involves two basic steps. The first is to measure the compressibility characteristics of the soil by laboratory or field testing, and the second is to estimate the resulting increase in effective stress in the soil. The latter step is not as simple as it appears, or is commonly assumed to be. If the water table is deep, then the initial pore pressure above the water table is not known with any degree of certainty. It may well be highly negative during dry summer weather and close to zero during wet winter weather. The construction of the building is likely to dampen down these variations as it will generally act as an impermeable boundary at the ground surface. The change in effective stress
IMPLICATIONS OF THE GROUNDWATER AND SEEPAGE STATE
49
governing settlement of the foundation will therefore be made up of two components: 1. The stress increase due to the foundation load 2. The stress change resulting from the change in the pore pressure in the ground, which may be positive or negative The latter is usually simply ignored in practice; in fact, most geotechnical engineers are probably unaware of its existence. However, its significance can be very important, as the net change in effective stress could be either positive or negative. If it is negative, the soil will swell and the foundation will rise rather than settle. An example is given in Section 4.8.2 illustrating the possible errors involved. 3.6.2
Ground Settlement Resulting from Groundwater Lowering
It sometimes happens that human activities, such as quarrying or deep excavations for highways (or other underground structures), result in the permanent lowering of the groundwater table in the surrounding area. This, in turn, causes an increase in effective stress in the soil and settlement of the ground surface. Depending on the nature of the ground, this settlement may be of little consequence, or it could be a major concern. It is important to appreciate that settlement will result not only from compression of the soil beneath the water table but also from compression of the layers above the water table. The pore pressure is likely to be lowered right through the soil profile, as illustrated in Figure 3.10 where the soil is assumed to be fine-grained uniform clay. Pore Pressure Negative Positive Ground surface
Initial water table H
Δhw Final water table re po
ore lp tia
Ini
al Fin
ure
ss pre
re
su
s pre
Figure 3.10 Change in pore pressure resulting from lowering of water table.
50
PORE PRESSURES AND SEEPAGE CONDITIONS ABOVE AND BELOW THE WATER TABLE
The groundwater table has been lowered by hw . This causes a change in pore pressure of γ w hw throughout the clay layer. The total stress is not significantly changed, so that there is an increase in effective stress equal to the change in pore pressure. The settlement of the ground surface is therefore given by δ = mv γw hw H where δ mv γw H
is is is is
the the the the
settlement, coefficient of compressibility, unit weight of water, and layer thickness.
3.6.3 Ground Settlement or Swelling Due to Covering the Ground Surface
It is possible that simply placing some form of barrier at the ground surface may result in “heave” from swelling of the soil, or even ground settlement. This will occur because any such barrier will tend to change the pore pressures from their current state and move them toward the hydrostatic state, as indicated in Figure 3.11. The figure shows a situation likely in a dry climate where the water table exists at a constant level due to recharge from a source some distance away where the climate is wetter. Evaporation at the surface creates a zone of high Pore Pressure Negative Positive
Ground surface Pore pressure state with uncovered surface
Water table
uil Eq ibr
Recharge from distant catchment
ium
Homogeneous clay
Upward seepage due to surface evaporation
Pore pressure state when surface is covered
re po ure
ss
pre
Figure 3.11 Pore pressure change resulting from covering the ground surface.
REFERENCES
51
negative pore pressure above the water table. Construction of a building or a high-quality pavement with an asphalt seal at the surface is likely to create an impermeable barrier at the surface and prevent evaporation. The result is that the pore pressure will tend toward the static equilibrium situation indicated in the figure. The accompanying decrease in effective stress will cause swelling of the soil. The opposite situation is also possible, although less likely, in which case settlement of the ground surface would occur. 3.6.4 Errors in Estimates of Slope Stability Ignoring Soil Suction Influence
Many slopes in residual soils are stable because much (or most) of the slope is normally above the water table. This means that much of shear strength of the soil is due to the “suction” in the pore water. This effect is often not recognized and is seldom taken account of in stability analysis. We will see later in Chapter 10 that rainfall can influence the stability of a slope, even if the water table remains below the toe of the slope. 3.6.5 Errors in Estimates of Slope Stability because of Simplified Assumptions Regarding the Seepage Pattern in the Slope
Most, if not all, computer programs allow the user to specify the pore pressure in the slope by simply defining the water table. The programs then use this information to estimate the pore pressure by calculating the vertical intercept between the water table and the point concerned. This implies that flow lines are horizontal and equipotential lines are vertical. For gentle slopes this is normally a reasonable assumption and also a conservative one. Slopes in residual soils can be large and quite steep, in which case the above assumption may not be a reasonable approximation and the calculated safety factors may be excessively conservative. This issue is considered further in Chapter 8. REFERENCES Kenny, T. C., and K. C. Lau 1984. Temporal changes of groundwater pressure in a natural slope of nonfissured clay. Canadian Geotechnical Journal 21(1): 138–146. Pun, W. K., and G. Urciuoli. 2008. Soil nailing and subsurface drainage for slope stabilisation. 10th International Symposium on Landslides and Engineered Slopes, Vol 1. Leiden, The Netherlands: CRC Press, 85–125. Urciuoli, G. 1998. Pore pressures in unstable slopes constituted by fissured clay shales. 2nd International Symposium on the Geotechnics of Hard Soils—Soft Rocks, Napoli, Vol.2, 1177–1185.
CHAPTER 4
CONSOLIDATION AND SETTLEMENT
4.1
INTRODUCTION
The most significant differences in behavior between residual and sedimentary soils are probably those associated with their consolidation characteristics. The conventional understanding or interpretation of the consolidation behavior of sedimentary soils is based on their formation mode, namely deposition in sea or lake, followed by compression due to self-weight. The very different mode of formation of residual soils, illustrated in Figures 1.1 and 1.2 of Chapter 1, means that this conventional understanding is of no relevance to residual soils. What is more, there is no tidy alternative framework for characterizing the consolidation behavior of residual soils. The comment was made in Chapter 1 (Section 1.2) that the use of the conventional e-log p plot for presenting the results of odometer tests on residual soils is an example of treating and interpreting their behavior as though they were sedimentary soils. The use of the e-log p plot in soil mechanics arises from the fact that the virgin consolidation graph of a sedimentary soil is approximately linear when plotted using a log scale for pressure. Residual soils do not undergo a sedimentation and consolidation process and do not have a virgin consolidation line, so there is no reason or justification for plotting the results of compression tests using the conventional e-log p plot. The use of such a plot can lead to serious errors in understanding and/or interpreting the compressibility characteristics of residual soils. For this reason we will begin this chapter by examining the e-log p graph in some detail. Somewhat surprisingly, we find that mistaken interpretation of compressibility behavior resulting from the log plot is not confined to residual soils. 53
54
CONSOLIDATION AND SETTLEMENT
4.2 INTERPRETATION OF STANDARD OEDOMETER TEST RESULTS AND THE ‘‘OMNIPOTENCE OF TRADITION’’
Figure 4.1 shows typical compression curves plotted using both log and linear pressure scales. Figure 4.1a shows the log plots: these are taken directly from the references shown in the figure, and include the determination of the preconsolidation pressure using the standard Casagrande construction. Figure 4.1b shows the same curves after they have been replotted (by the author) using a linear pressure scale. The graphs show several significant points: 1. The shape of the curves on the log plots (concave from below) suggests that the soil shows an initial zone of low compressibility followed by a steady transition to a zone of higher compressibility. 2. This behavior in turn suggests that there is a preconsolidation pressure separating these zones and that it is reasonable to apply the Casagrande construction to determine the preconsolidation pressure. 3. Examination of the curves plotted using a linear scale for pressure shows quite a different picture. The graphs are now concave from above, although the third one is very close to linear, except at high stress levels. There is thus no increase in compressibility with stress. The compressibility either remains constant or decreases as the stress level increases. 4. There is no longer any trace, on the linear plots, of the preconsolidation pressure determined from the log plots. There is simply no significant change of slope at the pressures identified as preconsolidation pressures according to the standard Casagrande construction. 5. The preconsolidation pressure determined from the log graphs is therefore not a soil property; it is an illusion created purely by the way the data are plotted. It is of course not the case that all log plots give a misleading picture of soil compressibility. If there is a clear and pronounced change of gradient at a particular stress level, then the same change will appear on a linear plot. This is especially true of many, if not most, highly sensitive soils. There is a further problem with e-log p plots that is apparent to some extent in the above figures, and that is the determination of compressibility parameters, in particular, the compression index Cc and the swell (or rebound) index Cs . In the absence of clearly defined linear sections in the graphs on the log plot, the determination of these parameters is a matter of judgment, or even pure guesswork. This difficulty is illustrated clearly in Figure 4.2, which shows the result of an oedometer test on a heavily overconsolidated sedimentary clay. Figure 4.2a shows the result in its original log form as presented by Lancellotta (1995), and Figure 4.2b shows it replotted by the author using a linear scale.
Pressure (linear scale)
Pressure (log scale) A
σ′c Void ratio
Void ratio
D B
after Craig (1992) σ′c
C Horizontal line
C
Tangent Cc B Cr
after Budhu (2000) σ′c
Void ratio
a
g
d c
after Das (1997)
A
σ′c
σ′c
b
f
Void ratio
Void ratio
A
σ′c
Bisector Void ratio
F
D
h
D
σ′c E
B
Void ratio
Void ratio
G F
after Wesley (2010) σ′c (a) Log plots
C (b) Linear (arithmetic) plots
Figure 4.1 Oedometer test results plotted using both logarithmic (log) and linear scales for pressure. 55
56
CONSOLIDATION AND SETTLEMENT
Stress range of interest to geotechnical engineers
Cc ?
0.8
0.8
0.6
Void ratio
Void ratio
Maximum stress level in most oedometer tests
Cs
0.6
0.4
0.4
0.2 25
100
1000
10000
Pressure (kPa)
(a) Logarithmic plot (after Lancellotta, 1995)
0.2
0
5000
10000
15000
20000
Pressure (kPa)
(b) Linear plot
Figure 4.2 Oedometer test on a highly overconsolidated clay (after Lancellotta 1995).
It would be tempting from the log plot to apply Casagrande’s construction to identify a preconsolidation pressure, and to carry on and determine the parameters Cc and Cs . There is uncertainty, however, in determining these parameters. The value of Cc is perhaps reasonably clear if we take it from the last part of the graph (on the log plot), which is linear over the last several readings. This determination is valid only for the stress range of this particular test, which was very high, up to 20,000 kPa. Most conventional oedometer tests are taken only to a pressure between about 800 and 1600 kPa. If this test had been limited to a lower stress level, then a straight-line section would be less apparent and the value of Cc would be substantially less, as indicated by the second line in the figure. The value of Cc for a soil is a commonly quoted parameter, and yet it seems to be both poorly defined and of little practical use. As originally conceived, it was the (log) slope of the virgin consolidation line of a clay sediment deposited in water and compressed by its own weight. It seems to have developed into meaning the slope of the tangent to the e-log p graph at the maximum stress to which the test is taken. It is then the slope of a line that lies almost entirely outside the measured data, as pointed out by Janbu (1998). It is thus of arbitrary value and of little or no discernable use, as it bears no relationship to the compressibility of the soil at stress levels relevant to foundation design. The determination of Cs is equally problematic. In Figure 4.1 none of the log graphs show a very well-defined linear section before the preconsolidation pressure is reached, and in Figure 4.2 the swell (rebound) line is
BEHAVIOR OF RESIDUAL SOILS
57
not parallel to the initial section of the consolidation line. The stress range of interest to a foundation designer is likely to be between about 50 and 400 kPa, so that the use of a Cs value determined from the rebound line would result in a gross overestimate of foundation settlement. It will be apparent from the above discussion that serious errors can easily arise, and indeed do arise routinely, in the interpretation of oedometer test results using the e-log p plot, even with sedimentary soils. In Chapter 1 reference was made to Terzaghi’s warning about “the omnipotence of theory”; perhaps we need a further warning about the “omnipotence of tradition”, as it is difficult to see any other explanation for the continued use of the e-log p graph apart from tradition. We will see later that there are also problems in determination of the coefficient of consolidation from oedometer tests, especially with residual soils.
4.3
BEHAVIOR OF RESIDUAL SOILS
4.3.1
Tropical Red Clay
The author’s first experience of the compression behavior of a residual soil was while working on the island of Java in Indonesia (in 1960, shortly after graduating). Java is dominated by a line of mainly andesitic volcanoes, and much of its soil cover is derived from the tropical weathering of volcanic deposits, especially lahar flow material and volcanic ash. Figure 4.3 shows the results of oedometer tests on three red clay samples derived from volcanic material, plotted using both log and linear scales for pressure. The graph using a log scale shows a “normal” shape, and possibly suggests a preconsolidation or yield pressure in the vicinity of 300 kPa. In
Pressure (kPa) 10
100
Pressure (kPa) 1000
0
250
500
750
1000
1.6 Compression (%)
Void ratio
2
1.4
4 6 8
1.2 10 (a) Log scale
(b) Linear scale
Figure 4.3 Oedometer tests on tropical red clay (Java, Indonesia).
1250
58
CONSOLIDATION AND SETTLEMENT
Compression %
0
100
Pressure (kPa) 200
300
2
4 5
Figure 4.4 Data from Figure 4.3 plotted over stress range relevant to foundation design.
direct contrast, the linear plot shows no suggestion at all of preconsolidation pressures. The point has already been made that stress history is of no relevance to residual soils and seeking preconsolidation pressures is a futile and misguided exercise. For residual soils the term yield pressure or vertical yield pressure should take the place of preconsolidation pressure. The behavior illustrated in Figure 4.3 is typical of tropical red clays, but should certainly not be regarded as typical of all residual soils. Some residual soils show very clear yield pressures. The red clays of Java are generally dense materials with liquidity indexes close to zero, and therefore unlikely to display any evidence of yield behavior. Before moving on to examine the behavior of other residual soils, we should note that the stress range likely to be of interest in foundation design is probably between zero and about 300 kPa. The undrained shear strength of many residual soils is between about 70 and 150 kPa, so that foundation pressures are likely to be in the range of 150–300 kPa. The data in Figure 4.3 are therefore shown again in Figure 4.4 as compression versus pressure on a linear scale, between zero and 350 kPa. Over this stress range the behavior is fairly close to linear, suggesting that the use of the linear parameter mv is a better choice than the log parameters Cc or Cs for estimating foundation settlement. Calculating the value of mv from the average slope of the graphs over the pressure range in Figure 4.3 is a simple procedure and unlikely to result in significant errors in the estimation of settlement. 4.3.2
Piedmont Residual Soil
Figure 4.5 illustrates the behavior of Piedmont soil in consolidation tests. Piedmont residual soils are found in the southeastern to mid-Atlantic region
2.0 B9-3M B9-4M B7-5M B8-7M B8-8M B7-9M
Void ratio
1.6
OCR = 4.0 OCR = 3.6 OCR = 3.4 OCR = 3.4 OCR = 3.3 OCR = 1.1
1.2
0.8
0.4
10
100
1000
10000
(a) Log scale Pressure (kPa) 0
500
1000
1500
2000
Compression (%)
10
20
30
(b) Linear scale
Figure 4.5 Consolidation behavior of Piedmont residual clay (adapted from Wesley 2000).
59
60
CONSOLIDATION AND SETTLEMENT
of the United States. They are derived from the weathering of Paleozoic metamorphic and igneous rocks, primarily schist, gneiss, and granite (Hoyes and Macari 1999). Figure 4.5a, taken from Hoyes and Macari (or Mayne and Brown 2003), shows oedometer test results plotted in the conventional manner using a log pressure scale. Figure 4.5b shows the same curves replotted using a linear pressure scale. While it is possible to infer preconsolidation or yield pressures from the log plot, the linear plot shows no evidence at all of such pressures. The indication of preconsolidation pressures in Figure 4.5a is purely the result of plotting the data on a log scale. Mayne and Brown (2003) discuss the oedometer test results from Piedmont soil and comment on the difficulty in determining the yield or apparent preconsolidation pressure. They quote experimental work by a number of authors who have determined preconsolidation pressures and apparent overconsolidation ratios (OCRs) for these soils. Apparent OCR values are stated to range from unity to approximately 4. Mayne and Brown (2003) also comment that the preconsolidation pressure or yield pressure might be better defined if improved sampling techniques were used. These attempts to determine apparent preconsolidation or yield pressures using e-log p graphs and the assumption that all soils should show such pressures are illustrations of the point made in Section 1.2 (Chapter 1), namely that the profession continues to expect residual soils to behave like sedimentary soils, and seeks to interpret their behavior within a framework developed for sedimentary soils. There is no basis at all for expecting all residual soils to display a yield or apparent preconsolidation pressure. As it happens, the expectation that all sedimentary soils should also show a preconsolidation pressure is also equally unfounded, as the behavior in Figure 4.2 clearly demonstrates. The behavior illustrated in Figure 4.5 is reasonably typical of behavior of a number of residual soils. Smooth curves are obtained whether a log or linear pressure scale is used; viewed from below they are concave using a log scale and convex using a linear scale. However, there is great variety in the behavior of residual soils and we will look at several more examples. 4.3.3
Waitemata Residual Clay
Figure 4.6 shows the behavior of a sample of a residual clay that covers a large area in and around the city of Auckland, New Zealand. The clay is derived from the weathering of a sandstone/mudstone formation known locally as the Waitemata series and is generally of moderate to high plasticity. The behavior is shown using both log and linear scales for pressure. The behavior is somewhat similar to that of the red clay in Figure 4.3. The log scale suggests the possibility of a yield pressure, but the linear scale again shows that this is a creation of the plotting method and not a property of the
BEHAVIOR OF RESIDUAL SOILS
61
Void ratio
1.1
1.0
0.9
0.8 10
40
100
400
1000
(a) Log scale Pressure (kPa) 0
500
1500
2000
Void ratio
1.1
1.0
0.9
0.8 (b) Linear scale
Figure 4.6 Consolidation test on a sample of Waitemata clay (found in Auckland, New Zealand).
soil at all. The linear plot shows a straight-line section, at least from zero to about 500 kPa, again indicating that the linear parameter mv is a more suitable means of expressing the compressibility than the log parameters. Figure 4.6 also illustrates the behavior during an unloading and reloading cycle after loading the sample to 200 kPa. This unloading cycle illustrates the point made earlier in Figure 4.2 that the slope of unloading and reloading cycles is not constant, so that the value of Cs , the swell coefficient, varies with stress level. The behavior illustrated in Figure 4.6 is the most common form of behavior found in Waitemata clays, especially in the nonsensitive
62
CONSOLIDATION AND SETTLEMENT
Pressure (kPa)
Vertical strain (%)
0
10
100
Pressure (kPa) 1000
0
500
1000
1500
5
10
14 (a) Log plot
(b) Linear plot
Figure 4.7 Further consolidation tests on Waitemata clay (taken from Pender et al. 2000).
or low-sensitivity materials. A further example is given in Figure 4.7, showing results selected from Pender et al. (2000). While the behavior shown in Figures 4.6 and 4.7 is typical of Waitemata clays, there are exceptions, especially in sensitive low-plasticity silt layers that are occasionally found interbedded within the clay. Results of consolidation tests on this material are illustrated in Figure 4.8. In this case, the impression created by the log plot that there is a yield pressure in the vicinity of about 200–300 kPa is confirmed by the linear plot. The linear plot shows clearly that there is an increase in compressibility at this stress level. This particular soil sample had a sensitivity of about 10, so it exists in a noncompact state and the presence of a yield pressure is not surprising. The normal interpretation put on this behavior is that the yield pressure marks the point at which the structure of the soil begins to disintegrate and it is thus less able to withstand the applied stress. Residual soils that demonstrate both yielding behavior and nonyielding behavior within the same profile are not uncommon. The volcanic ash soils described in the next section are a further example. 4.3.4
Volcanic Ash (Allophane) Soils
A detailed description of volcanic ash soils is given in Chapter 11 and only an elementary observation of their compression behavior will be made here. Figure 4.9 shows the results of consolidation tests made on three samples of volcanic ash clay taken as bulk samples from shallow depths in Java, Indonesia. The log plot suggests the behavior is rather similar, whereas the linear plot clearly illustrates three distinct types of behavior. Sample A shows a well-defined yield pressure of about 250 kPa. Sample B shows almost linear behavior, and sample C shows steadily increasing stiffness with stress level.
BEHAVIOR OF RESIDUAL SOILS
63
1.4
Void ratio
1.2
1.0
0.8
10
100
1000
(a) Log plot Pressure (kPa)
Compression (%)
0
250
500
750
1000
10
20 (b) Linear plot
Figure 4.8 Consolidation tests on sensitive silt found within the Waitemata clays.
4.3.5 Summary of Principal Aspects of Compression Behavior of Residual Soils
We can summarize this discussion of the compression behavior of residual soils by emphasizing the following points: 1. There is no sound basis, from either theoretical or practical considerations, for portraying the compression behavior of residual soils using a log scale for pressure.
64
CONSOLIDATION AND SETTLEMENT
Pressure (kPa)
Pressure (kPa) 10
100
1000
0
5000
100
200
300
400
500
A A
2 B
B 2
1
C
Compression (%)
Void ratio
3
4
C
6 8 10 12 14
(a) Log scale
(b) Linear scale
Figure 4.9 Consolidation tests on three volcanic ash soils (in Java, Indonesia).
2. The continued application of the e-log p plot to residual soils is a source of routine misunderstanding of the compression behavior of these soils. It is an example of the “omnipotence of tradition.” Even with sedimentary soils it is a frequent source of misunderstanding. 3. A more appropriate way of illustrating the compression behavior is shown in Figure 4.10. This shows three types of behavior regularly found in residual soils (and indeed also in sedimentary soils), namely linear, strain-hardening and strain-softening behavior. All three types of behavior can be found within the same residual soil type, even within the same profile. Despite the comment at the start of this chapter about the lack of a behavioral framework for understanding the consolidation behavior of residual soils, Figure 4.10 does provide at least the semblance of a framework for both residual and sedimentary soils. This is the relationship between the liquidity index and the shape of the compression curve. Soils with a low liquidity index generally show strain-hardening behavior with no trace of a yield pressure, while those with a high liquidity index generally show strain softening with a distinct yield pressure. At higher stresses, well beyond the yield pressure, the latter will also show strain-hardening behavior. 4. The linear parameter mv is a much more appropriate parameter to use in expressing the compressibility of residual soils. Despite the behavior illustrated in the above examples and summarized in Figure 4.10, it is often the case that the value of mv is reasonably constant over the pressure range of interest in settlement estimates.
CONSOLIDATION BEHAVIOR AFTER REMOLDING
65
Vertical yield pressure
Pressure Strain hardening is typical of dense soils with low liquidity index
Strain hard
ening
en oft ns
Strain
Lin
ra i St
Yield from structural breakdown
ing
Strain softening is typical of nondense soils with high liquidity index
ear
Figure 4.10 A conceptual illustration of the general compressibility behavior of soils.
4.4
CONSOLIDATION BEHAVIOR AFTER REMOLDING
It may be of value at this stage to consider the consolidation behavior of residual soils after thorough remolding and also after mixing with water to form a slurry. This can give us a useful picture of the influence that structure has on their behavior. Two soils have been examined in this way—a tropical red clay and a volcanic ash clay. Three consolidation tests have been carried out: the first is for the soil in its undisturbed state, the second is for the soil after thorough remolding, and the third is for the soil after mixing it to form a slurry. Consolidation curves are presented using both the conventional log scale and a linear scale. Figure 4.11 illustrates the behavior of a tropical red clay. The curve from the remolded soil is essentially the same as for the undisturbed soil, which is rather surprising. Remolding normally has a significant influence on the stiffness of most soils. The absence of any significant effect from remolding indicates that structure does not play a significant role in influencing the properties of the undisturbed soil. The consolidation line from the slurry sample has the usual linear form on the log plot and lies well above the undisturbed curve. Even at the maximum pressure to which the test was taken (2500 kPa) this line still lies well above the curve for the soil in its undisturbed or remolded state. This line should not be thought of as the “virgin” consolidation line of the original soil; as pointed out in Chapter 1 the undisturbed soil does not have a virgin consolidation line since it has not undergone a consolidation process. The fact that the slurry
66
CONSOLIDATION AND SETTLEMENT
2.5
2.5 Undisturbed Remolded Slurry
eLL
2.0
Void Ratio
Void Ratio
eLL
Undisturbed Remolded Slurry
ePL enat 1.5
2.0
ePL enat 1.5
1.1 10
1.1 100
1000
3000
0
Pressure (kPa)
1000
2000
2600
Pressure (kPa)
(a) Log scale
(b) Linear scale
Figure 4.11 Consolidation tests on a tropical red clay: undisturbed, remolded, and as a slurry suspension.
enat
enat
3.5 eLL
3.5 eLL Void Ratio
Void Ratio
consolidation line lies well above the undisturbed line indicates that in its natural state this is a dense soil, due presumably to the nature of the weathering process that formed it. Figure 4.12 illustrates the results from a sensitive volcanic ash clay. This soil behaves in an opposite manner to the red clay of Figure 4.11. Both the remolded and slurry samples have consolidation curves lying well below that for the undisturbed sample. The soil is therefore highly structured and
3.0 ePL
Undisturbed Remolded Slurry
Undisturbed Remolded Slurry
3.0 ePL
2.5
2.5
2.1 10
2.1 100 Pressure (kPa)
(a) Log scale
1000
3000
0
1000
2000
2600
Pressure (kPa)
(b) Linear scale
Figure 4.12 Consolidation tests on a volcanic ash clay: undisturbed, remolded, and as a slurry suspension.
TIME RATE AND ESTIMATION OF THE COEFFICIENT OF CONSOLIDATION
67
exists in a noncompact state. The effect of remolding is to destroy this structure and render the soil much less capable of withstanding stress than when in its undisturbed state. The structural influence is clearly significant and accounts for the well-defined yield pressure, and also the fact that after remolding the soil has lost most of its strength and its behavior is very similar to that of the soil after it has been mixed with water to form a slurry. To assist in understanding the behavior in Figures 4.11 and 4.12, the void ratios corresponding to the Atterberg limits and the natural water content are marked on the void ratio scale. It is seen that the red clay exists at a void ratio below the plastic limit, and thus has a very low liquidity index, while the volcanic ash clay exists at a void ratio above the liquid limit, and thus has a very high liquidity index. Figures 4.11 and 4.12 thus confirm the relationship between liquidity index and compression behavior illustrated in Figure 4.10. 4.5
VALUES OF STIFFNESS PARAMETERS FOR RESIDUAL SOILS
Table 4.1 shows values of Young’s modulus and the coefficient of compressibility (mv ) from a range of residual soils. Some of the data are from the author’s own files, and some from various other sources. The values are for the pressure range from zero to about 400 kPa, this being the range likely to be of relevance to foundation design. The values are also plotted in Figure 4.13. It is apparent from this information that the stiffness of many residual soils is fairly similar. Values ranging from about 1 × 10−4 to 2.5 × 10−4 kPa−1 are common. This is perhaps not too surprising as the undrained shear strength of residual soils is generally confined to a reasonably small range, between about 70 and 200 kPa. 4.6 TIME RATE AND ESTIMATION OF THE COEFFICIENT OF CONSOLIDATION
Residual soils are frequently of significantly higher permeability than sedimentary soils, which means that their coefficient of consolidation (cv ) values are also higher. This influences their behavior both in the field and in standard oedometer tests. In particular, it makes the determination of cv values of residual soils from standard oedometer tests somewhat problematic. The most common method for determining cv is probably that of Taylor, which uses a square root of time plot. The theoretical shape of root time plots for samples 20 mm thick, based on Terzaghi’s consolidation theory, is shown in Figure 4.14. The points on the graphs indicate the time intervals at which readings are normally taken, at least when the recording is done manually.
68
Tropical red clay
Soil Description
Volcanic ash (allophane) clay Sandstone Gray clay Basalt Basalt saprolite Granite Granite saprolite Granite Granite saprolite Gneiss Gneiss saprolite Gneiss Gneiss saprolite Gneiss Gneiss saprolite Gneiss Gneiss saprolite Gneiss Gneiss saprolite Gneiss Gneiss saprolite Gneiss Gneiss saprolite Gneiss Gneiss saprolite Piedmont formation Schist, gneiss, and granite
Andesitic lahar/ volcanic ash Volcanic ash
Parent Rock
Soil Type
47–84 10–13 20–60 9–10 28
Young’s Modulus E (MPa) from Plate Loading Tests
0.30–0.45 0.9 0.6–2.5 0.7–4.5
Wesley De Mello (1972) Lumb (1962) Hui (1972) Sandroni (1981) Werneck et al. (1979) Garga & Costa (1977) Napoles Neto (1954) Vargas (1979) Azevedo (1972) Campos (1980) Werneck et al. (1979) Hoyes & Macari (1999), Barksdale et al. (1982)
1–2 0.7–3.3 0.5–5 0.4–1.0 0.1–0.17 0.7–0.9 0.13–0.42 0.8–0.9 0.30
Wesley
Wesley
Reference Source
0.7–2
1–2
Coefficient of Compressibility mv (10−4 kPa−1 ) Oedometer Test Estimated from 0–300 kPa Plate Loading Test
Table 4.1 Compressibility Parameters for a Range of Residual Soils
Piedmont soil Volcanic ash (allophane) clay Red (halloysite) clay Basalt saprolite Weathered Sandstone
Granite saprolite
Gneiss saprolite
0
1 2 3 4 Coefficient of compressibility, mv (10−4 kPa−1)
Figure 4.13 Values of coefficient of compressibility from a range of residual soils.
Square root of time (min0.5) 0
0
1
2
3
4
5
Cv =0 .00
40
ec
2 /s
m
2 ec m /s
c 0.01
2 /sec
80
1c
Cv =
60
Cv = 0.1 cm
Average degree of consolidation
20
100
Figure 4.14 Root time plots for 20-mm-thick oedometer samples.
69
70
CONSOLIDATION AND SETTLEMENT
This figure shows that when the coefficient of consolidation is below about 0.01 cm2 /sec, the number of readings obtained will scarcely be sufficient to define the straight section of the graph that is fundamental for determining the cv value. Figure 4.15 shows root time graphs from three different residual soils. None of these graphs show the straight-line section predicted by consolidation theory illustrated in Figure 4.14. The reason for this is simply that the pore pressures dissipate so rapidly that the shape of the curves is no longer governed by the mechanics of pore pressure dissipation. It is probable that pore pressures are fully dissipated even before the first reading is made. With manual recording, it is not possible to take readings at closer intervals than those in Figure 4.15, although it is apparent from the shape of the curves that even continuous recording of deflection would not significantly alter the appearance of the curves. There is thus an upper limit to the value of the coefficient of consolidation that can be measured in conventional consolidation tests, this limit being approximately 0.1 m2 /day (= 36.5 m2 /year = 0.012 cm2 /sec.). Readings taken in the first minute will lie on a straight line only if the cv value is less than this. If reliable values of cv are required for soils with higher values, it is necessary to use a different method of measurement, such as a pore pressure dissipation test in a triaxial cell. Values of cv for the soil types in Figure 4.15 are shown in Table 4.2.
√ time 0
Compression (%)
2
√ min 2
4
6
Volcanic ash soil Waitemata clay (weathered sandstone) Tropical red clay
4
6
8
10
Figure 4.15 Root time graphs from oedometer tests on three residual soils.
TIME RATE AND ESTIMATION OF THE COEFFICIENT OF CONSOLIDATION
71
Table 4.2 Values of cv for the Soil Types in Figure 4.15 Soil Type Waitemata silts and clays Indonesian red clays Volcanic ash soils
Coefficient of Consolidation (m2 /day) 0.01–10 0.07–0.7 0.01–200
These values cover a wide range and lie both above and below the value of 0.1 m2 /day that can be measured in the standard consolidation test. We should note that these values are from laboratory tests, either conventional oedometer tests or triaxial pore pressure dissipation tests; the values steadily decrease as the stress level rises. The high values are obtained from the initial low stress increments and are thus values most likely to apply to the soil in its natural state. Because graphs of the shape shown in Figure 4.15 are common with residual soils, and not uncommon with sedimentary soils, their physical significance should be clearly understood. The important points are the following: 1. If the oedometer test does not produce a linear section of the deformation versus root time plot, it means that the deformation rate is not governed by pore pressure dissipation, and the coefficient of consolidation is greater than the limiting value mentioned above. 2. In terms of conventional concepts of primary and secondary consolidation, nearly all the compression is secondary compression, as it is occurring under constant effective stress. The curves therefore reflect the creep behavior of the clay when tested in this particular way. 3. The shape of the curves is not primarily a property of the soil. It is a function of the drainage path length in the test. If the soils in Figure 4.15 were tested in an oedometer with a much greater sample thickness (say 1 m instead of 20 mm), then the root time plots would almost certainly show clear linear sections in accordance with consolidation theory, followed by a section representing secondary consolidation. The proportions of settlement made up of primary and secondary settlement would then be very different, and governed by the thickness of the sample tested. 4. In practice, soil layers are generally thick, and the drainage path length may be several orders of magnitude greater than in the oedometer. The absence of primary consolidation in the laboratory test is therefore not an indication that there will not be primary consolidation in the field. Similarly, the presence of a large proportion of secondary consolidation in the oedometer test does not necessarily indicate that secondary consolidation will be important in the field.
72
CONSOLIDATION AND SETTLEMENT
√ time 0
√ min 2 √ t90(??)
4
6
√ t90
Compression (%)
2
4 Correct
6
Samp
le A
8 Incorrect
Sample B
10
Figure 4.16 Correct and incorrect interpretations of root time graphs from oedometer tests.
Mention should be made at this point of what appears (in the author’s experience) to be a common mistaken interpretation of the root time plot. This is illustrated in Figure 4.16, which shows the results of tests on two samples. Sample A, with a low value of the coefficient or consolidation, √ conforms to expected behavior, and the construction for determining t90 has been correctly applied. Sample B, on the other hand, has a much higher coefficient of consolidation, and does not conform to expected behavior; there is no linear section of the graph. Despite this, an arbitrary straight line has been drawn to “best fit” applied to determine √ some of the points, and the standard construction t90 . The value so obtained√is about 1.5 min0.5 , giving a value of t90 of 2.25 min. The true value of t90 is probably no greater than 0.25 min0.25 , corresponding to a t90 of only 0.06 min. The coefficient of consolidation calculated on this basis would be in error by a factor of nearly 40. This practice occurs because of an expectation that all soils should conform to “normal” behavior. Geotechnical engineers expect a straight line and laboratory technicians will obligingly provide one whether it exists or not! The proper interpretation in such cases should simply be that no straight line exists, and the test indicates that the value of cv is greater than the limiting value discussed above (0.1 m2 /day).
RATE OF CONSOLIDATION FOR SURFACE FOUNDATIONS ON DEEP SOIL LAYERS
73
4.7 RATE OF CONSOLIDATION FOR SURFACE FOUNDATIONS ON DEEP SOIL LAYERS
The well-known Terzaghi consolidation theory is not applicable to foundations of finite size placed on the surface of deep soil layers, which is often the case with residual soils. This is because neither settlement nor drainage of pore water is restricted to the vertical direction. The Terzaghi case and the actual drainage pattern beneath a surface foundation are illustrated in Figure 4.17. For circular or rectangular foundations, drainage will be three dimensional, while for a strip foundation (very long in comparison to width), it will be two dimensional. The analysis of these foundation situations is much more complex than the one-dimensional case, and analytical solutions are not available. Davis and Poulos (1972) have obtained solutions using numerical methods and computer programming. Davis and Poulos presented their solutions using the same time factor as the Terzaghi solution, which involves the thickness of the soil layer. With deep layers of firm to stiff residual clay (on which building foundations are likely to be built), the thickness of the layer is often deep and poorly defined. For this reason the solutions are presented here in a slightly different form, defining
Load of limited width
s
Uniform load of infinite width
Known boundary 3D consolidation & drainage
se
ep
ag
e
pa
th
1D vertical consolidation & drainage only
One-dimensional conditions assumed by the Terzaghi consolidation theory
Remote, and possibly unknown boundary Conditions applicable to most foundations, especially on deep residual soils
Figure 4.17 One-dimensional Terzaghi consolidation, and the situation beneath a surface foundation.
74
CONSOLIDATION AND SETTLEMENT
the time factor T in terms of the foundation dimensions, as follows: cv t where b is half the width of the foundation. b2 cv t Circular foundation: T = 2 where a is the radius of the foundation. a
Strip foundation: T =
The solutions are presented in Figures 4.18 and 4.19. Figure 4.18 gives the solutions assuming the foundation constitutes an impermeable barrier Time factor Ts = cv t / b2
Degree of consolidation U
0.1 0
0.3 0.5 0.7 1
3
5
7
10
30
50 70 100
Strip footing b
0.2 Impermeable base impermeable layer
0.4
h
10 5
0.6 Values of 0.8
h b 0.5
1
2
1.0 (a) Strip footing Time factor Ts = cv t / a2 0
0.04
0.1
0.3
0.5 0.7 1
3
5
7 10
30 40
Degree of consolidation U
Circular footing a 0.2 Impermeable base impermeable layer
0.4
0.6
0.8 Values of
h a
h
20 (=50) 10 5 2 1 0.5
1.0 (b) Circular footing
Figure 4.18 Degree of consolidation versus time factor for impermeable foundations (adapted from Davis and Poulos 1972).
RATE OF CONSOLIDATION FOR SURFACE FOUNDATIONS ON DEEP SOIL LAYERS
75
at the ground surface. This is the situation with most building foundations. Normal practice is to place a sealing layer of site concrete on the soil surface as soon as the excavation is made to the design depth. This prevents moisture change in the soil and also serves as a working platform. The foundation is then constructed on this layer of site concrete. Figure 4.19 gives the solution for permeable foundations. While the foundation itself is unlikely to be permeable, there are situations where a layer of hard-fill (crushed gravel) or sand is placed on the surface prior to construction of the foundation. This may be done to provide a firm working
Time factor Ts = cv t / b2
Degree of consolidation U
0.01 0
3
5 7 0.1
3
5 7 1
3
5 7 10
3 5 7100 Strip footing b
0.2 Permeable base Impermeable layer
h
0.4
10
5
0.6
2 Values of
0.8
h b
0.5
1
1.0 (a) Strip footing Time factor Ts = cv t / a2
Degree of consolidation U
0
0.001
3
5 7 0.01
3
5 7 0.1
3
5 7 1
3
5 7 10
Circular footing a
0.2
Permeable base Impermeable layer
h
0.4
0.6
0.8 Values of
h a
0.5
1
2
5
20
1.0 (b) Circular footing
Figure 4.19 Degree of consolidation versus time factor for permeable foundations (adapted from Davis and Poulos 1972).
76
CONSOLIDATION AND SETTLEMENT
platform or to accelerate the rate of consolidation by providing an outlet for seepage at the surface. Large storage tanks are often constructed in this way. We can make a limited check on the validity of Figure 4.19 by considering the case of a permeable foundation on a thin soil layer. We will therefore adopt the thinnest layer to which the charts can be applied, that is h/b and h/a = 0.5. With these conditions, consolidation should approach the one-dimensional situation since the easiest escape route for water is vertically upward to the permeable boundary of the loaded area. From the graphs in Figure 4.19 we can read off the value of the time factor T for the degree of consolidation U = 90%; this gives the following values of T90 . Strip footing: 0.17 Circular footing: 0.19 From these values we can determine the values of t90 in terms of h and cv . Strip footing: t90 =
T90 b 2 T90 4h 2 0.68 h 2 = = cv cv cv
Circular footing: t90 =
T90 a 2 T90 4h 2 0.76 h 2 = = cv cv cv
For a normal one-dimensional situation: t90 =
0.848 h 2 cv
These values agree well with what we would expect in this situation. The rate of consolidation is a little faster with these footings of limited size as some water can escape in a nonvertical direction at the edges of the foundation.
4.8 4.8.1
EXAMPLES OF SETTLEMENT ESTIMATES Foundations for a Multistory Building on Red Clay
A new 4-story building is planned for a site consisting of a deep layer of firm to stiff red clay. The unit weight of the clay is 16.9 kN/m3 . The building, a reinforced concrete framed structure is to be founded (hopefully) on surface footings. A suitable depth for the footings, to avoid surface shrinkage, is 1 m. The layout of the building is shown in Figure 4.20, and the loads carried by the columns are shown in Table 4.3.
3 @ 7.5 m = 22.5 m
EXAMPLES OF SETTLEMENT ESTIMATES
77
Internal column
Corner column
Intermediate column
3 @ 9.0 m = 27.0 m
Figure 4.20 Layout of building columns.
Determine the following: 1. Suitable dimensions for the foundations. Adopt a safety factor of 3 with respect to bearing capacity failure. 2. The expected settlement of the foundations. For settlement estimates use a load of G (dead load) + 0.4Q (live load). 3. Whether the differential settlement is within acceptable limits. Site investigation data are given in Figures 4.21 and laboratory test results from the red clay are given earlier in Figures 4.3 and 4.4. The coefficient of consolidation, determined from oedometer tests is about 0.15 m2 /day. Table 4.3 Column Loads Internal Column Intermediate Column Corner Column Dead load G (kN) Live load Q (kN)
1650 750
1200 600
900 450
78
CONSOLIDATION AND SETTLEMENT
Cone resistance (MPa) 0
5
10
15
20
Depth (m)
Stiff red clay Unit weight = 16.9 kN/m3
10 Water table
20
Figure 4.21 Typical CPT test from the site. Solution
Step 1 Calculation of allowable bearing pressure Estimation of undrained shear strength: Cone resistance qc = 1.8 MPa (average) Su = qu /16 = 113 kPa (using an Nk value of 16) Ultimate bearing capacity = Su Nc + γ D = 113 × 6.2 + 17 × 1 = 718 kPa Allowable bearing capacity = Su Nc /3 + 17 = 250 kPa Step 2 Calculation of foundation loads and foundation dimensions This is carried out in Table 4.4. Step 3 Estimation of compressibility parameter mv Examination of the oedometer test results in Figures 4.3 and 4.4 indicates that the compressibility is reasonably linear from zero pressure up to at about 300 or 400 kPa. We will therefore use the linear parameter mv . This is calculated from L = mv σ L
so that mv =
L Lσ
Calculating between zero and 350 kPa we have mv = 4.1/100 × 1/350 = 1.17 × 10−4 kPa−1 .
79
Dead Load G (kN)
1650 1200 900
Foundation
Internal Intermediate Corner
750 600 450
Live Load Q (kN)
2400 1800 1350
Bearing Capacity Load G + Q (kN)
9.6 7.2 5.4
Required Area (m2 )
3.1 2.7 2.4
Adopted Size of Square Foundation (m × m)
Table 4.4 Determination of Foundation Loads, Pressures, and Foundation Sizes
9.61 7.29 5.76
Actual Area (m2 )
1950 1440 1080
Settlement Estimate Load G + 0.4Q (kN)
203 198 188
Foundation Pressure for Settlement Estimate (kPa)
CONSOLIDATION AND SETTLEMENT
1
1m
2m
2
1m
3
2m
4
2m
1 2 3 5 7
5
4m
Depth in meters
80
11 6
4m
15 Start of hard layer
Figure 4.22 Sublayers used for calculation of settlement.
Step 4 Calculation of settlements To estimate the settlements we must divide the soil into a series of sublayers as shown in Figure 4.22. The foundation pressure calculated above should be reduced by the pressure from 1 m of excavation depth, that is, by 17 kPa, to give the values in the Table 4.5. Influence factors for a uniformly loaded rectangular foundation are obtained from elastic theory charts in the usual way. Table 4.6 shows the values and gives the complete calculation of the settlement. Step 5 Correct the estimated settlements so that they are valid for rigid footings The values calculated above are for the center of flexible foundations carrying uniform pressures. For a rigid foundation, the settlement (which is uniform) is about 75 percent of that for a flexible foundation. The above values must therefore be corrected Table 4.5 Details of Foundation Dimensions and Pressures
Interior column Intermediate column Corner column
B (m)
Foundation Pressure σ (kPa)
3.1/2 = 1.55 2.7/2 = 1.35 2.4/2 = 1.20
186 181 171
EXAMPLES OF SETTLEMENT ESTIMATES
81
Table 4.6 Calculation of Settlement (L = Lmv σ ) Layer
H (m)
Z (m)
Internal Column 1 1 0.5 2 1 1.5 3 2 3 4 2 5 5 4 8 6 4 12
B/Z = L/Z
Iσ
4Iσ
σ
H (mm)
3.1 1.03 0.52 0.31 0.20 0.13
0.245 0.18 0.090 0.040 0.018 0.008
0.98 0.72 0.36 0.16 0.072 0.032
182.3 133.9 67.0 29.7 13.4 6.0
21.3 15.7 15.7 6.9 6.3 2.8 = 69 mm
Intermediate Column 1 1 0.5 2 1 1.5 4 2 3 4 2 5 5 4 8 6 4 12
2.7 0.90 0.45 0.27 0.17 0.11
0.241 0.162 0.075 0.032 0.014 0.006
0.96 0.65 0.30 0.124 0.056 0.024
173.8 117.6 54.3 22.4 10.1 4.4
20.3 13.8 12.8 5.2 4.8 2.0 = 59 mm
Corner Column 1 1 2 1 3 2 4 2 5 4 6 4
0.5 1.5 3 5 8 12
2.4 0.8 0.4 0.24 0.15 0.10
0.239 0.146 0.060 0.026 0.011 0.004
0.96 0.58 0.24 0.10 0.044 0.016
164.0 99.2 41.0 17.1 7.5 2.7
19.2 11.6 9.6 4.0 3.5 1.3 = 49 mm
using a factor of 0.75. This gives the following estimates for the settlement: Center foundation: 52 mm Intermediate foundation: 44 mm Corner foundation: 37 mm These are fairly large settlements, but may still be acceptable. We need to check whether they are acceptable structurally by determining the angular distortion that will occur between foundations. Criteria commonly adopted usually limit this distortion to 1/500 or possibly 1/300. Between the center foundation and an intermediate foundation, the distortion is given by (52 − 44)/7500 = 1/938, which is acceptable. Between the center and the corner the distortion is (52 − 37)/11,700 = 1/780, which is also acceptable. Between
82
CONSOLIDATION AND SETTLEMENT
the intermediate and the corner the distortion is (44 − 37)/7500 = 1/1070, which is acceptable. We should also examine the likely rate of consolidation to check what proportion of these settlements can be expected to occur during construction. The maximum size of foundation is 3.1 × 3.1 m. We do not have a solution for the time rate with square foundations so we must adopt an equivalent circle. The radius of such a circle is 1.75 m. The value of h/a is about 15/1.75 = 8.6. Using this value in the chart in Figure 4.18b gives us a T90 value of about 6. The time for 90 percent consolidation is given by T90 = cv t90 /a2 , where a is the radius of the foundation. Therefore, t90 = T90 a2 /cv = 6 × (1.75)2 /0.15 days = 123 days = 4 months. It is highly improbable that a four-story building would be completed in this time. This means that the bulk of the expected settlement will occur as the structure is built. This makes the expected settlements even more tolerable. We have made one important assumption in carrying out the above calculation, namely that the pore pressure state in the ground above the water table can be ignored. This assumption is almost inevitable in practice as information is not available on the pore pressure situation. However, as we shall see in the next section, this is not necessarily a safe assumption as it may not be valid and can influence the results of the calculation. 4.8.2 Settlement Estimate Involving Nonlinear Compressibility and Pore Pressure Influence
In this example we will carry out a foundation settlement estimate involving the following: 1. Nonlinear compression curves from both linear and log plots 2. Account is taken of pore pressures above as well as below the water table A foundation is to be built on a layer of firm to stiff residual clay having a depth of 16 m, below which sandstone is found. The water table is measured at a depth of 9 m. The clay, which can be assumed to be fully saturated, has a unit weight of 17.5 kN/m3 and a void ratio of 1.150. The foundation is to be rectangular, with dimensions 3 × 6 m, and applied pressure of 150 kPa. Table 4.7 gives the results of a consolidation test on the soil. The following steps are involved: 1. Calculate the results of the consolidation test and plot log and linear graphs. Calculate the values of the compressibility coefficient, mv and plot these on a graph versus pressure.
EXAMPLES OF SETTLEMENT ESTIMATES
83
Table 4.7 Results of a Consolidation Test Stress (kPa) Sample Thickness (mm)
0 12.5 25 50 100 200 400 800 20.000 19.932 19.878 19.791 19.655 19.439 19.124 18.680
Stress (kPa) Sample Thickness (mm)
1600 800 400 200 100 50 25 12.5 18.027 18.214 18.224 18.331 18.456 18.568 18.680 18.828
2. Calculate the initial total and effective vertical stresses in the ground. This will involve making an assumption regarding the pore pressure state above the water table. Various assumptions are possible. 3. Divide the soil into suitable sublayers and calculate the stress increases that will occur in each sublayer as a result of construction of the foundation. 4. Calculate the final total and effective vertical stresses in the ground. 5. Apply the appropriate formula to estimate the compression of each sublayer. Solution
Step 1 Calculation of results of the oedometer test To do this we can use the following relationships: L e L = so that e = (1 + e0 ) 1+e L L L 1 L1 − L2 1 = mv = L σ L1 σ for each load increment We will also calculate the parameter C defined as follows: C =
e e1 − e2 for each load increment σ1 = log σ0 log10 σσ21
We will not define Cc or Cs at this stage. The reason for this will become apparent when we examine the shape of the compression curves. The calculations are set out in Table 4.8. Plotting the values in Table 4.8 gives the graphs in Figure 4.23.
84
CONSOLIDATION AND SETTLEMENT
Table 4.8 Calculation of Results of Consolidation Test Stress (%)
Sample Thickness L (mm)
0 12.5 25 50 100 200 400 800 1600 800 400 200 100 50 25 12.5
20.00 19.932 19.878 19.791 19.655 19.439 19.124 18.680 18.027 18.124 18.224 18.331 18.445 18.560 18.688 18.828
L (mm) Strain Change in Void mv 10−4 C (%) Void Ratio Ratio e kPa−1 e 0 0.068 0.122 0.209 0.345 0.561 0.876 1.320 1.973 1.876 1.776 1.669 1.555 1.440 1.312 1.172
0 0.34 0.61 1.05 1.73 2.81 4.38 6.60 9.87 9.38 8.88 8.35 7.78 7.20 6.56 5.86
0 0.0073 0.0131 0.0225 0.0371 0.0603 0.0942 0.1419 0.2121 0.2017 0.1909 0.1794 0.1672 0.1548 0.1410 0.1260
1.150 1.143 1.137 1.128 1.113 1.090 1.056 1.008 0.938 0.948 0.959 0.971 0.983 0.995 1.009 1.024
— 2.72 2.17 1.74 1.37 1.10 0.81 0.58 0.44 0.07 0.14 0.29 0.62 1.25 2.76 5.99
— 0.023 0.020 0.030 0.050 0.076 0.113 0.160 0.233 — — — — — — —
It is clear that in neither the linear nor the log plot is there a straight-line section of these graphs, at least not in the load application range. Parts of the unloading curves are reasonably straight on both plots. The values of the compressibility parameters are plotted versus stress level in Figure 4.24. The values of C are plotted at the average stress for each load increment in the oedometer test. Figures 4.23 and 4.24 illustrate several important points about the compressibility coefficient mv in comparison to the log coefficients Cc and Cs . There is no doubt from Figure 4.23a and the graph of mv in Figure 4.24 that the soil is becoming stiffer as the pressure increases. However, both Figure 4.23b and the graph of C in Figure 4.24 create the impression that the soil may be softening (yielding?) with increasing pressure, as the graph in Figure 4.23 steepens, and the value of the parameter C, which is a measure of its steepness, increases, as confirmed in Figure 4.24. We should recognize, of course, that the parameter C was never intended to be used in the way that it has been calculated and illustrated here. The concept on which the use of the parameters Cc and Cs arose was that soil compressibility could be represented by two straight lines on a log plot. Experience with undisturbed soils shows that
EXAMPLES OF SETTLEMENT ESTIMATES
85
Pressure (kPa) 0
200
400
600
800
1000
1200
1400
1600
400
800
1600
Compression (%)
2
4
6
8
10 (a) Linear scale (log) Pressure (kPa) 12.5
25
50
100
200
1.15
Void ratio
1.10
1.05
1.00
0.95 (b) Log scale
Figure 4.23 Results of the oedometer test using both linear and log pressure scales.
this concept is grossly simplistic, and why it still holds such a dominant place in soil mechanics is something of a mystery. Presumably, it results from the “omnipotence of tradition” mentioned earlier. If the log parameter is to be used in practical situations, then a range of C values such as shown in Figure 4.24 is likely to be necessary to adequately represent the behavior of the soil. The only
86
CONSOLIDATION AND SETTLEMENT
Pressure (kPa) 2
3
5 7 100
2
3
5 7 1000 2
0.28
Linear parameter mv Cc 2.0
Unloading
0.2
Loading
1.0
Log parameter C
0.1
Compression parameter C
Coefficient of compressibility mv (10−4 kPa−1)
2.8
5 7 10
Cs (?) 0
0
Figure 4.24 Graphs of linear and log compression coefficients versus pressure.
exception to this is the case of a soft, normally consolidated sedimentary clay. Rather than use a range of C values, it is generally more appropriate and convenient to use a range of mv values covering the stress range of interest. Step 2 Calculation of the stress state in the ground In the previous example the pore pressure state in the ground was ignored; in effect the assumption was made that it did not change during the construction of the foundation, and that its value was of no relevance to the settlement estimate. This is not necessarily a safe assumption, as mentioned earlier. Normally, the only information we have on the pore pressure state in the ground is the depth of the water table, and this does not necessarily fix the pore pressures above or even below it, as we saw in Chapter 3. Figure 4.25 shows three possible pore pressure states in relation to the water table, namely the equilibrium state and the summer (or dry season) and winter (or wet season) states. The winter state is a reasonable assumption, but the summer state is simply a guess. We cannot know which of these states applies at the time the foundation is built. Once the foundation is in place and the ground surface is “sealed”
87
EXAMPLES OF SETTLEMENT ESTIMATES
Foundation (1 m deep)
Pore pressure (kPa) −200
−150
−100
−215
−50
3m
0
50
100 0
Wet weather 1m −145
Dry weather
1
2m
2
2m
3
3m
4
3m
5
4m
6
2 4
−80
6 −25
Water table Equilibrium (hydrostatic) state
8 10
Depth (m)
−250
12 14 16
Figure 4.25 Possible states of the pore pressure state in the ground, and sublayers used in calculations.
from weather effects, the pore pressure can be expected to move toward, and ultimately assume, the equilibrium state, so we do at least know the final state. By investigating the implications of these possible initial states we can at least get a feel for the uncertainties involved in our estimate. It is convenient to calculate the stress state in the ground in the same table as the complete settlement estimate. This will be done in Tables 4.10 onward, but to complete these tables we need to know the stress increase in each sublayer resulting from construction of the foundation, so we will proceed to this step immediately. Step 3 Estimation of stress increase in each sublayer caused by the foundation load This is estimated in Table 4.9 in the normal way using the elastic solution for stress increases in the ground from a uniformly loaded flexible foundation. Note that the net foundation pressure q = 150 − 17.5 kPa = 132.5 kPa, B = 1.5 m, and L = 3.0 m. Steps 4 and 5 Calculation of final stresses and the compression of each sublayer Case A: We will start with a simplifying assumption that is often made, namely that the pore pressure above the water table is permanently zero. The initial and final stresses are then as shown in Table 4.10. We still have to decide how we will make use of the compressibility data presented in Figures 4.23 and 4.24, and in Table 4.8. To
88
CONSOLIDATION AND SETTLEMENT
Table 4.9 Calculation of the Stress Increase in Each Sublayer Resulting from the Foundation Load Layer 1 2 3 4 5 6
H (m)
z (m)
m = B/z
n = L/z
Iσ
4Iσ
σ = 4Iσ σ
1 2 2 3 3 4
0.5 2 4 6.5 9.5 13
3 0.75 0.38 0.23 0.16 0.12
6 1.5 0.75 0.46 0.32 0.23
0.248 0.165 0.086 0.043 0.028 0.015
0.992 0.660 0.344 0.172 0.112 0.060
131.4 87.5 45.6 22.8 14.8 8.0
obtain the most precise estimate we should use the average mv value for the stress range that applies to each sublayer. A reasonable but slightly less accurate estimate is obtained by simply using the average mv over the stress range applicable to all sublayers. We will start by using the value of mv applicable to each layer. The values are read off the graph in Figure 4.24. Case B: We will now repeat the calculation making another simplifying assumption, namely that the pore pressure state is hydrostatic above and below the water table. This assumption would be reasonably valid for soils of very low permeability of the type represented by Figure 3.3 of Chapter 3. This assumption does not alter the increase in effective stress in each layer but increases the actual values, (i.e., the stress level) because the negative pore pressure increases the effective stresses above the water table. This in turn reduces the average mv value (since mv decreases with increasing stress level) in each sublayer above the water table. The calculation is given in Table 4.11. We see that this gives us a lower estimated settlement. This is to be expected because the mv values steadily decline with increasing stress level. We can note in passing that we could carry out the above calculation using an average value of mv throughout the whole layer. The stress range is generally between about 50 and 200 kPa so we could use the average mv over this range, calculated as follows: L 1 L1 − L2 1 = L σ L1 σ 0.561 − 0.209 1 = 1.19 × 10−4 kPa−1 = 19.79 150
mv =
89
1 2 3 4 5 6
Layer
26.3 52.5 87.5 131.3 183.8 245.0
σo 0 0 0 0 14.7 49.0
uo o
26.3 52.5 87.5 131.3 169.1 196.0
σ
Initial Stresses (kPa)
131.4 87.5 45.6 22.8 14.8 8.0
σ (kPa)
157.7 140.0 133.1 154.1 198.6 253.0
σ1 0 0 0 0 14.7 49.0
u1 1
157.7 140.0 133.1 154.1 183.9 204.0
σ
Final Stresses (kPa)
131.4 87.5 45.6 22.8 14.8 8.0
σ (kPa)
1.29 1.25 1.20 1.08 1.00 0.95
17.0 21.9 10.6 7.4 4.4 3.0
L (mm)
Total: 65 mm
mv 10−4 kPa−1
Table 4.10 Case A: Settlement Estimate Assuming Zero Pore Pressure above the Water Table, and using the mv Value Applicable to Each Layer
90
1 2 3 4 5 6
Layer
26.3 52.5 87.5 131.3 183.8 245.0
σo 99.8 111.3 126.7 146.0 169.1 196.0
uo −73.5 −58.8 −39.2 −14.7 +14.7 +49.0
o
σ
Initial Stresses (kPa) 131.4 87.5 45.6 22.8 14.8 8.0
σ 157.7 140.0 133.1 154.1 198.6 253.0
σ1 −73.5 −58.8 −39.2 −14.7 14.7 49.0
u1 1
231.2 198.8 172.3 168.8 183.9 204
σ
Final Stresses (kPa) 131.4 87.5 45.6 22.8 14.8 8.0
σ
1.03 1.05 1.07 1.04 1.00 0.95
mv 10−4 kPa−1
Total: 56 mm
13.5 18.4 9.8 7.1 4.4 3.0
L (mm)
Table 4.11 Case B: Settlement Estimate Assuming Hydrostatic Pore Pressure above the Water Table before and after Construction of the Foundation, and using the mv Value Applicable to Each Layer
EXAMPLES OF SETTLEMENT ESTIMATES
91
If we use this value of mv and repeat the calculation in Table 4.10, we obtain a settlement of 64 mm, so the error involved in using an average value of mv is very small. We will now attempt to evaluate the possible influence of seasonal effects on our settlement calculation based on the following assumptions: Case C: The foundation is built during summer (dry conditions). Case D: The foundation is built during winter (wet conditions). The assumed pore pressure distributions are those shown in Figure 4.25. In making these assumptions we are investigating soils of the type represented by Figures 3.3 and 3.4 of Chapter 3. The water table position in Figure 4.25 is the only information we have on the pore pressure state in the ground, so the assumed profiles for winter and summer are best guesses only; the actual pore pressures could be different. The calculations are given in Tables 4.12 and 4.13 for the summer and winter conditions, respectively. The initial negative pore pressure values for the summer situation are read from the graph in Figure 4.24, and the winter values are taken as zero. The winter values imply a continuous supply of water at the surface and vertical seepage to the water table. In carrying out these estimates we must first give consideration to the appropriate value of mv for use in our calculations. If the soil above the water table experiences regular seasonal drying and wetting, then it has been subject to an almost unlimited number of cyclic changes in effective stress. In this case the appropriate mv may be that from the unloading curve in the oedometer test, depending on whether application of the foundation load will subject the soil to higher stresses than those it has experienced from weather effects. To illustrate this issue, profiles of the effective stress changes involved in our estimates are shown in Figure 4.26. The top two graphs are for cases A and B (Tables 4.10 and 4.11), when it was assumed the pore pressure above the water table was zero or constant. In those cases it was appropriate to use the mv from the loading curve as construction of the foundation applies a new stress level to the soil. The lower graphs show the pore pressure changes for cases C and D. The lower left graph for summer conditions
92
1 2 3 4 5 6
Layer
26.3 52.5 87.5 131.3 183.8 245.0
σo 241.3 197.5 167.5 156.3 169.1 196.0
uo −215 −145 −80 −25 14.7 49.0
o
σ
Initial Stresses (kPa) 131.4 87.5 45.6 22.8 14.8 8.0
σ 157.7 140.0 133.1 154.1 198.6 253.0
σ1 −73.5 −58.8 −39.2 −14.7 14.7 49.0
u1 1
231.2 198.8 172.3 168.8 183.9 208.2
σ
Final Stresses (kPa) −10.1 1.3 4.8 12.5 14.8 8.0
σ
0.83 0.83 0.83 0.83 1.00 0.95
mv 10−4 kPa−1
Total: 11 mm
−0.8 0.2 0.8 3.1 4.4 3.0
L (mm)
Table 4.12 Case C: Settlement Estimate Assuming Dry Weather Condition Initially and Hydrostatic Finally, and using an Average mv Value for all Layers
93
1 2 3 4 5 6
Layer
26.3 52.5 87.5 131.3 183.8 245.0
σo 0 0 0 0 14.7 49.0
uo o
26.3 52.5 87.5 131.3 169.1 196.0
σ
Initial Stresses (kPa) 131.4 87.5 45.6 22.8 14.8 8.0
σ 157.7 140.0 133.1 154.1 198.6 253.0
σ1 231.2 198.8 172.3 168.8 183.9 204.0
u1 −73.5 −58.8 −39.2 −14.7 14.7 49.0
1
σ
Final Stresses (kPa) 204.9 146.3 84.8 37.5 14.8 8.0
σ
0.83 0.83 0.83 0.83 1.00 0.95
mv 10−4 kPa−1
Total: 72 mm
17.0 24.3 14.1 9.3 4.4 3.0
L (mm)
Table 4.13 Case D: Settlement Estimate Assuming Wet Weather Condition Initially, and Hydrostatic Finally, and using an Average mv Value for all Layers
94
CONSOLIDATION AND SETTLEMENT
Vertical effective stress (kPa)
Vertical effective stress (kPa) 100
200
Depth (m)
5
Final effective stress assuming zero pore pressure above the water table at all times Water table
10
Initial effective stress assuming hydrostatic pore pressure above and below the water table
15
200
0
100
200
300
5 Final effective stress assuming pore pressure is hydrostatic above and below the water table
Water table
10
Depth (m)
Depth (m)
Case B
Vertical effective stress (kPa) 300
5
Water table
Case C
Final effective stress assuming pore pressure is hydrostatic above and below the water table
10
Initial effective stress assuming dry weather condition— high negative pore pressure above the water table, hydrostatic below
15
300
Final effective stress assuming hydrostatic pore pressure above and below the water table at all times
Water table
Vertical effective stress (kPa) 100
200
10
Case A
0
100
5
Shaded area indicates change in effective stress
Initial effective stress assuming zero pore pressure above the water table at all times, hydrostatic below the water table
15
0
300
Depth (m)
0
Initial effective stress assuming wet weather condition—zero pore pressure above the water table, hydrostatic below
15
Case D
Figure 4.26 Effective stress states in the settlement calculations, cases A to D.
(at the time the foundation is built), shows that construction of the foundation scarcely raises the stress level above what it was before the foundation was built. The increase in effective stress from the foundation load is essentially counterbalanced by the decrease in effective stress resulting from the change in pore pressure. This means that for both the summer (case C) and winter (case D) conditions, nearly all of the increase in stress will be within the range to which the soil has already been regularly subjected during summer seasons. We should therefore use the mv from the unloading part of the oedometer test. Calculating this for the same stress
EXAMPLES OF SETTLEMENT ESTIMATES
95
range (50–200 kPa) we obtain L 1 L1 − L2 1 = L σ L1 σ 1.669 − 1.440 1 = = 0.83 × 10−4 kPa−1 18.33 150
mv =
In this case there is almost no settlement. The situation we have just analyzed is similar to that of expansive clays, but on a very mild scale. There is slight swelling only in the top sublayer. The settlement for the situation when the foundation is built during winter conditions is 69 mm, so is only marginally greater than the value of 66 mm obtained from our first estimate when the pore pressure was assumed to be zero. The settlement estimate for the winter condition is the greatest, as expected, since it involves an increase in effective stress from two causes, namely the application of the foundation load and the change in pore pressure above the water table from zero to a negative (equilibrium) value. These calculations illustrate the possible uncertainties of settlement estimates when the water table is relatively deep, and the pore pressure state above the water table is uncertain. In the absence of any other information, the best we can do is to assume the pore pressures are in hydrostatic equilibrium above and below the water table. We saw above that there is a small difference between assuming no pore pressure and assuming the hydrostatic pore pressure state. 4.8.3
Significance of Time Rate Assumption in the Previous Example
In the example in Section 4.8.2 above, no attention was paid to the rate at which pore pressures would change as a result of building the foundation. As illustrated by the example, pore pressures may change as a result of either or both of the following two factors: 1. Application of the foundation load 2. Changes in the negative pore pressure state above the water table, due to covering the ground surface The calculations in Section 4.8.2 are correct, provided the changes due to these two effects occur at the same rate. In general, this is not likely to be the case. The rate of consolidation from the foundation load can be expected to be more rapid than the volume changes resulting from changes
96
CONSOLIDATION AND SETTLEMENT
in the negative pore pressure state. This is because the drainage path length will normally be smaller in the former. However, with many residual soils, because of their relatively high permeability, settlement due to application of the foundation load and volume changes resulting from changes in the negative pore pressure state are likely to occur quite rapidly and to be virtually complete during the construction period. In this case, it is probably sufficiently accurate to assume they occur at the same time. With low-permeability clays, there is likely to be a time lag between the two effects, with the influence of the foundation load being felt ahead of the influence of the changes in negative pore pressure. If the magnitude of settlement is a critical issue, this time effect needs to be considered in making the settlement estimates.
4.9 ACCURACY OF SETTLEMENT ESTIMATES BASED ON OEDOMETER TESTS
It will be readily apparent from Section 4.8 that there is always likely to be uncertainty in estimates of settlement of surface foundations on residual soils because of the unknown pore pressure state above the water table. This error is accentuated when the log parameter Cs is used for making the estimate, as will be demonstrated in the following section. However, despite these uncertainties, it is generally the case that settlement estimates based on oedometer tests commonly overestimate the actual settlement. 4.9.1 A Common Source of Error Arising from the Use of the Log Parameter (Cs )
Consider the situation illustrated in Figure 4.27. This shows a layer of clay 10 m deep with a water table at 8 m. The result of a consolidation test on a sample of the clay taken near the surface is also shown. This indicates a preconsolidation pressure of about 300 kPa, and assuming the foundation stress will be less than this, we will carry out our settlement estimate using the Cs parameter over this part of the curve. Reading off the void ratio between 10 and 100 kPa we can calculate the value of Cs as follows: Cs = e/ log(100/10) = e = 1.052 − 1.010 = 0.042 To carry out our settlement estimate we will divide the clay into a series of sublayers starting with a layer 1 m thick immediately beneath the foundation. The initial vertical effective stress at the center of this layer is then 1.5 × 16 = 24 kPa. We will assume the foundation pressure (at the base of the foundation) is a little greater than 100 kPa so that the stress increment in this
ACCURACY OF SETTLEMENT ESTIMATES BASED ON OEDOMETER TESTS
97
Pressure (kPa) 1.1
1
10
100
1000
10000
Δσ′ δ1
1.0
Void ratio
1.5 m
δ2
Foundation Sublayer
1.0 m 1.0 m
Δσ′
8m 0.9
10 m
0.8
Clay: Unit weight = 16.0 kN/m3
Water table
Rock
0.7
Figure 4.27 Settlement estimate using the log parameter Cs .
layer is 100 kPa. The stress increment is therefore from 24 to 124 kPa. The compression of this layer is then given by δ1 =
σ + σ 24 + 100 Cs 0.042 H log o log = 1+e σo 1 + 1.05 24
= 0.0146 m = 14.6 mm However, in making this estimate, we have assumed that there is no pore pressure in the soil above the water table. This is not likely to be the case. It is more likely to be negative, and the most reasonable assumption we can make is that it is in hydrostatic equilibrium with the water table at 8 m deep. In this case the pore pressure is given by u = −6.5 × 9.8 = −63.7 kPa. The initial effective stress is, therefore, σ o = σ − u = 24 − (−63.7) = 87.7 kPa. The stress increment is now from 87.7 to 187.7 kPa. The compression of the layer is now given by δ2 =
87.7 + 100 0.042 log 0.0068 m = 6.8 mm 1 + 1.052 87.7
There is thus a substantial difference, the new value being less than half the previous value. We could carry on through further sublayers, and would find differences that would become increasingly smaller with depth. The stresses and the corresponding void ratio change for the two assumptions above regarding the pore pressure state are shown in Figure 4.27. If the model of behavior is a log one with respect to pressure, then the compression for a given stress increment is very dependent on the initial stress, as the diagram and the calculations illustrate.
98
4.9.2
CONSOLIDATION AND SETTLEMENT
Actual Settlement versus Predictions
Experience suggests that settlement predictions based on oedometer tests are usually somewhat conservative; that is, the actual settlement is normally less than the predicted value. The author’s experience with the residual (Waitemata) clays derived from sandstone in Auckland, New Zealand has been that settlements appear to be less than predicted, although reliable records to confirm this are very few. The same is true of foundations on the residual clays of volcanic origin in Java, Indonesia. The raft foundation of a geothermal power station (at Kamojang in West Java) was expected to show some minor differential settlement when a second stage was built with an integral connection to the first stage. No field evidence was found to indicate that any significant differential movement actually occurred. Useful information on predicted versus actual settlement is given by Willmer et al. (1982), and Barksdale et al. (1982) for foundations in the Piedmont residual soil of the southeastern United States. Willmer et al. (1982) describe settlement estimates and actual measured settlements for five structures. The measured settlements were generally significantly below the predicted values, the ratio ranging from 30 to 115 percent, with an average of 73 percent. Willmer et al. draw the following conclusions from their study: “Computed settlements for these soils, using conventional one-dimensional consolidation test data, are typically higher than actual settlements measured in the field. In addition, a high percentage of the settlement usually occurs rapidly upon application of the load.” Barksdale et al. (1982) made settlement estimates for a tower building using several methods, described as the one-dimensional method (the normal oedometer method), the layer-strain method, and a proposed stress path approach. They state that while these methods all give “good estimates of the observed settlements,” in the residual silty sands at the site, “in general, however, settlements are overestimated on residual soils.” The oedometer tests presented in their paper are all plotted using the conventional log plot, and show curves similar to those shown earlier in Figure 4.5. Their settlement estimate was carried out using an overconsolidation ratio of 4.5, and the Schmertman construction to determine the in situ compression curves. The use of this construction for residual soils appears quite inappropriate as it is based entirely on the idealized compression behavior of sedimentary soils.
4.10 ALLOWABLE DIFFERENTIAL SETTLEMENT FOR SURFACE FOUNDATIONS ON RESIDUAL SOIL
Various criteria have been established to ensure that buildings do not suffer unacceptable damage as a result of distortion caused by differential settlement. Damage is generally divided into two categories, namely aesthetic or
REFERENCES
99
nuisance damage, and structural damage. Aesthetic damage involves cracking of finishing elements and decorative features, and possibly jamming of windows or doors. Structural damage means threatening or weakening the essential strength of the building. The criteria normally used relate to the angular distortion of the building frame; the values commonly used are as follows: Aesthetic damage— angular distortion should be less than 1/500 Structural damage— angular distortion should be less than 1/250 (or 1/150 according to some authorities) Structural damage is caused by the total settlement, whereas aesthetic damage is caused only by postconstruction settlement, since the components that suffer aesthetic damage are normally installed only after the building frame is complete. The limit to be used in practice will depend on the nature of the building, since some buildings are more susceptible to aesthetic damage than others, and on the rate of settlement. If the structure is built on a soil that consolidates rapidly, most settlement will occur during construction (this is when the major loads are applied), and postconstruction settlement causing aesthetic damage will be small. This generally means that tolerances for buildings on residual soils can be less stringent than those for sedimentary soils, since the rate of consolidation of the former is normally more rapid.
REFERENCES Azevedo, F. F. S. S. 1972. Estudo de Compresssibilidade de Solos residuais de Gnaisse. Thesis, PUC-Pontificia Universidade Catolica do Rio de Janeiro. Barksdale, R. D., R. C. Bachus, and M. B. Calnan. 1982. Settlement of a tower on residual soil. Proceedings, ASCE Specialty Conference on Engineering and Construction in Residual Soils, Hawaii, 647–664. Budhu, M. 2000. Soil Mechanics and Foundations. New York: Wiley, Figure 4.16, 177. Campos, M. T. P. de. 1980. Ensaios de laboratorio e Provas de Carga Superficiais Instrumentadas no Solo Residual Gnaissico Jovem do Campo Experimental da PUC-RJ, Gavea, RJ. Thesis, Pontificia Universidade Catolica de Rio de Janeiro. Craig, R. F. 1992. Soil Mechanics, 5th ed. London: Chapman & Hall, Figure 7.4, 249. Das, B. M. 1998. Principles of Geotechnical Engineering, 4th ed. Boston: PWS, Figure 8.8, 314. Davis E. H., and H. G. Poulus. 1972. Rate of settlement under two and three-dimensional conditions. Geotechnique 22, No 1, 95–114. De Mello, V. F. B. 1972. Thoughts on soil engineering applicable to residual soils. Proc. 3rd Southeast Asian Conf. on Soil Engineering, Hong Kong, 5–34.
100
CONSOLIDATION AND SETTLEMENT
Garga, V. K., and C. A. Costa. 1977. Stress–compressibility characteristics of a residual soil from gneiss. Proc. 9th ICSMFE, Tokyo, Vol.1, 105–108. Hoyes, L., and E. J. Macari. 1999. Influence of in situ factors on dynamic response of Piedmont residual soils. ASCE Journal of Geotechnical and Environmental Engineering 125(4): 271–279. Hui, T. W. 1972. Some properties of a Malaysian residual granite soil. Proc. 3rd Southeast Asian Conf. on Soil Engineering, Hong Kong, 67–71. Janbu, N. 1998. Sediment deformation. Bulletin 35, Norwegian University of Science and Technology, Trondheim, Norway. Lancellotta, R. 1995. Geotechnical Engineering. Rotterdam: Balkema, Figure 5.18, 108. Lumb, P. 1962. The properties of decomposed granite. Geotechnique 12(3): 226–243. Mayne, P. W., and D. A. Brown. 2003. Site characterization of Piedmont residuum of North America. Proc. International Workshop on Characterization and Engineering Properties of Natural Soils, Vol. 2 Singapore, 2002. Leiden, The Netherlands: A. A. Balkema, 1323–1339. Napoles Neto, A. D. F. 1954. Estudo Dos Recalques de um Grande Castelo d’agua Fundado sobre Solo Residual Mole. 1st Brazilian Congress on Soil Mechanics. Vol ll, 175–188. Pender, M. J., L. D. Wesley, G. Twose, G. C. Duske, and Satywan Pranjoto. 2000. Compressibility of Auckland residual soil. Proceedings GeoEng2000 Conference, Melbourne. (CD Rom) Sandroni, S. S. 1981. Solos Residuais Pesquisas Realizadas na PUC-RJ. Proc. Simposio Brasileiro Sobre Solos Tropicais em Engenharis, Vol. 2, COPPE-UFRJ, 30–65. Vargas, M. 1979. Settlements of footings, mat or raft foundations. Proc. 6th PACSMFE, Lima, Vol. 2, 333–340. Werneck, M. L. G., S. F. D. Jardim, and M. de Souza S. de Almeida 1979. Deformation modulus of a gneisic residual soil determined from plate load tests. Solos e Rochas 2(2): 3–18. Wesley, L. D. 2000. Discussion on paper: Influence of in situ factors on dynamic response of Piedmont residual soils. ASCE Journal of Geotechnical and Geoenvironmental Engineering 126(4): 384–385. Wesley, L. D. 2010. Fundamentals of Soil Mechanics for Sedimentary and Residual Soils. New York: Wiley, Figure 8.29, 166. Willmer, J. L., G. E. Futrell, and J. Langfelder 1982. Settlement predictions in Piedmont residual soil. Proceedings, ASCE Specialty Conference on Engineering and Construction in Residual Soils, Hawaii, 629–646.
CHAPTER 5
SHEAR STRENGTH OF RESIDUAL SOILS
5.1
INTRODUCTION
In this chapter we will make some fairly broad observations about the shear strength properties of residual soils, and present data to support these observations. As mentioned in Chapter 3, the engineering properties of residual soils are generally good, especially those derived from the weathering of igneous and volcanic rocks. Both the undrained strength and the effective strength parameters of these residual soils are fairly high. This is self-evident from observation of hill slopes consisting of residual soils; these normally remain stable at substantially steeper slopes than is the case with most sedimentary soils. Further observations of this behavior are made in Chapter 8 on slope stability. Soils derived from the weathering of soft sedimentary rocks, especially shale, and the black cotton soils described earlier (Chapters 1 and 3), are a quite different group and tend to have poor engineering properties. They are frequently rich in the clay mineral montmorillinite, giving rise to low shear strength and failures in slopes of very gentle inclination, as well as foundation problems caused by their shrink–swell behavior. As mentioned in Chapter 2, it is questionable whether soils derived from the weathering of shale should be classed as residual soils, since their predominant properties tend to be derived more from their former lives as sedimentary soils than from the recent weathering processes that qualify them as residual soils. Comments in this chapter, therefore, are focused mainly on those soils that are derived from igneous and volcanic parent material. 101
102
5.2
SHEAR STRENGTH OF RESIDUAL SOILS
UNDRAINED SHEAR STRENGTH
As the reader is probably aware, for fine-grained sedimentary soils there are some empirical correlations relating undrained shear strength to other soil parameters. The best know is that of Skempton (1957), which relates the undrained shear strength of normally consolidated soils to the effective consolidation pressure and the Atterberg limits. His relationship is shown in Figure 5.1, and is summarized by the expression Su = 0.11 + 0.0037(PI) σ where
Su = the undrained shear strength σ = the effective overburden pressure PI = plasticity index
This relationship is of no relevance to residual soils in their undisturbed state. However, for fully remolded soils, it is also possible to relate the undrained shear strength to their water content and Atterberg limits, or, more specifically, to the liquidity index of the soil. The undrained shear strength is considered to be about 170 and 1.7 kPa at the plastic and liquid limits, respectively (Sharma and Bora, 2003). Slightly different values have been proposed by other authors. These values and the assumed logarithmic curve relating undrained shear strength with liquidity index are shown in Figure 5.2. This curve represents the lower limit of shear strength at which a soil can exist. Most soils in nature will exist with higher undrained shear strength than that given by this curve. Only undisturbed soils that show no
0.6 Su = undrained shear strength where effective vertical stress = σ′ 0.4 Su σ′ 0.2
0
Su = 0.11 + 0.0037 PI σ′
20
40
60
80
100
120
Plasticity Index (PI)
Figure 5.1 Undrained shear strength related to confining stress and plasticity index (after Skempton 1957).
EFFECTIVE STRENGTH PROPERTIES
103
250
Undrained shear strength (kPa)
Tropical red clays 200
Weathered sandstone (Auckland, NZ)
100
Volcanic ash clays
Fully remolded soils
0 −0.2
0
0.5
1.0
(PL)
Liquidity Index
(LL)
1.4
Figure 5.2 Undrained shear strength versus liquidity index for several residual soils and graph for fully remolded soils.
loss of strength on remolding (i.e., nonsensitive soils) will lie on this line, and such soils are not at all common. Also shown in Figure 5.2 are approximate limits for undrained shear strength and liquidity index for three residual soils of which the author has some experience. As expected, there is no relationship between these values and those for remolded soils. The undrained shear strength of undisturbed residual soils generally lies well above the strength of the fully remolded soil—a fact that arises because of the contribution to shear strength arising from their structure. It appears that the undrained shear strength of fine-grained residual soils is seldom less than about 75 kPa, with the possible exception of black cotton clays, and is commonly above or well above 100 kPa. Even black clays are not particularly soft, though they are probably softer than most other residual soils. 5.3
EFFECTIVE STRENGTH PROPERTIES
The high effective strength parameters found in most residual soils arise from several factors, including the following: 1. Residual soils contain clay minerals that tend to have good frictional properties. The exceptions are those derived from the weathering of soft sedimentary rocks, such as shale, and the black cotton soils described in Chapter 2, in which montmorillonite is predominant.
104
SHEAR STRENGTH OF RESIDUAL SOILS
2. Most residual soils have significant microstructural effects, which contribute positively to the shear strength of the material. 3. The microstructure also generally contributes a significant cohesive component to the shear strength of the material, that is, a significant c value. The angle of shearing resistance φ is generally in the range of 25–35◦ , although allophane and halloysite clays can have φ values as high as 40◦ .
5.3.1
Influence of Discontinuities
In contrast to these positive factors, there is the negative influence of possible discontinuities within the soil arising from its parent rock. These can constitute planes of weakness, and make the determination of representative values of c and φ very difficult. Figure 5.3 shows results of triaxial tests on a clay derived from weathered sandstone containing discontinuities. It is seen that there is a wide variation in strength, reflecting the influence of discontinuities in the soil. Some residual soils rarely contain discontinuities. Volcanic ash clays are an example; Figure 5.4 shows results of triaxial tests on a large number of volcanic ash samples. It is seen in this case that there is only a fairly narrow scatter in the results.
600
Shear strength (kPa)
MIDDLE CLAY from weathered sandstone °
400
Pa
c′
=
= φ′
34
k 54
′= aφ
25°
P
c′ =
5k
200
0
200
400 600 Normal stress (kPa)
800
1000
Figure 5.3 Triaxial tests on a residual clay containing discontinuities, derived from weathered sandstone.
EFFECTIVE STRENGTH PROPERTIES
300
5°
Shear strength (kPa)
VOLCANIC ASH CLAYS (from Indonesia & New Zealand)
Pa
c′
=
k 34
3 ′=
φ
′ aφ
105
4°
=3
P
4k
1 ′=
c
200
100
0
100
200
300
400
500
Normal stress (kPa)
Figure 5.4 Triaxial tests on volcanic ash clays (New Zealand and Indonesia).
5.3.2
Correlation between φ Value and the Atterberg Limits
For intact materials, not influenced by the presence of discontinuities, the position of the soil on the plasticity chart provides a good guide to their likely φ value, at least for predominantly fine-grained soils of moderate to high plasticity. As indicated earlier in Chapter 2, it is the position of the point in relation to the A-line that provides the best indication of likely engineering properties. The φ value tends to be low for soils lying above the A-line and high for those lying below the A-line. Some limited data for both residual and sedimentary soils from the author’s own files are presented graphically in Figure 5.5. This shows φ values plotted against distance of the soil above or below the A-line. It is clear that there is a reasonably well-defined relationship that shows a marked decline in the φ value as the soil approaches and rises above the A-line. The soils labeled sedimentary in this figure are a mixed group and the term sedimentary may be misleading. Those lying well above the A-line and having the lowest φ values are believed to be derived from the weathering of shale, and their properties are considered to originate from their parent shale, rather than any weathering processes they have undergone. However, the topography where they are found is not particularly well drained, and the weathering process may have been of the sort that produces minerals of the montmorillonite group. The actual type of soil does not appear to be very important; the trend shown in Figure 5.5 may apply to all soils, regardless of whether they are sedimentary or residual. However, many more data are needed to verify this.
106
SHEAR STRENGTH OF RESIDUAL SOILS
50
Friction angle φ′ (degrees)
40
30
20 Sedimentary soils Residual soils Volcanic soils
10 Below A-line
0 −30
−20
−10
Above A-line
0
20
10
Distance below or above the A-line = PI − 0.73(LL − 20)
Figure 5.5 The friction angle (φ ) related to position on the plasticity chart.
5.3.3 Effective Strength Parameters of a Residual Soil Derived from Shale
Figure 5.6 shows strength measurements on a clay derived from the weathering of shale, described by Brenner et al. (1997). Conventional shear box tests and ring shear tests were carried out. The shear box samples were
Shear stress (kPa)
300 Shear box—peak Shear box—residual Ring shear—residual
200
8 kPa
c′ = 2
100
2°
φ′ = 1
a φ′ = 6°
c′ = 12 kP
c′ = 6 kPa φ′ =
0
100
200 300 400 Normal stress (kPa)
5°
500
600
Figure 5.6 Effective strength parameters from shear box tests on a residual soil derived from the weathering of shale in KwaZulu, Natal, South Africa (after Brenner et al. 1997).
EFFECTIVE STRENGTH PROPERTIES
107
essentially undisturbed and enabled the peak strength to be established. By conducting a number of reversals of the direction of movement a measurement of the residual strength was also possible. The ring shear samples were partially disturbed so that only the residual strength was obtained from these tests. The friction angles (φ ), both peak and residual, are very low. There is a small anomaly in the residual values from the two types of test. The lowest value is given by the shear box, rather than the ring shear test, as is normally the case. The peak φ value is compatible with the lowest values given in Figure 5.5. These tests confirm the general expectation of low shear strength parameter in soils derived from shale. 5.3.4
Stress–Strain Behavior in Triaxial Tests
Figures 5.7–5.10 illustrate the behavior of two particular residual soils in triaxial tests. Figures 5.7 and 5.8 show results of tests on both undisturbed and remolded samples of red clay, and Figures 5.9 and 5.10 show similar results for a silt derived from the weathering of sandstone. These are the same soils for which oedometer tests were presented earlier (Figures 4.3, 4.4, and 4.6–4.8, respectively). These tests show the following characteristics: 1. The red clay behavior is similar in both its undisturbed and remolded state. When it approaches failure, the deviator stress continues to increase—only slightly, but steadily. The pore pressure curves show a steady decrease, indicating that the soil is behaving in a dilatant manner, i.e., it is tending to increase in volume. This is perhaps not surprising as it is a dense material in both its natural and remolded state. It is still a little surprising as it is a very fine-grained, moderately plastic clay—such clays do not normally show dilatant behavior. 2. The behavior of the silt appears similar at first sight, but it is significantly different. In its undisturbed state, its strength reaches a peak, and then declines at a slow but steady rate. The pore pressure shows a slight increase or remains steady. When remolded, the behavior is different—it now behaves in a typical silt manner, showing dilatant behavior at all stress levels. 3. The stress paths indicate that both clays behave rather like moderately overconsolidated soils. At low stress levels they behave as overconsolidated materials, but their behavior changes gradually as the stress level is raised and tends toward normally consolidated behavior. 5.3.5
The Cohesion Intercept c
In Chapter 9 (on slope stability) the comment is made that the value of c usually plays a significant role in maintaining the stability of slopes
108
SHEAR STRENGTH OF RESIDUAL SOILS
Tropical red clay Liquidity Index = −0.17 500 Sensitivity = 1
500 Effective consolidation pressure (kPa)
Deviator stress (kPa)
400 400 250
300
300
200
100 50
100 Undisturbed Compacted 0
2
4
6
8
10
14
500
300
Pore pressure (kPa)
12
400
200
300 250
100 100 50 0
2
4
6 8 Strain (%)
10
12
14
Figure 5.7 Consolidated undrained triaxial tests on undisturbed and remolded samples of tropical red clay.
in residual soils, and this point is emphasized here. The true source of the c value is uncertain; the most common explanation is that it is due to some form of weak bonds between particles, and in most soils this is probably true. An interesting example when this may not be true is that of terraced rice fields that are permanently irrigated from a canal supply at the top of the slope. An idealized cross section of such rice fields is shown in Figure 5.11. The flow net is for an infinite slope, and is a good approximation (at least
109
EFFECTIVE STRENGTH PROPERTIES
300
σ1′ − σ3′ (kPa) 2
Tropical red clay Undisturbed Compacted
a kP
4
200
ak
1 ′=
φ′
7°
=3
ual
sid
Re
c
Pe
100
0
100 200 300 σ1′ + σ3′ (kPa) In situ vertical 2 effective stress
400
500
600
Figure 5.8 Stress paths from the triaxial tests on tropical red clay.
Deviator stress (kPa)
300 400 200 200 100 50
0
2
4
6
10
12
14
Clayey silt: Liquidity Index = 0.5 Sensitivity = 10
Undisturbed Remolded
Pore pressure (kPa)
8
300
400 Effective consolidation pressure (kPa)
200
200 100 50 0
2
4
6
8
10
12
14
Strain (%)
Figure 5.9 Consolidated undrained triaxial tests on undisturbed and remolded samples of sensitive residual silt.
SHEAR STRENGTH OF RESIDUAL SOILS
200
Clayey silt Undisturbed Remolded
′= aφ
°
33
P
′= kc
5k
a
Pe 2
σ′1 − σ′3
(kPa)
110
ual
sid
Re
100
0 100 In situ vertical effective stress
200 300 σ′1 + σ′3 (kPa) 2
400
Figure 5.10 Stress paths from the triaxial tests on sensitive residual silt.
Terraced slope for irrigated rice cultivation
B H D A
β
For H = 3 m β = 35° γ = 16 kN/m3 φ′ = 35° the required value of c′ = 8.4 kPa on plane A-B = 15.3 kPa on plane C-D C
Figure 5.11 Cohesion intercept in irrigated, terraced rice fields from backanalysis.
EFFECTIVE STRENGTH PROPERTIES
111
theoretically) for the central terraces on such slopes. We can back-analyze this slope in two ways: we can analyze individual terraces, and we can analyze the slope as a whole. The potential failure planes on which such analysis has been carried out are indicated in Figure 5.11. To obtain the c value, it is necessary to make an assumption about the φ value. A value of 35◦ is adopted here, which is considered to be an average value for halloysite and allophane clays on which these rice fields are normally built, at least in Indonesia and other parts of Southeast Asia. The terraces are formed on slopes as steep as about 40◦ , and the maximum terrace height is about 3 m, although most terraces are in the 1- to 2-m range. This back-analysis gives the following values of c : Plane A-B 8.4 kPa Plane C-D 15.3 kPa Without c values in this range, the individual terraces, and the slope as a whole, would not be stable. It is clear also that because of the way the terraces are formed, at least half of each vertical face must consist of remolded soil. In fact, a larger proportion probably consists of remolded soil as the rice farmers clear the faces of weeds and other vegetation from time to time, and in doing so scrape a layer of soil from the faces and dump it on the surface of the terraces immediately below. In this way a slow remolding of the whole slope appears to occur. The explanation for these significant c values may be that there are attractive forces between the very fine particles contained in these clays. Figure 5.12 is a further illustration of the real c intercept in residual clays, this time coming from triaxial tests on the residual clays of Auckland, New Zealand, mentioned earlier in Section 4.3.3. These clays are formed from the weathering of sandstone and mudstone formations. A laboratory study of the soil involved drained triaxial compression tests and extension tests, and also a number of tension tests to investigate the behavior of the clay at the origin of the Mohr-Coulomb failure envelope. Tension tests are a special type of test that should not be confused with extension tests. The latter are simple tests in which the lateral stress exceeds the vertical stress, so that the soil fails in extension, but all stresses are compressive. Both vertical and horizontal stresses are compressive and no part of the soil experiences tensile stress. Tensile tests are designed specifically to create a vertical tensile stress so that the Mohr-Coulomb envelope straddles the origin and failure takes place on planes of zero effective stress. The tension tests were drained and the procedure developed by Bishop and Garga (1969) was followed. This involved the use of dumbbell-shaped samples as indicated in Figure 5.12. A full account of the study is to be found in Meyer (1997) and Meyer et al. (1999). The tests demonstrate clearly that the soil has significant tensile strength, the actual values ranging from 7.7 to 12.0 kPa. They show also that the soil
112
SHEAR STRENGTH OF RESIDUAL SOILS
l xia tria s m fro sse pe stre o l ve ng 7° en nfini 29. e r ′= co lu fai her Pa, φ n g k i o si at h 3.7 ten 1 Ex tests c′ =
Shear stress (kPa)
40
20
T
Shape of triaxial sample to test soil in tension
T −20
0
20
40
Effective normal stress (kPa)
Figure 5.12 Mohr’s circles at failure from the tension tests on Auckland clay (after Meyer et al. 1999).
still has considerable strength on failure planes on which the effective normal stress is zero or less than zero. The average c value from these tension tests is in good agreement with the intercept value from the Mohr-Coulomb envelope established by conventional methods. The Mohr-Coulomb parameters c and φ given in Figure 5.12 are in good agreement with those obtained by back-analysis of actual failures in intact areas of the Auckland clay. Exponents of the Cambridge (England) “critical state” school of soil mechanics have from time to time expressed the view that the cohesive intercept component of soil shear strength cannot be relied on and ought not to be included in design analysis. Geotechnical engineers working with real soils, especially residual soils, have long known from simple observation that the Cambridge position is untenable, and the data presented here on the red clays of Java, Indonesia, and the residual clay of Auckland, New Zealand, merely add weight to that view. The steep slopes on which the terraced rice fields of Java exist, or the slopes on which much of Auckland’s suburbia has developed, would not exist if it were not for the cohesive component of their shear strength. The cohesive intercept contribution to shear strength appears to be just as real as any other contribution. 5.3.6
Residual Strength
The residual strength of residual soils varies widely, as it does in sedimentary soils, but it does not follow some of the correlations with Atterberg limits or clay fraction derived from sedimentary soils. Figure 5.13 shows a plot of φ against the plasticity index, which is often used to illustrate a
EFFECTIVE STRENGTH PROPERTIES
40
15 19 16 29 26
Residual friction angle φ′r (degrees)
113
28
59 30
25 22 53 27
60
14
17
20
55 2 56 10
18
24
Numbers beside points are sample identification (see Table 1, Wesley, 2002)
21
61
58
57
20
4 23
1 7
11 13
35
33
34
10
31
12 50 63
9
3 38 62 54 5 45 44 42 52 49
46
Clays in general Volcanic ash clays
6
39 32 3637 43 40 48 51 4741 30 0
20
40
60
8 64
80
100
Plasticity Index
Figure 5.13 Residual strength versus plasticity index for a wide range of soils.
Residual friction angle φr′ (degrees)
A-line
40
30
Silt
Silty clay
Clay
20
10 Clays in general Volcanic ash clays Below A-line 0 −100
Above A-line
−50 0 50 Distance above or below the A-line: ΔPI = PI − 0.73(LL − 20)
100
Figure 5.14 The residual friction angle (φ r ) related to position on the plasticity chart (after Wesley 2003).
114
SHEAR STRENGTH OF RESIDUAL SOILS
correlation between the two, as indicated by the dotted line on the chart. However, it is clear that if volcanic ash clays are included in the data, the correlation no longer applies. We can reexamine the data in Figure 5.13 by replotting it versus the distance of the soil below or above the A-line. This is done in Figure 5.14, which shows a much better correlation, applicable to both residual and sedimentary soils. REFERENCES Bishop, A. W., and V. K. Garga. 1969. Drained tension test on London clay. Geotechnique 19(2): 309–313. Brenner, R. P., V. K. Garga, and G. E. Blight. 1997. Shear strength behavior and measurement of shear strength. in Mechanics of Residual Soils, G. E. Blight, ed, Rotterdam, Netherlands: Balkema, 155–220. Meyer, V. M. 1997. Stress–Strain and Strength Properties of an Auckland Residual Soil. Thesis, the University of Auckland. Meyer, V. M., M. J. Pender, and L. D. Wesley. 1999. The very low effective stress behaviour of a residual soil. Proceedings 8th Australia–New Zealand Conference on Geomechanics, Hobart, Vol.2, 877–883. Sharma, B., and P. K. Bora. 2003. Plastic limit, liquid limit, and undrained shear strength of soil—reappraisal. Proc. ASCE Journal Geot. & Geoenvironmental Engineering, 774–777. Skempton, A. W. 1957. Discussion on the planning and design of the new Hong Kong airport. Proceedings Institution of Civil Engineers 7: 305–307. Wesley, L. D. 2003. Residual strength of clays and correlations using Atterberg limits. Geotechnique 53(7): 669–672.
CHAPTER 6
SITE INVESTIGATIONS AND THE MEASUREMENT OF SOIL PROPERTIES
6.1
INTRODUCTION
In this chapter I will be making occasional personal observations about site investigation practices, so I hope the reader (and my publisher) will bear with me if I do this in the first person. The focus of the chapter will be as much on matters of approach to site investigations and laboratory testing as on strictly technical matters. Site investigation methods and laboratory testing are not much different whether the soil is sedimentary or residual, and there is no shortage of publications describing the methods. Some of the methods are basic and others are highly sophisticated. Despite this abundance of tools available to geotechnical engineers, site investigation procedures routinely used in many parts of the world are still surprisingly primitive. There are a variety of reasons for this, including economic and cultural factors, and a general reluctance to change long-established practices (“inertia of the human mind,” to borrow another Terzaghi phrase). However, site investigations and laboratory testing are at the heart of geotechnical engineering, and if there are shortcomings in this area, much of the blame must lie with the geotechnical profession.
115
116
6.2
SITE INVESTIGATIONS AND THE MEASUREMENT OF SOIL PROPERTIES
APPROACHES TO SITE INVESTIGATIONS
The purpose of site investigations and associated laboratory testing is, presumably, to obtain a picture of the geology and soil conditions of the site and to measure specific soil parameters for use in design. If this is the case, then we can rightly ask who it is that needs to obtain this picture. Presumably (again), it is the designer, or the geotechnical engineer most immediately responsible for the geotechnical aspects of the project involved. Geotechnical engineers go about gaining this picture in various ways, depending on the practices of the organization they work for, the size of the project, and their own personal inclinations or preferences. Some (group A) give instructions to their staff or to external agencies, who implement a program of drilling, sampling, laboratory testing, and the like, and deliver a volume (or even several volumes) of data to the geotechnical engineer. Others (group B) involve themselves closely in the program, making site visits, inspecting drill cores and undisturbed samples, and keeping a close eye on all parts of the operation, including laboratory tests. Readers can make their own judgment as to who (in the above examples) gains the better picture. These examples highlight what I think are subtle issues involved in just what is meant by “gaining a picture of the site geology and soil conditions.” It is perhaps better described as getting a “feel” for the type of site and soil one is dealing with. Personally, I find it difficult to get a satisfactory feel for a soil or site from volumes of data, no matter how rigorously it has been obtained and how well it is presented (in many cases it is neither). Nothing conveys the nature of a soil as immediately and reliably as direct inspection of it in exposures, investigation pits, or in core samples, and actually handling and manipulating the soil. Those who appreciate the need for getting a feel for the soil and the site make up group B described above. Those who think a soil can be adequately appreciated from sets of data and written descriptions belong in group A. 6.3
ORGANIZATIONAL AND ADMINISTRATIVE ARRANGEMENTS
Which of the two approaches described above applies in any particular situation depends to some extent on the organizational arrangements under which site investigations and laboratory testing are carried out. Possible arrangements are indicated in Figure 6.1. Model (a) shows an organization that is entirely self-contained. It has its own laboratory and field investigation unit. From a technical point of view this is the ideal arrangement. Model (b) shows an organization that has its own laboratory but no site investigation capability, and model (c) shows an organization that has no investigation or testing capability of its own. Within these three models, there are other variations. Some geotechnical consulting companies have no capacity for machine drilling but have the ability to do manual boreholes using a hand auger. Also, some drilling and
ORGANIZATIONAL AND ADMINISTRATIVE ARRANGEMENTS
The Organization The Geotechnical Team Site investigation unit
Soil mechanics laboratory
The Organization The Geotechnical Team Soil mechanics laboratory Site investigation company
(a) Entirely “in-house”
(b) External site investigation facilities
117
The Organization The Geotechnical Team
Site investigation company
?
Laboratory testing company
(c) External site investigation and laboratory testing facilities
Figure 6.1 Organizational arrangements for site investigations and laboratory testing.
site investigation companies have their own laboratory facilities, so that in model (c) above there is only one external agency involved. Model (c) is the least satisfactory as the geotechnical team has no direct institutional contact with the site investigations or the laboratory testing. It is, of course, possible to set up the contract (or other arrangement) under which the work is done so that the geotechnical team can have members supervising or observing the work at all times as it proceeds. The author’s own experience has been in two government agencies (in New Zealand and Indonesia), a consulting company, and a university, each of which had its own well-equipped laboratory in the same building as the office I occupied. For me, this is the ideal arrangement; it meant I could wander into the laboratory whenever it suited me, and I could give directions to the laboratory staff to let me know when certain samples were arriving or tests being carried out. The opportunity to inspect samples before tests are carried out is a big advantage; it avoids the execution of misguided tests when the soil turns out to be different from expectations. Because of the path my career took (which was by chance rather than design), I would now find it rather frustrating to work as a geotechnical engineer for an organization that did not have its own laboratory facilities. Another obvious advantage of being able to inspect cores and undisturbed samples is that it provides a basis on which to judge the reliability of the results of laboratory tests. The best guide to the expected behavior of a soil is visual inspection and handling of the soil. A complaint I heard from Indonesian colleagues involved in soil testing on a couple of occasions was of geotechnical engineers from abroad who pointed out their measured water content values were impossibly high and told them to repeat the tests. The soils involved were volcanic clays and the high water contents were normal. If the engineers involved had visited the sites from which the soil came and observed it in the field, as well as actually handled it, they would have known it was an unusual material and would not have expected it to conform to normal behavior.
118
6.4
SITE INVESTIGATIONS AND THE MEASUREMENT OF SOIL PROPERTIES
PLANNING SITE INVESTIGATIONS
Carrying out a site investigation should involve the following: 1. Assembling all the information already available on the site. This is the proper starting point, and can involve a range of possibilities. In some cases the engineer may already be familiar with the probable soil conditions from past experience on neighboring sites. At the other extreme, the site may be in a remote area where there is little or no human habitation, and no engineering projects of significance. However, it is unlikely these days that there are no topographical or geological maps available from which some useful information can be gained. 2. Obtaining all relevant information on the nature of the project. This could involve loads and settlement tolerances, excavation depths, scope of earthworks, time constraints, and the like. 3. Visual inspection of the site. This may or may not be a productive exercise, depending on the nature of the site. In some situations, where the site is dead level and there are no buildings in the vicinity, the visit may be of little value. However, it is rare for a site visit not to produce some useful information; there are normally hill slopes or road cuttings to be inspected, or existing buildings in the vicinity whose owners or designers may be happy to part with information about their foundation types and performance. The author was once involved in a ground treatment project involving filling and preloading a soft clay site in northern Malaysia, to make it suitable for light industrial use. Site investigations were carried out involving the usual drilling, sampling, and laboratory testing, but by far the most useful piece of data came during a site visit when the consultants on a neighboring site with a well-monitored preload already in place were kind enough to make available the settlement records from their preload. The comment just made has relevance to quite a lot of research that is carried out into particular soil types. The research produces a report or a paper with impressive tables of data coming from field and laboratory tests. It may include many elegant figures, including stress paths from sophisticated triaxial tests. But it may be almost bereft of direct observations or information on how the soil behaves in the field. Such observations should be an integral part of investigation or research into a particular soil type. 4. Identifying the critical issues that should be the main focus of the investigation, and ensuring that the budget is at least large enough to cover these critical issues. Budgets sometimes appear to be spent inefficiently by a ritualistic approach to site investigations—“we did this or that on the last investigation so we should do the same here”
FIELD WORK
119
or “we always put down four boreholes on building sites.” Also, a considerable amount of laboratory testing sometimes appears to serve no useful purpose other than help decorate the pages of a report. It does not provide information that is helpful for design purposes. 5. Planning the actual program of tests, but doing so in such a way that there is adequate flexibility to allow changes to be made if soil conditions revealed during the course of the investigation show that such changes are needed. 6. Supervising the site work, particularly the boreholes and the logging of the soil profile. In the author’s view, the geotechnical team in charge of the project should always have its own representative at the drilling rig to monitor progress and log the cores. Leaving it to drillers to keep the log normally means a log of uncertain reliability. Drillers are seldom adequately trained in soil logging procedures, and are too busy operating the rig to have time or inclination to keep a careful log. 6.5
FIELD WORK
We will now consider some technical aspects. The most important site investigation tools are normally boreholes and penetrometer tests, so a few observations on these may be of value. 6.5.1
Hand Auger Boreholes
Hand auger boreholes obviously have considerable limitations, but can still be an appropriate method of investigating soil conditions in particular situations, especially those involving relatively light buildings on clay. They can be a very satisfying form of site investigation, especially if one is present on site taking part in the operation. The ease or difficulty of turning the auger and advancing the hole gives a good indication of the stiffness of the soil, and together with the extracted cuttings makes possible a fairly reliable description of the soil. Hand auger holes are without doubt vastly superior to wash drilling as a means of obtaining an accurate soil profile. The depth possible with hand auguring is limited, although in medium to stiff clays probably not as limited as is commonly supposed. In countries where labor costs are low, innovative techniques have been developed and it is not uncommon for hand auger holes to be taken to 10 m or more. This normally necessitates the use of a derrick and pulley system of some kind to lift the rods up and down. It may also make use of an extra large handle at the top of the drill rods, so that one or two people can perch on it to provide extra weight while it is rotated.
120
SITE INVESTIGATIONS AND THE MEASUREMENT OF SOIL PROPERTIES
6.5.2
Machine Boreholes
A rotary drilling rig makes possible a wide variety of drilling techniques that can be employed as needed to suit the type of soil being investigated. The easiest material in which to put down a borehole and recover intact cores is firm to stiff clay. In this material, coring can be done using a single tube core barrel (also known as an open barrel) by simply pushing the barrel into the clay without rotation or the use of drilling water, assuming a rig of substantial weight is doing the drilling. Friction or adhesion between soil and barrel can be relied on to retain the core in the barrel. A core is brought to the surface and extruded from the barrel; the barrel is reinserted in the hole and a subsequent core is taken. Rotary drilling rigs are equipped with water pumps, and water pressure can be made use of to extrude the core. With this procedure, a continuous core is obtained over the full depth of the hole. The procedure is illustrated in Figure 6.2a. Regarding the inspection and recording of cores from boreholes, especially clay cores, it is not good practice to log the cores from the outside, so
1. Core barrel pushed into soil at bottom of hole
2. Core barrel removed and core extruded
3. Hole again advanced by pushing in the sampler
(a) Advancing the hole by continuous coring with a single tube core barrel
(1) Cut a slot lengthways along the core
(2) Pull the two halves apart to expose the texture of the soil
(b) Examining the core by “opening” it longitudinally into two halves
Figure 6.2 Continuous coring and core examination in clays.
FIELD WORK
121
to speak, because the outside surface is smoothed by the action of the core barrel, and the texture of the soil is not apparent. My personal preference for inspecting and logging cores is illustrated in Figure 6.2b. This is to cut a slot along the core and then manually split the core into two halves. In this way the texture (layering and discontinuities, etc.) is clearly displayed, and can be accurately described in the log. Continuous coring and logging in this way provides a very complete and reliable record of the soil profile over the full depth of the borehole. In contrast to the above method, the common practice still found in many countries of putting down site investigations boreholes by wash drilling, even in clays, must be regarded as extremely primitive or backward, and provides only a crude indication of soil conditions. 6.5.3
Penetrometer Testing
The dominant penetration tests are undoubtedly the standard penetration test (SPT), and the Dutch cone penetrometer, commonly called the CPT test. Considered only in terms of technical details, the CPT test appears much superior to the SPT test. Advantages of the CPT include the following: 1. It tests the soil in an undisturbed state. 2. It is a static test, which is desirable for any test that is to be correlated with other soil properties of a static nature. 3. It provides a very detailed record of the soil profile. A glance at CPT results such as those given in Figure 9.9 (of Chapter 9) provides an immediate and clear picture of variations in the strength or density of the soil. The results in Figure 9.9 were both obtained using mechanical CPT devices, with readings taken at the usual interval of 20 cm. With today’s modern electrical cones, readings are almost continuous, so the definition of the profile is even more precise. 4. It has been adapted to measure soil properties other than cone point resistance, such as skin friction and pore pressure. Disadvantages of the CPT include: 1. It can only penetrate materials up to a certain stiffness or density. Advantages of the SPT test are the following: 1. It can be used in much harder or denser materials than the CPT test. 2. It takes a sample of the soil being tested. Disadvantages of the SPT test are the following: 1. It is not a standardized test, despite its name. Some countries, such as the United States, carry out the test using a cable permanently attached to the drop weight. This means about 40 percent of the energy of the
122
SITE INVESTIGATIONS AND THE MEASUREMENT OF SOIL PROPERTIES
blow is lost through friction because the hammer drags the cable with it. 2. It is carried out only at fairly wide intervals, so the record of the soil profile is far less complete than that obtained from the CPT test. 3. The soil being tested may not be undisturbed prior to the conduct of the test. This is because the test is carried out in a borehole and the process by which the borehole is put down may disturb the soil. An advantage of the SPT test that is not an attribute of the test itself is that a great deal of work has been done to correlate the test with other soil properties, in particular the relative density of sand and the susceptibility of sand to liquefaction. In contrast, there is less information available on correlations with the CPT. However, the CPT has the advantage that its point resistance in granular materials, especially sand, is a direct measure of the expected point resistance of a full-scale driven pile. The choice of either of these tests for use in residual soils is a matter of judgment based primarily on the type of soil involved and the objectives of the testing. If the soil is such that the CPT cone can penetrate it, then the CPT will be the better test to employ. However, if the material is coarse and dense, with unpredictable hard zones, then the SPT test is likely to be more appropriate. 6.6
BLOCK SAMPLING
The influence of sample disturbance on laboratory test results is well known and hardly needs further discussion here. It is perhaps worth emphasizing that the effect of disturbance is most significant when the material is of high sensitivity. This is only natural as high sensitivity implies a rather open, fragile structure that is likely to be broken down by disturbance even if the disturbance involves only very small deformations. Figure 6.3 illustrates the difference in degree of disturbance between tube samples and block samples. Figure 6.3a shows the results of triaxial tests on a soft, sensitive, Norwegian clay. The curves from the block samples have much sharper peaks than those from the tube samples. Figure 6.3b shows the results of standard oedometer tests on a clayey silt of moderate sensitivity. The average compressibility of the tube samples is about double that of the block samples. With nonsensitive clays the effect of sample disturbance is not likely to be very significant. The tropical red clays of Java are an example of the latter; completely remolded samples appear to have the same compressibility as the undisturbed soil. Residual soils are not generally of high sensitivity, although there are certainly varieties that can be of high or very high sensitivity. In particular, volcanic clays can be of very high sensitivity, but they can also be of low or “normal” sensitivity, as discussed in more detail in Chapter 9. There are
BLOCK SAMPLING
123
30 Pressure (kPa) 20
0
10
100
200
300
400
500
5
1
2 Strain (%)
Compression (%)
(kPa) σ′1 − σ′3/2
0
0
3
Block sample Piston sample
10
15 Block samples Tube samples
50 20
(b) Oedometer tests on a moderately sensitive clayey silt ( Auckland, New Zealand)
0
50 σ′1 + σ′3/2
100 (kPa)
(a) Triaxial tests on sensitive Norwegian clay (after Lacasse et al., 1985)
Figure 6.3 Influence of sample disturbance in tube samples.
good reasons, therefore, for obtaining block samples from some residual soils, especially when reliable strength or compressibility measurements are required. The advantages of block sampling with residual soil, however, are not confined to minimizing sample disturbance. Residual soils are often heterogeneous and may contain coarse particles in a clayey or silty matrix. To carry out sensible tests on such soils requires a large sample size, which is only possible with block sampling. Procedures for block sampling are indicated in Figure 6.4. The key to successful block sampling is experience and having the right tools and materials available. The right kind of cutting implement is particularly important. Cutting implements may be knives, ordinary saws, or wire saws, and it is a matter of trial as to which is the most appropriate. Readers who have undertaken block sampling of a range of soil types will know that some soils are a dream to cut and trim to shape and others are a nightmare. Residual soils are to be found in both groups. The properties that call for block sampling (coarse material and random
124
SITE INVESTIGATIONS AND THE MEASUREMENT OF SOIL PROPERTIES
Block sample hand trimmed from side of cutting or base of excavation—ready to be “pushed” onto the timber plank
Sample placed on plank and wrapped in plastic to prevent drying Sample placed in box and space filled with packing material to support and protect the sample
Timber plank suitable for carrying the block sample
Figure 6.4 Block sampling from a bank exposure.
discontinuities) are also the properties that tend to make sample trimming difficult. There are no certain answers as to the best way to trim such soils; experience alone is probably the only and best guide. Figure 6.4 shows some of the other items essential for block sampling. These include plastic wrapping to seal the sampling and prevent drying. Wrapping the sample in plastic also provides considerable support for the sample during further handling and transport. Once the sample is wrapped it should be placed in an appropriate container. The sample size should be tailored to suit the container, with just enough space around the sides to lower it into the box and then pack the space with sawdust or sand or some other suitable material. 6.7
IN SITU SHEAR TESTS
The problems of sampling and conducting laboratory tests on residual soils can be overcome to some extent by carrying out tests in situ. Compressibility can be measured by conducting plate-loading tests, which do not involve any great difficulty apart from the costs associated with the weights or anchors needed as reaction for the downward thrust. Sample preparation is extremely straightforward, as it only involves preparing a flat surface on which to place the plate. Measuring the shear strength is considerably more difficult because it involves substantial excavation and sample preparation, as well as special equipment. Brenner et al. (1997) provide an overview of several techniques that are used for such tests. A possible setup for the test along the lines of that developed by the Hong Kong Geotechnical Control Office (Brand et al. 1983) is shown in Figure 6.5. A pedestal of the intact soil is first
IN SITU SHEAR TESTS
Shear box—two halves clamped together
125
Original ground surface
Sample trimmed using saws and knifes—free standing or with aid of the shear box itself
(a) Sample trimmed in situ
Gauge to measure deflection
Vertical load applied using kentledge or other weights, or anchors and jacks Load cell or proving ring
Hydraulic ram
Frame of loading device (b) Loading frame assembled, vertical load applied, and sample sheared
Figure 6.5 In situ shear box test in a shallow pit.
carved with great care to fit the precise dimensions of the shear box, which is normally square with side dimension from about 0.3 to 0.5 m. In some soil conditions the box itself can be used to assist in trimming and preparing the pedestal. Once the box is in place over the pedestal, the device itself is assembled and the normal (vertical) load applied. This can be done by simply using weights (kentledge), or by installing anchors in the ground, and using a hydraulic jack. If the test is done in a tunnel, the roof of the tunnel can be used to jack against. With the arrangement in Figure 6.5, the reaction for the application of the horizontal force is provided by the frame of the device itself. An alternative is to use the wall of the pit to jack against. Tests of this kind have the advantage that a large sample of the soil is tested in its original state. On the other hand, because of the large sample area and the limited capacity for vertical load application, the maximum normal stress is relatively low. This may not be an important limitation, as the stress levels of interest may not be high. These tests are most commonly
126
SITE INVESTIGATIONS AND THE MEASUREMENT OF SOIL PROPERTIES
used in connection with slope stability issues where the relevant normal stress may be quite low. 6.8 6.8.1
LABORATORY TESTING Index or Classification Tests
The most important point to be aware of here is the possible effect of predrying before carrying out particle size or Atterberg limit tests. Volcanic clays with high allophane or halloysite content can change from moderately plastic clays to nonplastic sandy silts if oven dried, and some suffer a severe change in properties with air drying alone. Hence, sample preparation for testing should involve no drying beyond that needed to conduct the test. Particle size tests should be carried out starting with a sample at natural water content and “wet sieving” the material if sieving is necessary. The weight of dry material can be determined either by careful measurement of the water content at the start, or by retaining all sieved material and drying it (by evaporating the water) after the test. The latter procedure is tedious and time-consuming. As indicated earlier in this book (Chapter 3), particle size measurements of residual clays are not particularly informative, at least not in comparison to Atterberg limits. With regard to Atterberg limits, one small point worth mentioning concerns the particle size limit set by the test standard, namely 0.425 mm. There does not seem to be any real virtue in keeping rigidly to this limit. If the size and proportion of material coarser than this limit is not such that it interferes with the execution of the test, then the test standard should be ignored and the test carried out on the complete material. The reason for this is that if material is removed, then the test results are of less value, because they no longer apply to the actual material one is dealing with. For example, the liquidity index is a very useful parameter, and its value becomes problematical if the water content is measured on the complete sample but the Atterberg limits on only a fraction of the sample. 6.8.2
Tests on Undisturbed Samples
The following general points are worth noting: 1. Sample preparation can be a real challenge with some residual soils. With soils that are difficult to trim in the laboratory using knifes, saws, or wire saws, it is preferable to sample them in the field with the diameter required for the laboratory test, at least if the samples are being obtained from a drilling operation. The tube or core barrel used should be the same diameter as the triaxial samples to be tested in the laboratory.
LABORATORY TESTING
127
2. Because of their heterogeneity, it is often desirable to carry out tests on large-diameter samples. The higher permeability associated with residual soils generally means that this can be done without the slow testing rates that would normally be required for large samples with most sedimentary soils. With regard to oedometer tests, it is very important with residual soils that all possible precautions be taken to minimize sample disturbance. One common source of disturbance occurs when the sample is trimmed into the oedometer ring, or the ring pushed into the soil. It is surprising that very primitive methods are still commonly used for this procedure; the ring is often pushed into the soil by hand with the result that it “wobbles” somewhat from the vertical and a small gap is created between the sample and the inside of the ring. To ensure that this does not occur, it is highly desirable that more rigorous procedures are used, such as that illustrated in Figure 6.6. This shows a vertical lathe designed by the author in his early years as a geotechnical engineer. The sample is first trimmed carefully to a convenient size so it can be placed on the rotating platform shown. The oedometer ring is placed in the ring holder and is then lowered onto the top of the sample. Soil is then carefully trimmed from immediately below the oedometer ring, using a wire saw or a sharp knife, and the ring pushed down gently as Platform to rotate sample and apply vertical weight
Sample ring holder Openings to view top of sample Oedometer sample ring Soil sample Turntable Turntable bearing
Figure 6.6 Vertical lathe for preparing samples for oedometer tests.
128
SITE INVESTIGATIONS AND THE MEASUREMENT OF SOIL PROPERTIES
the trimming progresses. Alternatively, weights can be placed on the top platform to provide a small and constant downward force. Once the sample protrudes above the top of the sample ring, the ring holder is raised and the sample and ring removed. The top and bottom are then trimmed to provide flat surfaces. The ring holder has holes in its upper surface so that the progress of the sample into the ring can be observed as the trimming proceeds. This method of preparation has been found to be very convenient and ensures that sample disturbance is minimized. 6.8.3
‘‘Computer Errors’’ in Processing Laboratory Test Results
This section is a short diversion from the more serious matters that books of this nature are expected to restrict themselves to. Figure 6.7 shows two examples of laboratory test results, taken from completely separate sources; readers may find them mildly amusing. The first, Figure 6.7a, taken from Liquid limit
Plastic limit
Plasticity index
Soil A
96.05
45.50
50.55
Soil B
30.39
21.92
8.47
(a) “High precision” Atterberg limit tests 1.20
1.15
Void ratio
1.10
1.05
1.00
0.95
0.90
1
100
10 Pressure (kPa)
(b) Computerized curve fitting to a highly improbable oedometer test result
Figure 6.7 Some curious (and amusing?) laboratory test results.
1000
CORRELATIONS WITH OTHER PROPERTIES AND PARAMETERS
129
a technical paper, shows results of “high-precision” Atterberg limit tests on two samples. The second, Figure 6.7b, taken from a geotechnical report prepared by a very reputable international consulting company that specializes in geotechnical work, shows a highly improbable e-log p curve from an oedometer test. Determining Atterberg limits to an accuracy of two decimal places is an interesting thought, and an oedometer test result having the shape shown is even more interesting! An e-log p curve such as this is something that university lecturers possibly expect from first- or second-year students, but probably not from reputable testing laboratories. How is it that results such as these find their way into important documents? Some remarks were made earlier regarding the desirability of geotechnical engineers involving themselves directly in the supervision of both field investigations and laboratory testing, and perhaps these results provide some justification for those remarks. However, we could absolve geotechnical engineers of any responsibility and place the blame on computers. One of the great blessings of the computer age has been to make humans infallible, as we all know well! All mistakes, from incorrect bank statements to faulty flight bookings, are almost always the result of “computer errors”—nothing at all to do with human shortcomings! But, seriously, it is possible, indeed probable, that results such as those in Figure 6.7 are the result of overreliance on computer processing. If the raw Atterberg limit test data are fed into a computer for processing, then the computer may well produce results to two decimal places—in fact, to as many decimal places as it is capable of. However, it is hard to understand how anyone familiar with Atterberg limit tests could imagine they could have an accuracy of two decimal places! Similarly, an oedometer curve of the shape shown is hardly likely to have been drawn by human hand— it is likely to have been totally processed and printed out by a computer, without even a passing glance from human eyes. With respect to Atterberg limits, it should perhaps be emphasized that these limits are normally whole numbers, according to definition, convention, and standards. Hopefully, with more direct involvement of geotechnical engineers in laboratory supervision and report preparation, and less reliance on computer processed results, anomalies such as those in Figure 6.6 will be less likely to occur. 6.9 CORRELATIONS WITH OTHER PROPERTIES AND PARAMETERS
With respect to correlations between field tests and other soil properties, residual soils in many cases may conform well to established correlations, but in some cases may depart from them dramatically. The two properties
130
SITE INVESTIGATIONS AND THE MEASUREMENT OF SOIL PROPERTIES
of most importance in geotechnical design for which various correlations with field tests have been established are probably the undrained shear strength of clays and the relative density of sand. A few comments on these correlations are made in the following sections. 6.9.1
Undrained Shear Strength
The cone resistance from the CPT test is normally related to the undrained shear strength by the following equation: Su = where Su qc σ vo Nk
= = = =
qc − σvo Nk
undrained shear strength cone resistance total vertical stress correlation coefficient
The value of Nk is generally between about 10 and 20. Its value depends both on the type of soil involved and the way in which the undrained shear strength to which it is being related is measured. Values of Su can be measured in a variety of ways, the most common of which are unconfined compression or undrained triaxial tests, and vane tests. The value from triaxial tests will depend on whether the test is an extension or compression test, and neither value can be expected to agree exactly with the vane value. Figure 6.8 shows the results of some investigations into the qc versus Su relationship carried out by the geotechnical staff at the University of Auckland. The investigations were carried out at three sites where the clay was of volcanic origin, but of considerable age and not containing a high proportion of the distinctive clay mineral allophane. The clay was generally of moderate to high sensitivity. The best-fit line is close to an Nk value of 12, which is in the range of values normally encountered in soils of moderate to high sensitivity. If the data in Figure 6.8 are examined in more detail, it is evident that the correlation for the Hamilton site has a lower degree of variability than at the other two sites. This is apparently due to the fact that the soil at Hamilton is more homogeneous than at the other sites. Using the above value of Nk , the values of undrained shear strength calculated from the CPT tests and the vane tests are shown together in Figure 6.9. It is seen that the correlation is very good over that part of the profile where there are no sudden changes in the strength of the soil. When the CPT plot shows sharp fluctuations the agreement is less satisfactory. This behavior suggests that the large departures from the correlation line in Figure 6.8 are not due to variations in the correlation itself but to the fact
CORRELATIONS WITH OTHER PROPERTIES AND PARAMETERS
131
4000 Tauranga Ramarama Hamilton
qc = 12Su + σvo
(qc − σva )
(kPa)
3000
2000
1000
0
50
100
150
200
250
Undrained shear strength from field vane (kPa)
Figure 6.8 Correlation between cone resistance and undrained shear strength in volcanic clays (courtesy M. J. Pender of the University of Auckland). Undrained shear strength (kPa) 0
1
200
400
600
800
CPT (Nk = 12) Field vane CU triaxial test Laboratory vane
2 3
Depth (m)
4 5 6 7 8 9
Hamilton site
10 11
Figure 6.9 Undrained shear strength from CPT tests and from direct measurements (courtesy M. J. Pender of the University of Auckland).
132
SITE INVESTIGATIONS AND THE MEASUREMENT OF SOIL PROPERTIES
that the soil testing in each case is not the same. The distance between the two tests was not more than 1 m, but with this distance and the practical difficulty of testing at identical depth, the element of soil tested in each case may well not have the same properties. 6.9.2
Relative Density of Sand
Sand is not actually a residual soil, but sands found in the climatic conditions that produce residual soils tend to have “non-conformist” properties that geotechnical engineers need to be aware of; for this reason mention is made of them here. Sands found in the wet tropics, especially those derived from volcanic materials, are seldom clean, hard-grained materials. They are likely to contain a significant proportion of both soft grains and silt-sized material. Correlations between cone resistance and relative density found in soil mechanics literature have been developed primarily from the study of clean quartz sands of a fairly uniform particle size. Such correlations are thus unlikely to be valid for sands of different particle hardness and grading. An extreme example of the unusual behavior of soft-grained sands is given in Chapter 9 (on volcanic soils). This describes the properties of
0.0
0
10
Cone resistance qc (MPa) 20 30 40
50
Vertical effective stress σ′vo (MPa)
Pumice sand, loose Pumice sand, dense 0.1
0.2
0.3 1
3
1
3
0.4 Dr = 40% 2
Dr = 80% 2
0.5 1 Schmertman (1976) Hilton Mines sand—high compressibility 2 Baldi et al (1982) Ticino sand—moderate compressibility 3 Villet & Mitchell (1981) Monterey sand—low compressibility
Figure 6.10 Influence of compressibility on correlations (adapted from Lunne et al. 1997).
REFERENCES
133
a pumice sand and an investigation into correlations between CPT cone resistance and relative density. Pumice particles are primarily quartz in composition, but contain a dense network of tiny holes known as vesicules, making them sufficiently soft that they can be crushed between fingernail and a glass surface. The startling fact that emerges from the investigation is that the cone resistance is not significantly affected by the relative density of the pumice sand. The investigation confirms what is already well established (although not always recognized) in the literature, namely that correlations between cone resistance, overburden pressure, and relative density are valid only for the sand from which the correlation was established. Figure 6.10 shows the curves correlating cone resistance to relative density and overburden pressure for three sands of varying compressibility. This illustrates the dependence of the correlation on the hardness of the particles, since this is the principal factor controlling compressibility. Also shown are the results for the pumice sand, which was probably of considerably higher compressibility than the Monterey sand. REFERENCES Baldi, G., R. Belloti, V. Ghionna, M. Jamiolkowski, and E. Pasqualini, 1982. Design parameters for sands from CPT. Proc. 2nd European Symposium on Penetration Testing, ESOPT 11, Amsterdam, Vol, 2, 425–432. Brand, E. W., H. B. Phillipson, G. W. Borrie, and A. W. Clover. 1983. In situ shear tests on Hong Kong residual soil. Proc. 2nd Int. Symp. Soil and Rock Investigations by in-situ Testing, Paris, 13–17. Brenner, R. P., V. K. Garga, and G. E. Blight. 1997. Shear strength behaviour and the measurement of shear strength in residual soils. Chapter 7 in Mechanics of Residual Soils, G. E. Blight, ed. Rotterdam: Balkema, 196–203. Lacasse, S., T. Berre, and G. Lefevbre. 1985. Block sampling of sensitive clays. Proceedings 11th International Conference on Soil Mechanics and Foundation Engineering, San Francisco, Vol. 2, 887–892. Lunne, T., P. K. Robertson, and J. J. M. Powell. 1997. Cone Penetration in Geotechnical Practice. London: Blackie Academic and Professional. Schmertman, J. H. 1976. An updated correlation between relative density, Dr , and Fugro-type electric cone bearing, qc Contract Report DACW 39-67 M 6646, Waterways Experiment Station, Vicksburg, Mississippi. Villet, W. C. B., and J. K. Mitchell, 1983. Cone resistance, relative density and friction angle. Cone Penetration Testing and Experience: Session at the ASCE National Convention, St Louis, ASCE, 178–207.
CHAPTER 7
BEARING CAPACITY AND EARTH PRESSURES
7.1
INTRODUCTION
We now move on from considerations of soil properties and their measurement to the three main design situations faced by geotechnical engineers: 1. Design of foundations of structures 2. Design of retaining walls 3. Assessment of slope stability These are primarily stability issues; each case is governed by the shear strength of the soil, and the design process involves the analysis of a possible failure mechanism or a number of possible mechanisms. In the case of foundation design, this mechanism is normally based on theoretical considerations and is of a fixed predetermined theoretical form. No trial process is involved. In the case of retaining wall design, the failure mechanism may also be predetermined on the basis of theoretical considerations. However, this is the case only when the soil is homogeneous, no seepage pressures are present, and the ground surface behind the wall is level. For more general conditions not involving these factors a trial process involving failure wedges is necessary (the Coulomb wedge method). Finally, in the case of slope stability, there is no predetermined failure mechanism, and the stability analysis involves a tedious trial process to establish the mechanism with the lowest safety margin (safety factor). 135
136
BEARING CAPACITY AND EARTH PRESSURES
In addition to assessing stability, the question of deformations may also need to be addressed. In the case of foundation design, this is likely to be a more important consideration than stability, and is often the controlling factor that determines the feasibility of a surface foundation. With retaining walls, deformation is not often a controlling factor, partly because small deformations of the wall are of little consequence, and partly because the designer has control over the stiffness properties of the wall, and can adopt a design that controls deformations. With slope stability, deformations are rarely of significance; the deformations of a slope with an adequate margin of safety against slip failure are negligible for practical purposes. We can state at the outset that the concepts and methods used in the above design situations are not fundamentally different between residual and sedimentary soils. For this reason, the reader need not look for new bearing capacity theories or radically different slope stability methods, and may find the material covered here rather on the light side. As already observed, residual soils are generally fairly well behaved, and they present only limited challenges as far as bearing capacity and retaining wall design are concerned. However, the issue of slope stability is an ever present challenge in a great many projects in residual soils. The challenge arises for various reasons, the main ones being the following: 1. The difficulty of determining soil strength parameters in heterogeneous materials 2. The lack of a constant steady-state seepage regime in the slope, and the difficulty of estimating the most adverse (worst-case) seepage situation For these reasons, we will not spend a lot of time on bearing capacity or retaining wall design, which will be covered in this chapter. We will spend considerably more time on slope stability in a separate chapter. 7.2
BEARING CAPACITY AND FOUNDATION DESIGN
For surface foundations on clays, conventional practice is to base bearing capacity estimates on undrained shear strength, and there are no adequate reasons to use a different approach with residual soils. The following points should, however, be noted: 1. The coefficient of consolidation of residual soils is generally much higher than that of sedimentary soils, and it is probable that with many foundations built on residual soils the pore pressures dissipate as rapidly as the load is applied, and postconstruction settlement is likely to be very small. If this is the case, then the loading is drained rather than undrained, and an argument can be made that the bearing
BEARING CAPACITY AND FOUNDATION DESIGN
2.
3.
4.
5.
6.
7.
8.
137
capacity will be more dependent on effective stress parameters than on undrained strength. The counter argument is that the use of effective strength parameters from residual soils leads to unrealistically high values of bearing capacity, in the same way that it does with sands. Even if the soil could support the pressure determined from an effective stress estimate, the deformation involved would be far greater than the structure could tolerate. A further argument in favor of using undrained strength is that many countries where residual soils are predominant are also in earthquake zones, and earthquake loading will always be essentially undrained, even in residual soils. There is also a purely pragmatic argument in favor of the undrained strength approach, and that is that it is much easier to measure this strength than the effective stress parameters. With cone penetrometer testing, for example, a comprehensive picture of the undrained strength, and its variation across the site, is easily obtained, whereas similar comprehensive information on the effective stress parameters would not be possible. As observed in the introduction, the feasibility of using surface foundations for buildings normally depends on settlement considerations rather than bearing capacity, so there is little to be gained from seeking higher bearing pressures than those given by undrained analysis. Some residual soils are rather heterogeneous so that the determination of parameters for bearing capacity estimates is not straightforward. There are thus sound arguments for using plate-loading tests, or some other form of in situ testing to try to determine reliable settlement behavior and bearing capacity estimates. The influence of discontinuities in the form of planes of weakness on bearing capacity appears to be much less than it is on slope stability. This is to be expected, as the shear surfaces associated with the bearing capacity failure mechanism are very restricted in location and inclination, whereas slope failure can occur on almost any plane of weakness. Construction projects in residual soils are often in hilly areas and foundations are built on platforms cut into the slope. Bearing capacity estimate must therefore be made taking account of the slope inclination.
The measurement of undrained strength can be made by various means, similar to those used for sedimentary soils, and will not be covered here. However, one word of warning may be appropriate. It is likely that in some climates there may be significant changes in the undrained strength near the ground surface due to changes in water content caused by seasonal influence. Figure 7.1 shows the result of a cone penetration test (CPT),
138
BEARING CAPACITY AND EARTH PRESSURES
Cone resistance (MPa) 0
5
10
15
20
?
Depth (m)
Seasonal effect?
10
20
Figure 7.1 Cone penetrometer test (CPT) showing possible dry weather influence near the surface.
which shows a stiffer zone of soil near the surface. With a result such as this, we should be wary of accepting the result at face value. It may be a dry weather effect, and for design purposes we may be better to assume that the mean value deeper down extends back up to the surface. The issue of settlement estimates has been covered fairly extensively in Chapter 4, with additional comment in Chapter 6. The main points made include the following: 1. High-quality undisturbed samples are needed to obtain reliable measurements of compressibility from laboratory tests. 2. We should put aside interpretations of compressibility based on the behavior of sedimentary soils. In particular, we should abandon the conventional practice of using a log scale for pressure. This practice presents a distorted picture of behavior, and routinely leads to wrong assumptions and misinterpretations. 3. The term vertical yield pressure should be used in place of preconsolidation pressure to indicate the stress level at which some residual soils show increased compressibility, presumably due to some form of structural collapse. Some residual soils show clear yield pressures and others show no indication of yield. 4. The conventional oedometer test is likely to be an unreliable way of determining the coefficient of consolidation of many residual soils. Their rate of consolidation is such that pore pressures dissipate too quickly for useful measurements to be made.
EARTH PRESSURE AND RETAINING WALL DESIGN
139
Height of fill placed (m) 0
10
20
30
40
4U
2U
Vertical strain (%)
1
3U
2
3M 3L
3
2L
U = Upper layer M = Middle layer L = Lower layer
4
Figure 7.2 Compression of foundation layers at a dam site in Fiji (after Prusza et al. 1983).
An interesting record of settlement on a residual soil is shown in Figure 7.2. This shows settlement as fill is placed on a dam site in Fiji. The graphs all show a yield stress at about 20 m of fill, which would be a vertical stress of about 350 kPa. The soil involved was weathered gneiss. A final observation regarding settlement estimates for foundations on residual soils is that the actual settlement is generally less than predicted. This issue is discussed in greater detail in Section 4.9.2 of Chapter 4. 7.3
EARTH PRESSURE AND RETAINING WALL DESIGN
Methods for estimating earth pressures and designing retaining walls are no different in principle with residual soils than what they are with sedimentary soils. At the same time, there are characteristics of residual soils that stimulate or necessitate differences in the design approach. Two of them will be described briefly here. The first relates to estimates of earth pressures when the heterogeneous nature of the soil makes determination of strength parameters by normal methods exceedingly difficult. The second is the use of clays in reinforced earth retaining walls. 7.3.1
Earth Pressure to Retain Cuts in Steep Slopes
The design of retaining walls with residual soils involves the same difficulties as those faced in estimating slope stability by analytical means, namely uncertainty regarding soil strength parameters and pore pressure conditions.
140
BEARING CAPACITY AND EARTH PRESSURES
It is possible in some situations to overcome this difficulty by back-analysis methods. The following is an example. It is an interesting example because it overcomes uncertainties with respect to both soil strength and pore pressure conditions. A not infrequent challenge faced by geotechnical engineers (including the author, on a number of occasions) is the retention of vertical (or near vertical) cuts made in steep slopes of sufficient extent that they are essentially infinite in relation to the size of the cut being made. The situation is illustrated in Figure 7.3. The cut may be only to create a little more level ground at the back of a house (for the owner’s barbecue), or it could be for a major new highway. In either case, the challenge is the same. Visual inspection, plus a little common sense, may indicate that the steepness of the slope is such that any cuts made in it are likely to cause a slip, or even a major landslide. Thus, some form of retention is considered essential. Whatever the form, its design (if it is to have some semblance of rational basis) requires a knowledge of the shear strength parameters of the soil and the seepage conditions in the slope. Trying to determine appropriate shear strength parameters for the residual soils of which many steep slopes are composed can be an almost impossible challenge, as also can trying to determine the worst seepage condition. A possible way to overcome this problem is to assume the slope has a safety factor of unity, and by back-analysis determine the shear strength parameters of the soil, or at least put some constraints on the values of these parameters. This can be done by analyzing the equilibrium of a soil segment above a possible failure plane, as shown in Figure 7.4. A comprehensive account of the complete analysis is to be found in Wesley (2001), and only an outline is given below. nt
te
x ”e
d
ite
im
pe
ep
e St
P
slo
of
l un
ta
r te
a
W
e
bl
“
e
l ab
)
(?
ri
va
Residual soil or highly weathered rock c′ = ? φ′ = ? Required P = ?
Figure 7.3 Retention force for cuts in steep “infinite” slopes.
EARTH PRESSURE AND RETAINING WALL DESIGN
141
β
Hw H e
w Flo
ce
fa
ur
s nd
ou
Gr
β
e
c rfa
ne
la ep
lur
uip
Ph
su
Eq
tic
a re
lin
ote
e ibl
fai
al
nti
ss
Po
e
lin
Figure 7.4 Stability of soil above a possible failure plane in an infinite slope.
The analysis produces the following expression for the safety factor: c γw Hw tan φ SF = + 1− (7.1) 1− γ H cos β sin β γ H tan β For the case of limiting equilibrium, (SF = 1), this becomes c γw Hw tan φ =1− 1− 1− γ H cos β sin β γ H tan β
(7.2)
For the case of a slope in which no water table is present, the expression becomes c tan φ =1− (7.3) γ H cos β sin β tan β And for the case of a slope with a water table at the ground surface, the expression becomes c γw tan φ =1− 1− (7.4) γ H cos β sin β γ tan β where γ and γw are the unit weights of the soil and water, respectively. These equations are essentially the same as those given by Taylor (1948), in a slightly different form. They show clearly that for a given value of φ the value of c needed to maintain equilibrium is proportional to the depth H, as pointed out by Taylor (1948). For our present purpose they define
142
BEARING CAPACITY AND EARTH PRESSURES
a relationship between c and φ at which the slope is stable, for a given slope inclination, depth of cut, depth of water table, and soil unit weight. Provided we make an assumption about the value of one or the other of these parameters we can proceed to determine the value of the necessary retention force by using a Coulomb analysis as shown in Figure 7.5. The expressions for the forces involved are as follows: P = W tan α − φ + U
sin φ cos φ − C cos (α − φ ) cos (α − φ )
(7.5)
cos α cos β 1 W = γH2 2 sin (α − β)
(7.6)
γw cos2 β 1 U = γH2 2 γ sin (α − β)
(7.7)
cos2 β sin β tan φ 1 1− for a slope without pore pressures C = γH2 2 sin (α − β) tan β (7.8) 2 γw tan φ 1 2 cos β sin β 1− 1− C = γH 2 sin (α − β) γ tan β for the phreatic surface at ground level
β
C W φ H
l
R
P α
U
Figure 7.5 Wedge analysis to determine the retaining force P.
(7.9)
EARTH PRESSURE AND RETAINING WALL DESIGN
143
In deriving the value of U it is assumed that the making of the cut does not affect the seepage condition, that is, seepage continues toward the cut with seepage lines still parallel to the ground surface. This is a conservative assumption, as in practice some drawdown of the phreatic surface is likely to occur, at least near the face of the cut. In each of the above expressions the term 1/2γ H 2 appears, so that it is easy to directly calculate the active pressure coefficient Ka rather than the force Pa . Making use of the relationships in Eqs. 7.3 and 7.4, the cohesive force C is expressed in terms of φ , and its value is thus dependent on the value chosen for φ . There are an infinite number of combinations of c and φ , as well as groundwater conditions; hence, some restrictions need to be placed on the cases to be considered in order to obtain solutions. It is easier to make a reasonable assumption about the value of φ than c so we will start by adopting a value of φ . We could adopt a fixed value, such as 35o , which would be reasonable for many residual soils on steep slopes. Alternatively, we could assume that the shear strength parameters increase as slopes become steeper and relate c and φ to the slope angle β. For simplicity, we will start with an arbitrary assumption, namely that the value of φ is somewhat less than the slope angle, and related to it by the following relationship: tan φ = 0.7 tan β The cohesion component c then takes on whatever value is needed to maintain stability, in accordance with Eqs. 7.3 and 7.4. We also need to make assumptions about the water table; two cases will be considered, the first assuming no water table (zero pore pressures), and the second assuming the water table is at the ground surface. We can now go ahead using a series of trial wedges and determine values of the pressure coefficient Ka . By plotting Ka against the wedge angle we can expect to determine the critical wedge angle and the corresponding maximum value of Ka . In this case it is more convenient to use the difference in inclination between the wedge angle and the slope angle, that is α − β. The graphs thus obtained are shown in Figure 7.6. These graphs are surprising, at least at first sight, for several reasons. First, the graphs do not have the form normally obtained by a Coulomb analysis. There is no peak value of Ka . Instead, Ka increases linearly as the wedge angle approaches the slope angle (i.e., as α – β approaches zero). The peak value is obtained by extrapolating the graph to intersect the vertical axis. The critical wedge thus becomes a “slab” of soil of thickness H extending an infinite distance up the slope, with its failure plane running parallel to the slope surface. Second, the value of Ka decreases as the slope angle increases. The third surprising result is that the force P turns out to be identical, whether or not the water table is present. The values of Ka are plotted against the slope angle in Figure 7.7.
144
BEARING CAPACITY AND EARTH PRESSURES
1.0
b = 10° b = 20°
Active pressure coefficient Ka
0.8
b = 30°
0.6
b = 40°
0.4 b = 50°
0.2
b = 60° b = 70°
0
1
2
5
10
15
Angle (a − b ) (degrees)
Figure 7.6 Results of the wedge analysis.
Active pressure coefficient Ka
1.0 Assumed: fÄ = 35° (constant), with or without seepage
0.8
0.6 Sand, no seepage 0.4 Assumed: tan fÄ = 0.7 tanb, with or without seepage
0.2
0
20
40 Slope angle b
60
80
Figure 7.7 Values of the active pressure coefficient Ka plotted against slope angle.
EARTH PRESSURE AND RETAINING WALL DESIGN
145
In addition to the curve obtained from the above analysis, two further graphs are shown in Figure 7.7. One is for clean, dry sand, and the other is for an assumed constant value of φ = 35o , which is considered to be typical of many residual soils found in steep terrain. The curve for clean, dry sand is an analytical solution, which the author was not aware of at the time he first investigated this problem. The analytical solution can be found in Jumikis (1962). For the limiting case when β = φ, the solution gives Ka = cos2 φ
(7.10)
αc = φ (= β)
(7.11)
where αc is the critical value of the wedge angle α. This analytical solution thus confirms the trends found with the “manual” wedge analysis used here for a material having both c and φ values as well as seepage in the slope. These apparent anomalies can be explained as follows. It is often intuitively assumed that the steeper the slope behind a retaining wall, the greater will be the force acting on the wall. Similarly, the steeper the slope in which a cut is to be made then the greater will be the force needed to retain it. This is, however, a logical assumption only if the material is the same in each situation. It is not a logical assumption in relation to natural soil slopes. The reason slopes in some natural soils are steeper than others is primarily because the material they consist of is stronger. Slopes in sedimentary clays in the United Kingdom tend to be rather flat, while those in volcanic ash soil in Indonesia are generally very steep, reflecting in each case the relative strength of the soils. If then we are considering the relative magnitude of forces needed to retain cuts in London clay and volcanic ash, for example, there is no a priori reason to assume that a higher force will be needed for the volcanic ash soil. The results summarized in Figure 7.7 are therefore neither surprising nor illogical. In fact, they are the opposite; the following simple logic demonstrates that they must have the form they have. If the slope being retained is flat and is at limiting equilibrium then clearly the material has the properties of a liquid, and the horizontal stress will equal the vertical stress, that is, Ka = 1. On the other hand, if the slope is stable at 90o , then no force is required to retain it, that is, Ka = 0. Hence, the Ka value must start at unity for a level slope and decrease to zero for a vertical slope. The fact that Ka is the same regardless of whether seepage is present is also not illogical. The back-analysis with seepage present produces a higher value of c than produced without seepage, which counterbalances the influence of pore pressure in the wedge analysis. There are a number of idealizing assumptions involved in the above analysis, which may not apply in practice: 1. The slope is at the point of failure. If this is not the case then the soil strength will be higher than the back-analysis produces, and the Ka values will be less.
146
BEARING CAPACITY AND EARTH PRESSURES
2. The slope is infinite. If this is not the case, then both the back-analysis and the wedge analysis would be affected, presumably by comparable amounts and the net result not much different to that from the infinite slope assumption. 3. The seepage pattern remains the same after the cut is made. In reality, there will probably be drawdown to some extent near the face of the cut. This will have the effect of lowering the force level required to retain the slope and hence the Ka values in Figure 7.7 for the cases involving seepage will be conservative. 4. The critical failure surface is planar. Nonplanar failure surfaces would result if the basic assumptions were altered and this would influence the values of Ka somewhat. 5. No frictional forces between soil and retaining wall have been included in the analysis, although it would not be difficult to repeat the analysis taking them into account. In many cases, cut slopes are retained using anchor systems without a rigid facing on the slope, and in this case the issue of wall friction would not arise. Finally, we can note that this approach could be used for nonvertical walls, and would be expected to result in somewhat lower values of Ka than those applicable to vertical walls. 7.3.2
The Use of Residual Soils for Reinforced Earth Construction
Reinforced earth (RE) walls appear to have become so commonplace in recent years that old-style cantilever reinforced concrete walls are almost a thing of the past. The advent of new reinforcing materials, especially geogrids, has further stimulated the growth of reinforced earth by making possible the use of clay as the fill material. The cost reductions associated with the use of clay are significant, especially in countries of the wet tropics, where the intensity of weathering means that residual clays are often readily available, while supplies of clean granular material are available only at high cost. The residual clays generally have good engineering properties, namely high strength, low compressibility, and low shrink/swell characteristics. There has been resistance to the use of clays in reinforced earth walls from various quarters, and some codes place restrictions on their use as fill material. The reasons for this resistance are unclear; they appear to come more from force of habit and “mental inertia” than logic and reason. Clay of all types is routinely used in important engineering structures, especially earth dams, which are much more critical structures than reinforced earth walls.
EARTH PRESSURE AND RETAINING WALL DESIGN
147
The relative advantages and disadvantages of granular versus clay fill can be summed up as follows: Granular Fill • • • • •
It is easy to handle and compact in all weathers. It has good frictional properties, and so provides a good bond between soil and reinforcement. It has generally high strength and low compressibility, so both wall dimensions and deformations can be minimized. It is free-draining, so it acts as a drain to intercept any seepage coming from the retained soil. It is relatively expensive.
Clay Fill • • • • • • •
It is often readily available and much cheaper than granular fill. Clay requires much more rigorous control of compaction than granular fill. Its frictional characteristics are not as good as granular fill, so greater bond area will be needed to provide anchorage. Clay is less rigid than sand or gravel, and therefore wall deformations will be greater. It can be handled and compacted only in dry weather. Clay is not permeable and drainage blankets are needed if there is likely to be any seepage coming from the soil retained by the wall. The actual stress state in the compacted clay is likely to be uncertain, so design assumptions may not apply in practice.
None of the arguments against the use of clay are insurmountable, and some are of little consequence, as the following comments indicate. 1. The advent of geogrids means that the contact area between reinforcement and soil is increased enormously compared to that with metal strips, and the need for high frictional characteristics in the fill material is much reduced. It is still desirable for the fill material to have good frictional characteristics, as this reduces the forces to be taken by the reinforcement. In this respect, residual soils have a significant advantage over sedimentary soils. The friction angle φ of many residual soils is between 30o and 40o , and so is not greatly different from the values found in granular materials.
148
BEARING CAPACITY AND EARTH PRESSURES
2. Although compacted clay is less rigid than granular fill, the likely deformation is still not large and for many situations is of no consequence. In situations where it is of consequence, allowances can normally be made for it during design and construction. 3. It is true that clay can be handled easily and compacted only in dry weather, and in some situations this could rule out its use. However, embankments for highways and building platforms, not to mention earth dams, all involve the use of clay, so this argument is no stronger for reinforced earth walls than other structures. 4. Drainage blankets are regularly installed as part of any retaining wall construction, so this is not an exceptional requirement. 5. It is true that the actual stress state in the compacted clay fill is unlikely to be known with certainty, but unless the wall has been poorly designed and constructed, the actual stress state is likely to be more favorable than assumed in design, as will now be explained. Reinforced earth walls are normally designed using an effective stress analysis. The pore pressures in the fill material are assumed to be zero. This is perfectly reasonable when the fill is a free-draining granular material, but when it is compacted clay the situation is quite different. There is little doubt that the permeability of undisturbed residual clays is normally much higher than that of sedimentary clays, but this is much less likely to be true when they have been excavated and recompacted. The compaction process destroys the structure that gives them high permeability, and is likely to lower their permeability by an order of magnitude or more, as for example is the case with volcanic ash clays (Figure 9.8 of Chapter 9). This means that the compacted clay fill is likely to behave in an undrained manner during the construction period, and possibly for a considerable time longer. The pore pressure state in clay at the time of compaction is likely to be in the range of 200–400 kPa, assuming the clay has been compacted in a controlled manner with a water content close to its standard Proctor optimum value. The pore pressure will steadily rise as fill layers are added, but will not become positive until the confining stress exceeds the initial negative pore pressure. The pore pressure parameter B is less than unity for a compacted clay. This means that pore pressure in the fill is unlikely to become positive unless the wall height exceeds about 10–15 m. Hence, for walls of moderate height the pore pressure at the end of construction is normally expected to be negative. This adds to the margin of safety in the wall as it increases the effective stress on any potential failure plane. The long-term pore pressure state in the wall will depend on the topography and seepage state of the environment where it is built. Figure 7.8 shows two possibilities. One is a level site with twin walls retaining a highway embankment, and the other a hill slope retained by a wall. For the wall built on level ground with a sealed surface, the long-term pore pressures will move toward the equilibrium state shown, which is
EARTH PRESSURE AND RETAINING WALL DESIGN
Reinforced earth
149
Reinforced earth Pore pressure
Road pavement
Negative Retained fill, pore pressure normally negative
Ground surface
Equilibrium pore pressure
Possible vertical seepage implies zero pore pressure Water table (a) Level site
Drainage layer to intercept seepage from retained slope, sealed at top to prevent entry of surface run-off Reinforced earth
e
fac
nd rou
sur
G
tic rea
face
sur
Ph
pe hill slo e from e wall g a p th See ed by retain
Pipe outlet (b) Wall retaining a hill slope
Figure 7.8 Seepage and pore pressures in reinforced earth walls built of clay.
governed primarily by the water table depth. The sealed surface of the road pavement minimizes the influence of dry or wet weather, but is unlikely to totally prevent passage of water to the soil below. Thus, the pore pressure may increase somewhat from the initial negative value at times, but can be expected to remain negative for the long term. The situation with the wall retaining the hill slope is different. Seepage from the hillside and the need for an interceptor drain means that there is no longer a direct controlling influence from the water table depth. The equilibrium pore pressure state will still tend to be negative and governed by the height of the wall. However, without the advantage of a sealed pavement at the ground surface, and possible contact with water in the drainage layer, the pore pressures may at times become zero or close to zero. This is the situation for which the wall has presumably been designed so is not of any great consequence. Such a wall does not have the luxury of an extra margin of safety as is the case of the wall on a level site.
150
BEARING CAPACITY AND EARTH PRESSURES
The above comments mean that compaction of the clay must be done with rigorous quality control and the drainage and pore pressure situation properly evaluated and controlled by appropriate measures, especially when the wall is supporting a natural slope from which seepage is coming. The drainage cutoff layer at the back of the wall illustrated in Figure 7.8b must be carefully constructed to ensure it is continuous and of clean free-draining material, and is sealed at the top to prevent entry of surface runoff. The author is aware of a number of reinforced earth walls built of clay in New Zealand and Indonesia, and has observed the construction of one wall 8 m in height built within 100 m of the Engineering School at Auckland University. Brief details of some of these walls are as follows: New Zealand Location Lucas Creek, Albany, Auckland Grafton Gully, Auckland State Highway 3 Ngaere Overbridge
Height (m)
Soil Type
13
Medium to high plasticity clay (weathered sandstone/mudstone) Medium to high plasticity clay (weathered sandstone/mudstone) Taranaki brown ash (presumably allophonic) clay
8 7.5
Indonesia Location
Height (m)
Soil Type
Bintaro Viaduct, Jakarta
7.8
Outer Ring Road,Ceger-Hankam Raya, Jakarta
7.3
Tropical red clay, weathered from volcanic material—andesitic lahar and ash. Tropical red clay, weathered from volcanic material—andesitic lahar and ash.
REFERENCES Jumikis, A. R. 1962. Soil Mechanics. Princeton, NJ: Van Nostrand, 570–585. Prusza, Z., De. Kleiner, and A. V. Sundaram. 1983. Design, construction and performance of large dams on residual soils. Proc. 7th Pan Am. Conf. Soil Mechanics and Foundation Engineering. Vancouver, Vol. 1, 185–198. Taylor, D. W. 1948. Fundamentals of Soil Mechanics. New York: Wiley, 429–431. Wesley, L. D. 2001. Coulomb wedge analysis of cuts in steep slopes. Canadian Geotechnical Journal 38(6): 1354–1359.
CHAPTER 8
SLOPE STABILITY AND SLOPE ENGINEERING
8.1
INTRODUCTION
While the general principles of slope stability that apply to sedimentary soils are equally applicable to residual soils, there are various aspects of slope behavior that are peculiar to, or characteristic of, residual soils. These include the following: 1. Slopes in residual soils (excluding black cotton soils, and soils derived from soft sedimentary rocks such as shale) generally remain stable at much steeper angles than those in most sedimentary soils. Slopes of 45o or steeper are not uncommon. Cuts in volcanic ash (allophane) clays can often be made as steep as 60o and at least 10 m high, without danger of slipping. 2. Slope failures in residual soils, especially when steep slopes are involved, are unlikely to be deep-seated circular failures. They are more likely to be relatively shallow, often with slightly curved, or almost planar, failure surfaces. However, the volume of material involved may still be very large. 3. The value of the cohesion intercept (c ) usually plays a significant role in maintaining stability; it appears to be due to some form of weak bonds between particles. 4. The residual strength is generally closer to the peak strength than is the case with most sedimentary soils, especially in clays containing allophane or halloysite. 151
152
SLOPE STABILITY AND SLOPE ENGINEERING
5. The contribution to shear strength arising from the zone of negative pore pressure above the water table may be a significant factor in the stability of some slopes in residual soils. 6. With some residual soils (possibly the majority), the presence of discontinuities may be the dominant factor governing the stability of slopes. 7. The extent to which the stability of slopes in residual soils can be evaluated by analytical methods is often very limited, because of uncertainties in the soil strength parameters and in the seepage conditions. 8. Slips and landslides in residual soils are generally triggered by heavy rainfall, and are thus the result of temporary increases in the pore water pressure in the slope. Severe storm events are much more likely to trigger slope failures than normal seasonal effects. 9. Strong earthquakes may also be the trigger for slips or landslides 10. The actual cause, as distinct from the trigger, of a great many landslides in residual soils is in fact human activity. Excavation into slopes, the placing of fill on slopes, the interference with natural drainage and seepage patterns, and deforestation are all factors that generally reduce stability and possibly lead to failures, especially in urban areas.
8.2
FAILURE MODES
As mentioned above, slope failures in residual soils, especially when steep slopes are involved, are unlikely to be deep-seated circular failures. They are more likely to be relatively shallow, with fairly planar failure surfaces. In large slopes with a limited depth of weathered material overlying sound rock, they are likely to be predominantly translational slides. Also, it is not uncommon in volcanic areas for volcanic material to slide at the interface between volcanic deposits and the underlying sedimentary soils. The slip surface in this case may be fairly linear so that the slide is essentially a translational slide. However, the volume of material involved may still be very large. Some modes of failure are illustrated in Figure 8.1. 8.3 THE PLACE OF ANALYTICAL AND NONANALYTICAL METHODS FOR ASSESSING THE STABILITY OF NATURAL SLOPES
It should not be imagined that assessing the stability of natural slopes is essentially an analytical exercise. Hopefully, it will be clear from what will be said later in this chapter that there are severe limitations on the extent to which analytical methods can be applied to natural slopes. They may or
THE PLACE OF ANALYTICAL AND NONANALYTICAL METHODS
Shallow circular slide (very common)
er table
Peak wat
153
Block slide (fairly common)
ater table
Normal w
Har
yer
Hard la
Large translational slide (common)
d la
Deep seated circular slide (very unlikely)
yer
Figure 8.1 Failure modes in residual soils.
may not be an important part of slope stability assessment, depending on the nature of the slope, in particular its geology, topography, soil conditions, and past history. Other, nonanalytical methods, however, are always an essential part of any assessment of slope stability. These methods may appear primitive and not technically satisfying to geotechnical engineers, but that does not lessen their importance. They include the following: 1. 2. 3. 4.
Visual inspection of the slope Geological appraisal of the slope and surrounding area Inspection of aerial photos Inspection of existing slopes in similar materials to the slope in question
Careful visual inspection of slopes, along with geological knowledge can give a good guide as to whether a particular slope is stable or not. Slopes with smooth contours (Figure 8.2) indicate that they have been formed by surface erosion processes, without slip movement. On the other hand, irregular surfaces suggest that some form of slip movement may have been involved. Inspection of aerial photographs can often show features of a site that are not evident from a direct visual inspection. They can show scarp lines or changes of vegetation indicating boundaries of old slip movement. Inspection of any existing cuts in the area of interest can tell us two things— how
154
SLOPE STABILITY AND SLOPE ENGINEERING
Irregular contours suggest instability
Smooth contours indicate stability
Possible slip or slump movement Shape is formed by steady surface erosion
?
? Shape appears to be formed by mass movement
?
Figure 8.2 Stability indications from surface features of slopes.
the cut slope itself is performing, and what sort of material it is made of. It is probably true that most assessments of the stability of a natural slope are based 80 percent or more on (1) to (4) in the previous list and less than 20 percent on the results of analytical methods. 8.4
APPLICATION AND LIMITATIONS OF ANALYTICAL METHODS
The limitations of applying conventional analytical methods to residual soil slopes arise from uncertainties with respect to both soil properties and seepage conditions in the slope. Soil conditions include the basic geology, in particular the stratigraphy and structure of the site, and the shear strength parameters for each distinct material. Seepage conditions govern the pore pressure in the slope, on which the stability analysis depends just as much on soil properties. These factors are considered in turn in the following sections. 8.5
UNCERTAINTIES IN MATERIAL PROPERTIES
In considering the use and limitations of analytical methods it is convenient to divide slopes into three categories: 1. Slopes consisting of uniform, homogeneous materials 2. Slopes containing distinct continuous planes of weakness 3. Slopes of heterogeneous material, but without distinct planes of weakness, as, for example, in a weathering profile of the Little kind (Figure 2.1a of Chapter 2). 8.5.1
Slopes Consisting of Uniform, Homogeneous Materials
With such slopes, the determination of accurate safety factors by conventional slip circle analysis would appear to be a reasonable expectation. However, there are still uncertainties that cannot easily be eliminated. These
UNCERTAINTIES IN MATERIAL PROPERTIES
155
uncertainties relate first to the shear strength of the soil and second to the seepage and pore pressure state in the ground. With respect to shear strength, which has already been covered extensively in Chapter 5, the following points are reiterated: • •
•
The value of φ can usually be determined with reasonable accuracy using normal measurement methods, such as triaxial testing. The value of c , although it may not be very large, is often significant (due to some form of weak bonds between particles), but it cannot easily be determined with the same degree of reliability as φ (see Section 10.3 and Figure 10.3.2 of Chapter 10). Careful triaxial testing at low confining stresses is needed to accurately determine c . The residual strength is likely to be fairly close to the peak strength, especially in clays continuing allophane or halloysite.
With regard to the seepage pattern and pore pressure state in the slope, the relatively high permeability of most slopes in residual soils means that the seepage state is likely to be continuously changing, depending on the weather conditions. The worst-case seepage pattern is clearly the one that governs the long-term stability of the slope. Unfortunately, because our knowledge of soil conditions will always be partial, this worst case will never be known with certainty, although we will see shortly that there are some methods that we can adopt to try to estimate this worst case. 8.5.2
Slopes Containing Distinct, Continuous Planes of Weakness
The behavior of many slopes in residual soils is dominated by the presence of random discontinuities in the form of distinct planes of weakness, as illustrated in Figure 8.3a. This is likely to be the case with soils that have been subject to tectonic deformations and shearing, or derived from rocks subject to such deformation. The presence of these discontinuities makes the determination of the likely failure mode, and the values of the soil strength parameters, extremely difficult, and thus reduces the likelihood that
Planes of weakness
(a) Random discontinuities —indeterminate influence on stability
(b) Regular discontinuities —quantifiable influence on stability
Figure 8.3 Possible discontinuity patterns and influence on slope stability.
156
SLOPE STABILITY AND SLOPE ENGINEERING
analytical methods will produce reliable results. Only in rare situations is it likely to be possible to determine the location, orientation, and strength of discontinuities with the degree of reliability needed for the use of analytical methods. The exception to this observation is the situation when the fissures or bedding planes are generally orientated in a particular direction. Some residual soils derived from sedimentary soils may contain planes of weakness that reflect particular weak layers in the parent material, as indicated in Figure 8.3b. In this case it may be possible to determine the shear strength parameters within these weak layers and make use of them in sensible stability analysis. 8.5.3 Slopes of Heterogeneous Material, but without Distinct Planes of Weakness
The weathering of igneous rocks, such as granite, does not generally create distinct planes of weakness, so that this is quite a different situation to that just described above. The soil profile will consist of zones of partly weathered material containing remnants of the parent rock, and zones of fully weathered material (soil). Determination of the strength parameters applicable to the material as a whole is still very difficult, if not impossible, by conventional sampling and laboratory testing. This may not entirely rule out the use of analytical methods, as it may still be possible to determine the strength parameters from back-analysis methods applied to existing slips or slopes. Some examples of these methods are given in a later section. 8.6 8.6.1
UNCERTAINTIES IN THE SEEPAGE AND PORE PRESSURE STATE Influence of Climate and Weather
As already observed, slips and landslides in residual soils triggered by periods of prolonged or intense rainfall are the result of temporary increases in the pore water pressure in the slope. Any measurement of pore water pressure in the slope is valid only at the time it is made and cannot be assumed to be relevant to long-term stability estimates. For such estimates, it is the worst seepage condition likely to occur in the future that will determine the long-term stability of the slope. As pointed out in the introduction to this chapter, we should be careful when evaluating the causes of slope failures in residual soils to recognize that while rainfall is the trigger that initiates failures, the fundamental cause is often that the slope in question has been significantly affected by human activity. It is virtually impossible to build cities, highways, dams, or even pipelines, without undertaking earthworks that involve cuts and fills. It is this reshaping that frequently reduces the margin of safety of slopes and is the real cause of the subsequent instability.
UNCERTAINTIES IN THE SEEPAGE AND PORE PRESSURE STATE
157
One important reason (which should be clearly recognized) that a significant number of slopes in residual soils remain stable at steep angles is because the phreatic surface (water table) is often deep, and the pore pressure above the surface is negative (suction or pore water tension) as described in Chapter 3. This zone of pore water tension, which may include most of the slope, increases the effective normal stress across any potential failure surface, thus increasing the shear strength and the safety factor of the slope. The influence of intense rainfall on this zone is to increase the pore pressure from its negative value toward zero (i.e., to reduce or destroy the suction above the water table), or possibly turn it into a positive value if the phreatic surface rises. However, it is not necessary for the phreatic surface to rise at all for rainfall to induce failure in a slope. The reduction in the negative pore pressure without change in the water table may induce failure in the slope. An example of such a situation is given later in Section 8.9. We should appreciate that the existence of high negative pore pressures above the water table and their contribution to the stability of the slope do not necessarily imply that the soil is not fully saturated. As already noted, clays become partially saturated only as a result of evaporation at the ground surface, not because water drains out of them under the influence of gravity. Coarse residual soils, such as the weathered granites found in Hong Kong, behave differently; they appear to be sufficiently coarse that water drains out of them under gravity, and during dry periods the zone above the water table is likely to become less than fully saturated. 8.6.2
Response of Seepage State and Pore Pressure to Rainfall
The influence of rainfall on the water table and the pore pressure state in a slope arises from two different weather effects: 1. Regular seasonal influence. This is cyclical in nature, and for many climates is reasonably predictable, as indicated in Figure 3.7 of Chapter 3. 2. Isolated storm events. These are generally unpredictable, both in timing and intensity, and are more likely to be the direct trigger of landslides than normal seasonal changes. The mechanics by which the pore pressure and water table changes take place vary between two distinct types, as described in Chapter 3: 1. Coarse-grained materials of relatively high permeability, especially coarse sands and gravels. In this case water flows into and out of the void space and displaces air. The degree of saturation changes by variable amounts, depending on the particle size involved. With coarse sands and gravels, the change is likely to be from virtually zero saturation to full saturation. Volume change of the soil itself is considered
158
SLOPE STABILITY AND SLOPE ENGINEERING
to be negligible and is not part of this mechanism. The parameters involved in this mechanism are the coefficient of permeability, the porosity, and degree of saturation of the soil. 2. Fine-grained, low-permeability soils, especially clays. Except for a shallow zone near the surface where evaporation effects are important, this mechanism will commonly involve fully saturated soil, and the same mechanics apply as in the familiar consolidation (or swelling) process in clays. Changes in pore pressure are accompanied by changes in volume of the soil. The parameters involved are the coefficients of permeability and compressibility of the soil, or in their combined form, the coefficient of consolidation. An example of the use of Terzaghi consolidation theory to estimate changes in pore pressure caused by rainfall is given in Section 3.2.6 of Chapter 3. A reasonable representation of the situation with low-permeability soils, especially for slopes in clays of moderate to high plasticity, is illustrated in Figure 8.4. This figure is, in principle, a version of Figures 3.4 and 3.7 of Chapter 3, with amendments to include the influence of intense storm events. The depth of penetration of seasonal effects will normally be greater than that of the storm events, as shown. This figure shows that antecedent rainfall can be expected to have some influence on the situation during an intense rainfall. An intense rainfall storm that occurs when the seasonal water table is at its highest level is more likely to trigger a landslide than one occurring when the water table is at its lowest level.
Pore pressure
Seasonal and storm influence Seasonal influence nce only
al influe
Season
Expected trend with increasing depth
Storm influence
No climatic influence
Dry season
Wet season Time (one year)
Figure 8.4 Summary of pore pressure response to climatic effects in clay slopes.
THE WORST-CASE ASSUMPTION REGARDING THE WATER TABLE
8.6.3
159
Comparison with Sedimentary Soils
It will already be apparent that the long-term stability of cut slopes in residual clays is much more problematic than is the case with sedimentary clays. With sedimentary clays, the permeability is normally very low and it is likely to be a number of years, or even decades, after the cutting is made before the seepage pattern and pore pressures reach a steady state. Once this steady state is reached, it is unlikely to be significantly affected by seasonal weather effects or storm events and the long-term stability can be estimated from this steady-state situation. In contrast, the permeability of residual soils is relatively high, and the seepage pattern and pore pressures are likely to be subject to seasonal changes. The most probable situations for cut slopes in the two soil types are illustrated in Figure 8.5. This shows graphs of pore pressure and effective stress versus time for a typical point P on a potential slip surface, and the safety factor with respect to failure on this surface. With sedimentary slopes it is normally assumed that behavior is undrained during excavation of the cutting, after which the pore pressure trends toward its long-term steady-state value. This represents the worst-case situation for stability analysis. With residual soils, it is likely that behavior will not be strictly undrained during excavation, and also the long-term pore pressure will be subject to seasonal and storm influences. The worst-case pore pressure state will thus be that arising from the most adverse storm event. There is no way to be sure what this worst condition will be, but theoretical approaches can provide us with some guidance, as described in the following sections. 8.7 THE WORST-CASE ASSUMPTION REGARDING THE WATER TABLE
One way to estimate the worst-case pore pressure condition is to assume the water table rises to the ground surface during the most severe storms the slope is likely to encounter. This approach is sometimes used in practice and does not appear to be unreasonable, although it runs the risk of being severely overconservative. This risk arises from two factors, one of which can be largely avoided, as we shall see shortly: 1. The assumed worst case is significantly worse than the slope will ever actually experience. 2. The use of conventional slope stability computer programs involves an assumption about the relationship between pore pressure and water table that is unrealistic and can give rise to serious errors in the estimate of the safety factor. The second factor arises because almost all conventional slope stability programs (or even manual methods) assume that the pore pressure can
Potential failure surface
Long term steady state —typical of low permeability (sedimentary) clays
P Fluctuating water table —typical of medium to high permeability (residual) clays
Pore pressure
Storm events
Seasonal influence
Time Sedimentary clays
Long term
End of construction
Time
Safety factor
Effective stress
Residual clays
Time
Figure 8.5 Conceptual illustration of the short- and long-term stability of a cut slope in sedimentary and residual soil.
160
THE WORST-CASE ASSUMPTION REGARDING THE WATER TABLE
161
be calculated with sufficient accuracy by adopting a hydrostatic relationship with respect to the phreatic surface. In other words, the pore pressure is calculated using the vertical distance from the phreatic surface to the point concerned. We will refer to this as the vertical intercept assumption. This implies a seepage state in which the flow lines are horizontal and the equipotentials are vertical. This is probably a reasonable assumption with most sedimentary soils, where slopes and phreatic surfaces tend to have gentle inclinations. With residual soils, however, this assumption needs some rethinking, or at least reanalyzing. Slopes in residual soils are generally considerably steeper than those in sedimentary soils, as illustrated in Figure 8.6. We will proceed to investigate this issue by estimating the stability of a range of idealized slopes of varying inclinations, assuming two different pore pressure states, both of which are based on the worst-case assumption that the water table is at the ground surface. 1. Case A: The vertical intercept assumption is made, in line with most computer programs. This implies a seepage state with vertical equipotentials and horizontal flow lines. 2. Case B: The most probable flow net compatible with the water table at the surface is determined, and the stability analysis repeated using pore pressures from this flow net.
r table
Wate
Critical circle (a) Sedimentary soil able
ater t
ase w
tc Wors
Critical circle
le
Average water tab
(b) Residual soil
Figure 8.6 Typical slopes in sedimentary and residual soils.
162
SLOPE STABILITY AND SLOPE ENGINEERING
The only way the water table can exist at the ground surface is for rain to be continuously falling on the surface. For case B we will therefore determine the flow net assuming continuous rainfall equal to, or in excess of, the maximum infiltration the soil can take. We will adopt a slope height of 20 m, and a range of inclinations. Full details of the slopes being analyzed are as follows: Height: 20 m Slope inclination: from 0.25 (H):1 (V) to 2.5:1 Unit weight: 16 kN/m3 Shear strength parameters, as in Table 8.1. The shear strength parameters have been selected to give an average safety factor of unity (or close to unity) for each inclination from the two pore pressure cases analyzed. This means varying the strength parameters from large to small as the slope angle is decreased, as the table shows. The parameters are also believed to be representative of residual soils with slopes of these inclinations. The results of the analysis are shown in Figure 8.7. Figures 8.7a and 8.7b show typical results for one of the slopes analyzed, namely the 1:1 slope. In Figure 8.7a, a flow net has been created using the SEEP/W program and then used in SLOPE/W to calculate the safety factor. The cross section actually used in the seepage study extended in the horizontal distance well beyond the boundaries shown in Figure 8.7a to minimize edge effects. Figure 8.7b shows the seepage situation implied in virtually all computer programs, namely that equipotential lines are vertical and horizontal flow lines are horizontal. This gives an ru value of 9.8/16.0 = 0.61. It is seen that the position of the critical circles determined by the slip circle analysis is not very different but there is a large difference in the safety factor. The value using the flow net is 50 percent higher than the value obtained using the vertical intercept assumption (hydrostatic pore pressure). Figure 8.7c shows the safety factors obtained from all six slopes analyzed. As expected, the difference is greatest with the steepest slope and gradually decreases to become very small at inclinations of 2.5:1 or less. The phreatic surface becomes flatter and the equipotentials move close to vertical. The values of ru equivalent to the pore pressure state from the flow nets have also been calculated and are shown in Figure 8.7d. It is seen that for steep Table 8.1 Shear Strength Parameters Used in the Analysis Slope inclination c (kPa) φ (degrees)
0.25:1 70 45
0.5:1 50 45
1:1 30 40
1.5:1 16 35
2:1 15 33
2.5:1 13 30
163
THE WORST-CASE ASSUMPTION REGARDING THE WATER TABLE SF = 0.81
SF = 1.22
1:1
1:1
20 m 30 m
Unit weight = 16 kN/m3 cÄ = 30 kPa fÄ = 40° (a) Flow net with continuous rainfall on surface
(b) Flow net assumed in computer analysis when input is phreatic surface at ground surface ru = 0.61 - Water table at surface, vertical equipotentials 0.7
1.6
Based on flow net during continuous rainfall
1.2 1.0 0.8 Based on ru = 0.61 - water table at surface, vertical equipotentials
0.6
0.6 Pore pressure ratio (ru)
Safety factor
1.4
0.5 0.4 0.3 Equivalent values of ru to give safety factors based on flow net
0.2 0.1
0.4 0.25:1
0.5:1
1:1
1.5:1
2:1
Slope inclination
2.5:1
0 0.25:1
0.5:1
1:1
1.5:1
2:1
2.5:1
Slope inclination (c) Safety factors versus slope angle
(d) Values of ru equivalent to the flow net from continuous rainfall on the ground surface
Figure 8.7 Influence of seepage assumptions on the safety factor of slopes.
slopes the value of ru is only a small fraction of the value that normal computer programs would assume in this situation. It needs to be emphasized that the flow net implied in conventional computer programs is a physical impossibility in slopes with steep phreatic surfaces. To put it more strongly, it is a gross travesty of the true situation, and can in no way be claimed to be even a reasonable, albeit conservative, approximation. The flow nets used in this study, such as that in Figure 8.7a, thus represent the most probable worst-case seepage pattern in a slope during continuous rainfall on the slope and surrounding area. The conclusion from this study is that while it may be reasonable to assume the worst-case pore pressure situation is that with the water table at the ground surface, the use of normal slope stability programs in this situation is likely to result in grossly conservative estimates of safety factor for steep slopes. This is because the vertical intercept relationship between pore pressure and water table no longer applies. For the 0.5:1 slope, the
164
SLOPE STABILITY AND SLOPE ENGINEERING
safety factor from a “normal” computer analysis is about 0.53, while that from a realistic seepage pattern is about 1.48. The above analysis has demonstrated a theoretical method by which a reasonable estimate of a worst-case pore pressure state can be estimated. This leaves open the question of the probability of this state actually occurring. In the next section a method is presented that helps answer this question. It involves a transient seepage analysis of the way in which the water table and pore pressure change in a slope during continuous rainfall.
8.8 TRANSIENT ANALYSIS OF RAINFALL INFLUENCE ON THE STABILITY OF A HOMOGENEOUS CLAY SLOPE
An example of a clay that generally belongs in category 8.5.1 above (homogeneous soil) is the tropical red clay found widely in the island of Java in Indonesia. It is not completely homogeneous, but the variations in its properties are sufficiently small that for practical engineering purposes it can often be considered to be homogeneous. The author has previously described and analyzed a riverbank slope in this clay (Wesley 1977). The stability analysis was limited to examining the slope with the relatively deep water table that was present at the time of the investigation. No attempt was made to establish the most probable seepage pattern or the worst case. Our present purpose is to reanalyze the slope in greater detail, taking account of changing pore pressures resulting from rainfall, and at the same time illustrate that theoretical transient analysis in uniform slopes can produce sensible and informative results. Figure 8.8 shows a series of cross sections along the river bank that were actually measured, together with the idealized section used in the analysis. This idealized section is 10 m high with an inclination angle of 70 degrees. The computer program SEEP/W is used here to carry out the transient seepage analysis. This is based on the conventional transient form of the continuity equation (Lam et al. 1987) expressed as follows: k
∂ 2h ∂ 2h + ∂x 2 ∂y 2
+ Q = mw γ w
∂h ∂t
(8.1)
where Q is the rate of flow into a soil element from an external source, and mw is the slope of the volumetric water content with change in pore pressure u. The volumetric water content (θ) is the volume of water per unit volume of soil. It is directly related to the water content as normally defined in soil mechanics. Hence, mw =
∂θ ∂u
TRANSIENT ANALYSIS OF RAINFALL INFLUENCE
165
Red clay
70° 10 m
Mottled red and grey clay Reddish grey clay
(a) Measured river bank cross sections Initial state SF = 2.14
Final state SF = 0.81
Long term steady state flow net
a Idealized cross section
10 m 70°
15 m
l water table
Assumed initia
b (b) Initial and final pore pressure conditions
Figure 8.8 Transient analysis of a river bank slope in tropical red clay.
For fully saturated soils it is easily shown that mw = mv , the coefficient of compressibility of the soil. For the case we are studying here the term Q disappears as the rate of flow into the soil elements is determined by the hydraulic conductivity of the soil and the hydraulic gradient at the soil surface, and does not have a predetermined value. The above equation then becomes 2 ∂ h ∂ 2h ∂h k + 2 = mw γ w 2 ∂x ∂y ∂t
166
SLOPE STABILITY AND SLOPE ENGINEERING
which with a little manipulation becomes k ∂h = ∂t mv γ w
∂ 2h ∂ 2h + ∂x 2 ∂y 2
(8.2)
Readers will recognize Eq. 8.2 as having a similar form to the well-known Terzaghi consolidation equation. The only difference of substance is its two-dimensional form. The similarity is to be expected, since the soil parameters controlling the mechanics in the two situations are the same, namely the coefficient of permeability, k, and the compressibility coefficient, mv , or their combined form, the coefficient of consolidation, cv . The Terzaghi equation is simply a special case of transient flow. The objective of the analysis is to determine how the pore pressures and the safety factor of the slope change as a result of continuous rainfall on the slope and surrounding ground. The analysis is similar to that in Section 8.7, but includes both transient states and the ultimate steady state. The transient seepage states at a sequence of time intervals obtained from the SEEP/W analysis are transferred to a SLOPE/W analysis to obtain safety factors. The soil properties used are those in the original (1977) analysis, namely ◦
Unit weight = 16.2 kN/m3 , c = 14 kPa, φ = 37 In addition, for the transient analysis, the following parameters were adopted: Coefficient of permeability k = 0.01 m/day, Coefficient of compressibility mv = 0.0001 kPa−1 . The results are shown in Figures 8.9 and 8.10. Figure 8.9 shows graphs of pore pressure on one particular vertical section through the slope, namely section a-b in Figure 8.8, at a series of time steps. The similarity of these curves to Terzaghi consolidation curves is clearly evident. There is a notable difference however, as the final equilibrium situation is not one of hydrostatic equilibrium. It is an equilibrium seepage state, so that the pore pressures are well below the hydrostatic values. This is another illustration of the point made in the last section regarding the error involved in the common vertical intercept assumption method used by computer programs to calculate pore pressures. These contours illustrate an important point about the way the water table rises. It does not rise at a uniform rate; instead, it rises slowly at first and then very rapidly in its final stages. This is because of the shape of the contours. From the start until time step 1.1 it rises from its initial depth of 10 m to 8 m; it then rises from 8 m to the surface between time steps 1.1 and 2.7. Figure 8.10 shows the rise in water table with time as well as the rise in pore pressure at a depth of 15 m (an assumed impermeable boundary). The water table reaches the surface after only 2.7 years, while the pore pressure at 15 m takes about 20 years to reach an equilibrium “steady” state.
TRANSIENT ANALYSIS OF RAINFALL INFLUENCE
167
Pore pressure (kPa) −60 0 0.1
2
−40
−20
0
20
40
60
80
100 Final water table
0.2 0.4 0.6 1.1
r yd H
4
ta os tic te
yd H ro
2.7
tic
8
a st
6.2
20 (final steady state)
e
at
st
Depth (m)
a st
6
10
Initial water table
12
14
Figures on contours are time steps (days)
16
Figure 8.9 Pore pressure changes with time on section a-b of Figure 8.8.
Figure 8.10 also shows the change in safety factor with time. The initial value of safety factor is 2.14, taking into account the negative pore pressure above the water table. It falls to unity in about 3 years and continues to decline to reach its steady-state value of 0.81 in 20 years. If the long-term stability is estimated assuming a worst-case condition with the water table at the surface and using a conventional computer stability program, then the safety factor is only 0.11. The safety factors are summarized in Table 8.2. This example, involving an actual field situation, illustrates a number of important points: 1. The analysis produces a sensible result, as it indicates that three days of continuous heavy rainfall is necessary for the safety factor to fall to unity and initiate failure. The island of Java does have very heavy rainfall, but it is most unlikely to be continuous for three days. It is very unlikely to last more than one day, so the likelihood of the worst-case pore pressure state actually occurring is very low. 2. Adopting a worst-case condition of the water table at the ground surface and carrying out a stability analysis using routine computer programs that incorporate the vertical intercept assumption to estimate pore pressure produces a hopelessly unrealistic result. The
168
SLOPE STABILITY AND SLOPE ENGINEERING
Safety factor
2.2 1.8 1.4 1.0 0.6
0
5
10
15
20
15
20
5
10
15
ble
0
2 4 6
Water ta
Depth of water table (m)
0
13 11 Pressure
head
9
8
7
10
5
Pressure head at base of layer (m)
Time steps (years)
Figure 8.10 Transient changes in water table depth, pore pressure at 15 m, and safety factor.
banks of the stream concerned here had been stable for years at the time of the investigation in 1975 (and as far as the author is aware are still the same). Thus, an analysis that produces a safety factor of 0.11 is clearly nonsensical. 3. The results of the analysis are essentially the same as those in the author’s 1977 paper, in that it shows the slope to have a safety factor of unity when the value of ru is quite low. The 1977 paper states: “the safety factor falls to unity when the ru value rises to just under 0.1.” The current analysis gives the value of ru as 0.07, which is not too different (the author derives some satisfaction from this—the stability analysis in the 1977 paper was done manually using a slide rule). 4. In the 1977 paper the statement is made that “the groundwater level could rise substantially during periods of heavy rainfall to give higher values of ru ,” a statement that reflects the author’s (mistaken) belief at the time that the pore pressure was related directly to the level of the water table (the vertical intercept assumption). 5. The shear strength parameters c and φ used in this study are believed to be reliable, as also is the assumption that the soil is reasonably homogeneous. However, the parameters mv and k used
MODELING STABILITY CHANGES RESULTING FROM VARYING RAINFALL INTENSITIES
169
Table 8.2 Details of the Analysis and Corresponding Safety Factors Situation
Safety Factor Comment
Initial condition—water table as shown in Figure 8.8b
2.14
After three time steps (days) Pore pressure ratio ru = 0.07
1.03 1.01
Long-term steady state flow net shown in Figure 8.8b Water table at ground surface and vertical intercept assumption. ru = 0.60
0.81 0.11
Analysis includes effect of negative pore pressure above water table Slope on point of failure This is the ru value equivalent to the seepage pattern after three time steps The most probable worst-case pore pressure state Normal software method— implies vertical equipotentials and horizontal flow lines
in the steady-state analysis are of much less certain reliability. The coefficient of permeability adopted is higher than most test results indicate. Both coefficients (of permeability and compressibility) are based primarily on conventional oedometer tests. The situation involved here is one where the soil has been subject to endless cycles of seasonally changing effective stresses, and much more detailed laboratory testing is needed to establish reliable values of the parameters. The time steps in the above analysis should therefore be regarded with considerable reservation. They could be in error by an order of magnitude. It is generally the case that cv values measured in the laboratory tend to be a poor representation of those that apply in the field, so this cannot be ruled out in the present case. 8.9 MODELING STABILITY CHANGES RESULTING FROM VARYING RAINFALL INTENSITIES
The above example involved steady rainfall of an intensity equal to or greater than the rate at which water could flow into the ground. With highly permeable slopes, the rate at which water can flow through the ground may by such that only with very intense rainfall will the seepage and pore pressure state change significantly. An example of such a situation has been provided by the makers of the software programs SEEP/W and SLOPE/W (GeoSlope Direct Contact 2005), and is illustrated in Figures 8.11 and 8.12. This example is a good demonstration of the capabilities of the programs, and emphasizes the important influence that the negative pore pressure state above the water table has on the safety factor. The site involved experienced a period of intense rainfall over a four-day period in February 2005,
Elevation (m)
30
20 Initial water table 10
0
10
20
30
40
50
60
Distance (m)
(a) Assumed slope profile and initial water table
SF = 1.35 before intense rainfall
Elevation (m)
30
20
Factor of safety in late Jan 2005 after seven months wet period with 58.4 cm (23 inches) AND before intense four day 22.9 cm rainfall
Unchanged water table 10
0
10
20
30
40
50
60
Distance (m)
(b) Computed safety factor after seven months rainfall
SF = 1.00 after 4 days intense rainfall
Elevation (m)
30
20
Factor of safety in late Jan 2005 after seven months wet period with 58.4 cm (23 inches) AND after intense four day 22.9 cm (9 inches) rainfall
Rise in water table during intense four day rainfall
10
0
10
20
30
40
50
60
Distance (m)
(c) Computed safety factor after four days intense rainfall
Figure 8.11 Changes in water table and safety factor in relation to rainfall (from GeoSlope International 2005).
170
MODELING STABILITY CHANGES RESULTING FROM VARYING RAINFALL INTENSITIES
171
25 Suction profile Feb 11 2005 Suction profile Feb 15 2005
Elevation (m)
20
15 5 m = rise in water table
10
5
0 −150
−100
−50
0
50
100
150
Suction (pore pressure) (kPa) (a) Suction profile change at mid-slope during intense rainfall 1.4 80
70
1.0
0.8
0.6
Jan, 2005
Feb, 15 - 05
60
Feb, 11 - 05
Factorof Safety
Rainfall
50
Cumulative rainfall (cm)
Factor of Safety
1.2
40
Feb, 2005
Date (b) Factor of safety and rainfall history
Figure 8.12 Changes in the pore pressure (suction) and factor of safety related to rainfall history (from GeoSlope International 2005).
triggering a number of severe landslides. There had been steady high rainfall, totaling 58 cm, over the previous seven months from July 2004 till February 2005, but it was the intense rainfall totaling 23 cm in four days between February 11 and February 15 that triggered the major landslides. The program SEEP/W was used to determine the changes in the seepage state and pore pressure due to the 58 cm of rain over the 7-month period followed by 23 cm in 4 days. The pore pressures obtained using SEEP/W
172
SLOPE STABILITY AND SLOPE ENGINEERING
were then transferred to the SLOPE/W program to analyze the safety factor as a function of the rainfall and pore pressure, in the same way as in the previous example. The changes in water table and safety factor are shown in Figure 8.11. It is seen that the 58 cm of rainfall in the 7-month period does not alter the position of the water table, and the safety factor has not fallen below 1.35 at the end of this period. The permeability of the soil is such that during this time the water entering the slope can drain freely through the slope and out again without altering the water table level. It may have some influence on the negative pore pressure (suction) above the water table but not of sufficient magnitude to significantly influence the stability of the slope. Figure 8.12a illustrates the pore pressure conditions on a vertical section through the middle of the slope just before and just after the intense four-day rainfall event. The water table rises by about 5 m, that is, from RL8 to RL13. The changes in safety factor related to rainfall are shown in Figure 8.12b. The slope of the rainfall line in this diagram is a measure of the intensity of the rainfall. It is evident that the intensity prior to February 11, 2005 was insufficient to significantly alter the safety factor, but at the intensity between February 11 and February 15, there is a direct impact on the seepage state and on the safety factor. In this example, the water table has still not risen to the level where it is within the sliding mass. This appears to be an example of a slope consisting primarily of a coarse material of high permeability, where pore pressure changes accompany changes in the degree of saturation. To conclude this section on theoretical analysis of the influence of rainfall on the stability of natural slopes, the following observations are made: 1. The initial “healthy” theoretical factors of safety of both slopes described above are largely due to the influence of the (assumed) negative pore pressures above the initial water table. 2. The above examples are not intended to suggest that it might be possible to predict when a slope will fail from knowledge of the slope and soil characteristics, together with rainfall data. Neither the slope characteristics nor the rainfall pattern will ever be known with the degree of reliability needed to make realistic prediction. In rare situations, where past records enable soil parameters to be determined from back-analysis, and where rainfall records are comprehensive, it might be possible to make predictions that are not entirely unrealistic.
8.10
THE HONG KONG SITUATION
We will now look at some actual records of water table and pore pressure response to changes in rainfall. These are the records from Hong Kong, which are probably the most comprehensive available. Hong Kong, along
THE HONG KONG SITUATION
173
with many parts of the Far East, is subject to extremely intense rainfall from time to time, because it is in the path of typhoons; these typhoons have been the trigger for many large disastrous landslides, resulting in loss of life and severe damage to property. For about the last four decades, Hong Kong has had a specialist geotechnical unit responsible for investigating slope failures and setting up guidelines for new developments. In considering the Hong Kong data, however, we must recognize that we are not dealing with the typical clays found in many parts of Asia, especially the wet tropics of Southeast Asia. The residual soils from Hong Kong granites are quite sandy materials with high permeability, while those of countries such as Malaysia and Indonesia are clays of moderate to high plasticity (see Section 10.2 of Chapter 10). This presumably reflects the more tropical climates of these latter countries compared to Hong Kong. We should note also that the statement of Brand (1982) that “residual soils are invariably unsaturated and of relatively high permeability” may be largely true with respect to permeability but is certainly not true with respect to the degree of saturation The weathered granites of Hong Kong are apparently partially saturated because water drains out of them under gravity forces. This does not happen with the moderate- to high-plasticity clays found in the wet tropics (or in wet temperate climates such as New Zealand), which remain fully saturated except for a surface zone affected by evaporation. 8.10.1
Measurements of Pore Pressure Response
Hong Kong measurements of pore pressure response using stand-pipe piezometers show highly variable behavior, reflecting the two climatic influences mentioned above, namely regular seasonal cycles (wet and dry seasons), and intense, short-duration storms. The forms of response are summarized in Figure 8.13, taken from the Hong Kong Geotechnical Manual for Slopes (2000). This information is very interesting, as it shows that groundwater regimes in Hong Kong respond in quite different ways to the same storm event. Some piezometers respond only to seasonal effects, and some respond only to storm events, some do not respond at all, and there are various combinations of these. This behavior does not appear to have been analyzed in sufficient detail to establish just what the factors are that determine this behavior. Possible factors that contribute to this behavior include the following: •
•
The Hong Kong materials are such that the mechanism of pore pressure change is likely to lie somewhere between the two mechanisms (1) and (2) described in Section 8.6.2, so that the wide variation in response is perhaps not unexpected. Piezometers that show no response of any sort may be located in places where the phreatic surface is fixed by nearby boundary conditions. It
174
SLOPE STABILITY AND SLOPE ENGINEERING
Seasonal response 1 - Little or None
Storm response
2 - Multiple Peaks
3 - Single Peaks
P A - Multiple Peaks T P B - Single Symmetrical Peak T P C - Single Asymmetrical Peak T P D - Slight T P E - None T Legend: Seasonal response Storm response
P - Piezometric level T - Time
Storm event
Figure 8.13 Typical piezometer responses to seasonal changes and storm events in Hong Kong (from Geotechnical Manual for Slopes 2000).
•
is also possible that they may be in zones of very low-permeability material. Piezometers that show seasonal response but no storm response are likely to be located in layers of low permeability, where a long period of steady or regular rainfall is needed before the groundwater system shows any change.
THE HONG KONG SITUATION
•
175
Piezometers that show no seasonal response but some storm response are likely to be in soils of relatively high permeability, so that in normal seasonal conditions water entering the slope can find a way out just as quickly as it enters the slope. It is only in very intense rainstorms that the rate of entry exceeds the rate of exit, with the consequence that the pore pressures increase and the water table rises.
The influence of the negative pore pressure (often called soil suction) within residual soil slopes has long been recognized, and some Hong Kong studies have investigated this factor. Sweeney (1982) describes suction measurements in several slopes, and Ching et al. (1984) examine the contribution this suction makes to the stability of the slopes. Their records from two particular slopes are shown in Figure 8.14. One of the slopes is in weathered volcanic material and the other in weathered granite. The records for the granite material were from September 1980, till March 1981 while those for the volcanic material were primarily during the Soil Suction (kPa) 80
60
40
Soil Suction (kPa) 20
Legend 22.3.1980 13.4.1980 Rainy 7.6.1980 season 2.9.1980 15.12.1980
0
100
5
5
10
10
15
15
20
Depth (m)
100
80
60
40
20
0
20
End of caisson
25
25
30
30
35
35
Hydrostatic pore pressure with respect to water table at 27 m Water table range during suction measurements
(a) Soils derived from volcanic material
1.3.80 - 31.5.1980 13.9.1980 29.11.1980 1.12.1980 28.3.1980 Water table May, 1980
(b) Soil derived from granite
Figure 8.14 Records of soil suction (negative pore pressure) from two sites in Hong Kong (adapted from Sweeney 1982).
176
SLOPE STABILITY AND SLOPE ENGINEERING
rainy season, from March 1980 till November 1980. The following appear to be the most significant features of these records: 1. There is remarkably little variation in the pore pressures over the period the readings were taken. 2. The pore pressures were negative in both materials, and at no time did they “rise” to zero, except for a very shallow zone at the ground surface of the volcanic soil. 3. The behavior of the two soils appears to be quite different. The curves for the volcanic material show high negative pore pressures (suctions), up to 90 kPa, while those for the granite material show only very small suctions, hardly more than 20 kPa, except in the top several meters. 4. The shape of the curves suggests that the volcanic material is behaving like a clay, similar to that illustrated in Figures 3.4 and 3.6 of Chapter 3, while the granite material is behaving more like a free-draining sand. This is consistent with the difference in particle sizes of the materials. The average silt and clay fraction of the materials is about 60 percent for the volcanic material and 30 percent for the granite. 5. With the volcanic soil there is an interesting anomaly in the relationship of the negative pore pressure changes to the changes in the water table. We saw in Chapter 3 (Figures 3.3 and 3.4) that changes in the water table go hand in hand with changes in the pore pressure. The water table change is a consequence of the pore pressure changes. In Figure 8.8, however, substantial changes in the water table are occurring (from about 36 to 27 m), but are not accompanied by changes in the pore pressure. This is not to suggest that there are errors in the recordings. Rather, it is highly likely that some natural variations or discontinuities in the soil account for this behavior. It is an unfortunate fact of life that soils, especially residual soils, seldom behave according to tidy theoretical predictions. 8.10.2
The Wetting Front Method for Estimating Water Table Rise
Mention should perhaps be made here of a method for predicting the water table change during rainy periods put forward by Lumb (1975), which appears to have gained some acceptance in Hong Kong (see the Geotechnical Manual for Slopes 2005). The method is very simplistic. It treats the soil as a rigid material, and involves a wetting front, the thickness of which can be calculated from the rainfall duration together with the permeability and porosity of the soil and the changes in degree of saturation that occur as the front progresses. This wetting front travels downward and the water table rises by the thickness of the front. The mechanics assumed in the
THE HONG KONG SITUATION
177
method are such that it cannot have any validity for clays, and as far as the author is aware, the concept is not supported by field evidence. The records in Figure 8.14 show no evidence of such a front. 8.10.3
Importance of Antecedent Rainfall
Brand (1995) presents an analysis of the relationship between Hong Kong rainfall intensity, including rainfall prior to the main storm event (antecedent rainfall), and slope failures. He concludes that there is no significant influence from the antecedent rainfall, and suggests this is likely to be the case with other residual soils of high permeability. This suggests that any pore pressures that build up during rainfall dissipate rapidly and do not influence subsequent events. The records in Figure 8.13, however, show that water tables do change seasonally at some sites, indicating that pore pressures do not dissipate immediately after rainfall events. Antecedent rainfall is likely to be of considerable importance in clay slopes. 8.10.4 Results of Stability Analysis and Assumptions Regarding the Pore Pressure State
We will return briefly to the point made in Section 8.7, as it is of considerable relevance to the Hong Kong situation. Brand (1982) states: It is probably true to say that the majority of the cut slopes which now exist in residual soils were not designed in the engineering sense and that such slopes defy rational analysis by classical means. This is certainly so for many old cut slopes in Hong Kong, where analysis indicates theoretical factors of safety of less than unity even though the slopes have remained stable for many years under conditions of extremely heavy season rainfall.
It is probable that the theoretical factors of safety mentioned above were obtained using the conventional vertical intercept method for estimating the influence of pore pressure. Beattie and Attewell (1977) describe a study of landslides in Hong Kong that included a tentative method for estimating the influence on stability of changes in the phreatic surface. Figure 8.15 shows a slope cross section and the method used for estimating pore pressure changes. The method makes use of the vertical intercept assumption. The cross section in Figure 8.15 appears to be reasonably typical of Hong Kong slopes; they are very steep and the piezometric surface is also very steep in many slopes. In this situation, the vertical intercept assumption will have a strong influence on the values of safety factor. It seems likely, therefore, that the reason calculated safety factors are less than unity in stable slopes is because of the errors arising from this assumption, and that perhaps the slopes do not “defy rational analysis” after all.
178
SLOPE STABILITY AND SLOPE ENGINEERING
Ground surface Critical piezometric surface Segment for stability analysis
Critical slip surface
Vertical intercept used to estimate pore pressure
Figure 8.15 Hong Kong slopes and influence of piezometric surface on pore pressure (after Beattie and Attewell 1977).
8.10.5
Recommended Safety Factors for Hong Kong Slopes
Some additional information from the Hong Kong Geotechnical Manual for Slopes that is of interest is the values of safety factors recommended in relation to storm intensity and acceptability of risk. These are shown in Figure 8.16. They seem surprisingly low, but this may reflect the fact that they are, in fact, generally conservative because of the vertical intercept assumption made regarding the pore pressures discussed above. 8.10.6
Triaxial Tests and Back-Analysis of Landslides
Malone and Shelton (1982) describe an analysis of landslides in Hong Kong that occurred between 1978 and 1980. This included a back-analysis study of a large number of landslides in various materials and a comparison with the results of triaxial testing of the materials. The results for soils of volcanic origin and those weathered from granite are summarized in Figure 8.17. The Mohr-Coulomb envelopes are from a large number of triaxial tests on samples from a variety of sites within the respective materials, so they are not specific to the landslides. The tests were carried out over a wide range of confining pressures. It is evident that the values from back-analysis lie both above and below the Mohr-Coulomb line. This is perhaps surprising, but we should remember that Hong Kong soils are heterogeneous, and the Mohr-Coulomb line is an average from many triaxial tests. The scatter in these tests is not indicated in the single failure line in the figure. A significant feature illustrated dramatically in Figure 8.17 is the relatively low stress levels involved in the failures, and the necessity for triaxial tests to be conducted at low stress levels to be relevant to actual field situations.
THE HONG KONG SITUATION
179
New slopes:
Risk to life
Recommended factor of safety against economic loss for a 10yr return period storm
Economic risk
Recommended factor of safety against loss of life for a 10yr return period storm Negligible
Low
High
Negligible
> 1.0
1.2
1.4
Low
1.2
1.2
1.4
High
1.4
1.4
1.4
Note: (1) In addition to a factor of safety of 1.4 for a 10 year return period rainfall, a slope in the high risk-to-life category should have a factor or safety of 1.1 for the predicted worst groundwater condition (2) The factors of safety given in this Table are recommended values. Higher or lower factors of safety might be warranted in particular situations in respect of economic loss.
Figure 8.16 Recommended design safety factors for Hong Kong slopes.
8.10.7
Concluding Remarks on the Hong Kong Situation
Shear stress (kPa)
150
100
Ä ,f
7°
=3
a
cÄ
50
0
=1
kP
Points from back-analysis —each point represents one landslide
Shear stress (kPa)
The Hong Kong records are the most comprehensive available for particular soil types within a small area. Despite this, they do not lead to any tidy conclusions regarding the mechanisms of pore pressure change during rainfall, or means for predicting changes and their influence on slope stability. They appear to suggest that the mechanism of change for Hong Kong soils is somewhere between that for a partially saturated, rigid material and
150
100 cÄ
50 100 150 Normal stress (kPa)
200
= fÄ
°
35
k
50 Volcanic soils
Granite soils 0
=5
, Pa
0
50 100 150 Normal stress (kPa)
200
Figure 8.17 Results of back-analysis of landslides compared to triaxial tests for Hong Kong soils (adapted from Malone and Shelton 1982).
180
SLOPE STABILITY AND SLOPE ENGINEERING
that for a clay. The weathered granites appear to belong in the former category, while the weathered volcanic soils are closer to the latter category. Whatever the case, the Hong Kong data demonstrates clearly that slope design in residual soils will normally involve a large measure of judgment and experience, hopefully assisted to some extent by the use of analytical methods. 8.11 BACK-ANALYSIS METHODS TO DETERMINE SOIL PARAMETERS
There are various methods by which the shear strength parameters c and φ can be determined from back-analysis. This approach overcomes some of the difficulties associated with making direct measurements, in particular the following: 1. The great difficulty or impossibility of taking any sort of undisturbed samples in some residual soils, especially those containing substantial remnants of unweathered rock. 2. The difficulty of knowing whether samples are truly representative of the soil in situ, even when undisturbed samples of apparent good quality can be obtained. Some residual soils are homogeneous, but more often they are not, and large-diameter samples are needed to take adequate account of variations in properties. Several back-analysis methods are described in the following sections. However, we should recognize at the outset that there are still quite severe limitations on the usefulness of these methods. In particular, they can be applied only to homogeneous materials, or the assumption has to be made that the material is homogeneous. Also, there is normally considerable uncertainty about the seepage and pore pressure conditions in the slope. There may also be some uncertainty regarding the geometry and the unit weight of the soil, but this is minor compared to the pore pressure uncertainty. 8.11.1
Back-Analysis of a Single Slip or a Single Intact Slope
Consider the slope shown in Figure 8.18. If there is an existing slip in the slope, then we can assume the safety factor is unity, and by a back-analysis of this circle we can determine shear strength parameters c and φ that give SF = 1. However, there is not a unique combination that satisfies this criterion, only a range of combinations of values. Even if there was not an existing slip in the slope, we could still assume it to have a safety factor of unity and by back-analysis obtain another set of combinations of c and φ that give SF = 1. The two sets of values
BACK-ANALYSIS METHODS TO DETERMINE SOIL PARAMETERS
181
Center of slip circle
Ground surface and phreatic surface
Position of slip circle
Figure 8.18 Slope for back-analysis to determine the strength parameters c and φ .
obtained in this way are shown in Figure 8.19. It is seen that the values are different, although they coincide at one point. We would not expect to get the same range of values because the first set (from the known slip) has been obtained from a single fixed slip—the one shown in the figure. The second set has been obtained without any constraints on the location of the slip circle, so that this set represents a range of different circles. Figure 8.19 suggests the means by which we can obtain a unique set of values from the analysis of the slope with the actual slip in it. If we take each of the sets of values obtained from the actual slip, and then reanalyze the slope assuming it is an intact slope (no existing slip in it) seeking to determine the critical circle, we will obtain a series of critical circles in different locations. This is illustrated in Figure 8.20.The true combination of c and φ is that which produces a critical circle coinciding with the actual failure circle. The values obtained in this way are c = 18 kPa and φ = 30◦ . There are several other ways in which to determine the true values of c and φ . For example, they are given by the point at which the two graphs coincide in Figure 8.19, although this point is poorly defined because of the tangential nature of the intersection. Other methods are described by Wesley and Lelaratnam (2001). 8.11.2
Analysis of a Number of Slips in the Same Material
It is a big advantage when more than one slip is available in the same material. To obtain the strength parameters c and φ we could apply the method described above to each slip individually and then use an averaging procedure to obtain the most representative values. A better way is that
182
SLOPE STABILITY AND SLOPE ENGINEERING
Cohesion intercept cÄ (kPa)
60
(b) Intact slope 40
20 (a) Slope with actual slip
0.5
1.0
1.5
tan fÄ
Figure 8.19 Range of values of c and φ obtained by back-analysis of an intact and failed slope.
fÄ = 45°
Line of centers of slip circles showing corresponding f values
35°
Center of actual slip circle 25° 15°
5°
Position of actual slip circle
45°
Ground surface and phreatic surface
35° 25°
15°
f = 5°
Figure 8.20 Circles corresponding to combinations of c and φ .
BACK-ANALYSIS METHODS TO DETERMINE SOIL PARAMETERS
183
BROWN LONDON CLAY First-time Slides
30
Shear strength (kPa)
Long-term
Equilibration may not be complete
ru = 0.35
20
10
0.3
Grange Hill
0.25
cÄ = 1.0 kPa f = 20°
Northolt
West Acton
Crews Hill
Cuffley Hadley Wood Sudbury Hill
cÄ = 0 fÄ = 20°
0
10
20
30
40
50
Effective normal stress (kPa)
Figure 8.21 Values of c and φ obtained from back-analysis of a number of slips in brown London Clay (after Chandler and Skempton 1974).
illustrated in Figure 8.21, which is for brown London clay (after Chandler and Skempton 1974). This is not a residual soil but the method is equally valid for residual soils. In this example data are available from seven different sites in the same material. By back-analysis the average shear strength needed to maintain stability and also the average normal stress on the slip surface on which failure has taken place have been determined. These values have then been plotted on a graph of shear stress against effective normal stress, and a best fit line drawn though the points to establish the Mohr-Coulomb failure line and the c and φ values. The horizontal line through some of the data points in the graph reflects uncertainty about the seepage condition and pore pressures in the slope. The line indicates the range of possible effective normal stress values arising from this uncertainty. 8.11.3
Analysis of a Large Number of Intact Slopes (No Previous Slips)
It is possible to collect data on slope heights and slope angles for a particular geological formation or soil type, that is, for any material that is reasonably homogeneous, and use these data to deduce the strength parameters by a curve-fitting procedure. The data should be gathered from those slopes considered to be closest to failure, in other words the steepest slopes for any particular height. The data are then plotted in graphical form as shown in Figure 8.22 and a curve drawn defining the upper limit of combinations of slope height and angle that will remain stable.
184
SLOPE STABILITY AND SLOPE ENGINEERING
High fÄ low cÄ
C
Cohesion intercept cÄ
Slope height
Low fÄ high cÄ
tA
B
in
“Best fit” envelope of available data –upper limit of stable height /angle combinations
Po
A
Po
int
B Point C
P
tan fÄ Point P defines the appropriate values of fÄ and cÄ
Slope angle
Figure 8.22 Curve fitting to height/slope data to determine c and φ .
In addition to the curve fitted to the field data, two curves are also shown in Figure 8.22 to indicate the way in which the shape of the curves varies with the relative magnitude of c and φ . For any given values of c and φ , and fixed seepage condition (defined by an ru value), there will be a unique combinations of slope heights and slope angles that will be stable. A procedure involving trial and error can then be used to fit a curve to the field data. This procedure can be quite tedious, but systematic methods can be used to avoid time-consuming trial-and-error procedures. For example, we can select two or three points on the curve, such as A, B, and C, and then use the procedure in Section 8.11.1 to determine combinations of c and φ for each point and plot these as graphs on a common graph, as shown in Figure 8.22. The intersection of these graphs (the point P) defines the values common the whole curve and thus the values we are seeking.
8.12 8.12.1
SLOPE DESIGN Selection of the Profile for a New Cut Slope
It is perhaps appropriate to close this chapter by revisiting and reemphasizing what was said in the opening sections, namely that the selection of an appropriate profile for a new cut slope in residual soil is a matter of judgment based more on nonanalytical approaches than on analytical estimation. Despite this, much of the chapter has been spent looking at theory
SLOPE DESIGN
185
and analytical methods, particularly in relation to the influence of climate and rainfall on slope stability. This has not been done to stimulate the use of analytical methods as a design process in preference to nonanalytical methods. Rather, it has been done because estimating the influence of rainfall is a predominant issue in selecting stable slopes, and knowledge of the theoretical mechanism (or mechanisms) by which rainfall influences stability ought to be an aid in the process of using judgment to determine slope profiles. A further point that should be emphasized here is that the use of nonanalytical methods should in no way diminish the importance of site investigations, especially investigations aimed at providing a comprehensive picture of the geology of a site. A simple illustration of the importance of this is given in Figure 8.23. The prime objective of a site investigation in relation to the design of cut slopes must be to determine an accurate soil profile at the location of the cut, especially in weathered igneous rocks such as granite. In many situations, especially in highway construction, it is inevitable that slopes will be steep and safety factors will not be high. In this situation it is imperative to take maximum advantage of the stronger materials, especially any unweathered rock. The cut should be vertical or near vertical in competent rock to minimize earthworks and to make room for more gentle slopes in the soil layers in the upper levels of the cut, as indicated in Figure 8.23. Profiles of the sort illustrated in Figure 8.23 are common in weathered granites, such as those found in Hong Kong and Malaysia. It is highly desirable to determine the profiles prior to commencement of construction rather than during excavation. For practical reasons slopes are cut from the
Original ground surface
Co
m
“S Su
ap
ple
te
ro
rfa
ce
of
lite
”—
so
ly
we
at
un
we
ath
dr
er
oc
k
he
ed
re
d
ro
ck
so
il l
ay
Steep slope in sound rock minimizes earth works and allows gentler slopes in softer layers near the surface er
Figure 8.23 Profile of a cut slope in weathered igneous rock such as granite.
186
SLOPE STABILITY AND SLOPE ENGINEERING
top down in their final profile, and any adjustments to this profile made necessary by soil conditions revealed during excavation pose construction difficulties. It is not an easy matter getting excavation equipment back up to the top of a cut slope to reshape the profile. For determining the surface of the sound rock, geophysical methods can be a better approach than conventional boreholes. In volcanic materials, the increase in strength with depth found in weathered granites may be very small or insignificant, in which case a uniform slope angle is likely to be the most appropriate. However, volcanic material is likely to be rather unpredictable, which again emphasizes the need for thorough site investigations. 8.12.2
To Bench or Not to Bench a Slope?
Figure 8.24 shows a slope that has incorporated benches, or berms, into its design. These are not infrequently considered to be an aid to improve the stability of a slope, or at least a means to control and minimize erosion. Whether benches (berms) really are a desirable feature of slope design is a question that is almost invariably raised during discussions or presentations on the design of cut slopes, at least in the countries of the wet tropics. There is no simple or single answer to this question, but the following comments may be useful: 1. Benches do not normally have a significant influence on the general stability of the slope. If the slope is cut without benches but with the same average inclination as the benched slope (as indicated in Figure 8.24) the stability would be the same. It can be argued that benches may have an adverse influence on stability because water will tend to pool on the benches and result in greater infiltration into the slope.
Benches—to intercept run-off and control surface erosion 3m
10 m
Figure 8.24 Benched slope versus unbenched slope.
REFERENCES
187
2. The only useful function that benches can have is to control erosion and provide a means of access to the slope. Their usefulness in controlling erosion will depend very much on the installation of properly designed sealed surface drains on the benches and on regular maintenance to keep the drains functioning as intended. 3. The author is a somewhat less than enthusiastic advocate of benches on slopes because he has inspected a very large number of benched slopes in which the benches are clearly not performing any useful function. The drains that were incorporated at the time of design and construction have become blocked with eroded material or vegetation, and in many cases surface slips of the benches have rendered them ineffective. Where such slips occur they tend to promote concentrations of surface runoff and lead to rapid increases of surface erosion. 4. For highly erodible soils, such as weathered granite, it is undoubtedly the case that control measures are needed and benches may be the most practical measure available. However, it is imperative that measures are adopted to ensure regular and effective maintenance of the benches. 5. For erosion-resistant soils, such as allophane clays, there is no benefit to be gained from the use of benches, and they probably do less good than harm. 8.12.3
A Note on Vegetation Cover on Slopes
Vegetation generally has a positive effect in helping to stabilize slopes. Its influence is threefold: 1. Vegetation reduces the amount of water seeping into the ground, and thus helps to minimize pore pressures. 2. Vegetation extracts moisture from the ground, which also assists in minimizing pore pressures. 3. Vegetation helps to minimize surface erosion. This may not have a direct influence on the stability of the slope, but is beneficial because a well-vegetated surface is much less likely to allow seepage into the slope than a bare eroded surface. REFERENCES Brand, E. W. 1982. Analysis and design in residual soils. Proceedings ASCE Specialty Conference on Engineering and Construction in Tropical and Residual Soils, Hawaii, 89–143. Brand, E. W. 1995. Slope stability in tropical areas. Proceedings, International Conference on Landslides, Balkema, 2031–2050.
188
SLOPE STABILITY AND SLOPE ENGINEERING
Beattie, A. A., and L. J. S. Attewill. 1977. A landslide study in Hong Kong residual soils. Proceedings 5th Southeast Asian Conference on Soil Engineering, Bangkok, 1955, 177–187. Chandler, R. J., and A. W. Skempton. 1974. The design of permanent cutting slopes in stiff fissured clays. Geotechnique 24(4): 457–466. Ching, R. K. H., D. J. Sweeney, and D. G. Fredlund. 1984. Increase in factor of safety due to soil suction in two Hong Kong slopes. Proceedings 4th International Symposium on Landslides. Toronto: BiTech, 617–623. Geo-Slope International. 2005. Why do slopes become unstable after rainfall events? Illustration Example in Internet Newsletter: Direct Contact, May 2005. Geotechnical Manual for Slopes, 2nd ed. 2000. Geotechnical Engineering Office, Civil Engineering Department, The Government of Hong Kong. Lumb, P. 1975. Slope failures in Hong Kong. Quarterly Journal of Engineering Geology 8: 31–65. Malone, A. W., and J. C. Shelton. 1982. Landslides in Hong Kong. Proceedings ASCE Specialty Conference on Engineering and Construction in Tropical and Residual Soils, Hawaii, 425–443. Sweeney, D. J. 1982. Some in situ soil suction measurements in Hong Kong’s Residual soil slopes. Proceedings 7th Southeast Asian Geotechnical Conference, Hong Kong, 91–105. Wesley, L. D. 1977. Shear strength properties of halloysite and allophane clays in Java, Indonesia. Geotechnique 27(2): 125–136. Wesley, L. D., and V. Lelaratnam. 2001. Shear strength parameters from back-analysis of single slips. Geotechnique 51(4): 373–374.
CHAPTER 9
VOLCANIC SOILS
9.1
INTRODUCTION AND GENERAL OBSERVATIONS
Soils of volcanic origin are a broad group of materials with widely divergent properties. Those found in dry climates, and even in some of the dryer temperate climates, tend to be low-plasticity or nonplastic materials with properties that are not particularly unusual. Those found in wet temperate climates and especially those in the wet tropics tend to be very different, with unique characteristics. Since the author’s experience has been primarily with the latter group, and this is the one that is so puzzling to geotechnical engineers, this group will be the main focus here. 9.2
ALLOPHANE CLAYS
Allophane clays, perhaps more than any other soil, illustrate the misconceptions that can arise when conventional soil characterization procedures are applied to residual soils. On the one hand, many engineering projects involving allophane clays, such as Cipanunjang dam in Indonesia (mentioned in Chapter 1), have been successfully completed without difficulty, and, on the other hand, the clays have something of a reputation as problem soils (Gidigasu and Bani 1972; Morin and Todor 1975). This designation has arisen primarily because their behavior does not fit comfortably into what geotechnical engineers have come to regard as normal. Their behavior may be problematic in the sense that it does not fit comfortably into normal patterns, but this does not necessarily qualify them as problem soils. The very fine-grained nature of the soils, together with their extraordinarily 189
190
VOLCANIC SOILS
high natural water contents, creates the expectation that they will be highly plastic, of high compressibility, and of low strength, and that they may even be a source of shrinkage and swell problems. None of these is the case; in fact, the opposite is generally true, so that conventional geotechnical characterization indicators are shown to be faulty. Their unusual characteristics do, however, give rise to some problems, one of the most prominent of which has been in the area of compaction and compaction control. 9.2.1
Performance of Natural Hill and Mountain Slopes
The stability of these soils on steep slopes has been observed and commented on by various investigators, including the author (Wesley 1973), with respect to the steep, irrigated, terraced rice fields of Indonesia and other Southeast Asian countries. Rouse et al. (1986) report that in Dominica slopes remain stable at inclinations of around 40◦ , even when used for crop cultivation. Belloni and Morris (1991) report on slopes in Ecuador with inclinations ranging from 35◦ to 55◦ . The wet climates in which these slopes exist, and their use as rice fields mean that there must be significant seepage pressures within them for much of the year. Peak φ values of the soils are in the range of 30◦ to 40◦ , so clearly their shear strength must have a significant cohesion (c ) intercept for the slopes to remain stable. Comment on irrigated rice fields has already been made in Section 5.3.5 (Chapter 5); back-analysis showed c values of 8–15 kPa. There are, of course, limits to the stability of these soils. The data quoted above from Rouse et al. (1986) were obtained in the course of investigations into massive shallow landslides resulting from two devastating hurricanes, while the data from Belloni and Morris (1991) were obtained during investigations into similar shallow slides induced by an earthquake. 9.2.2
Formation of Allophane Clays
The formation and composition of allophane clays is complex, and most of the research on the subject comes from the discipline of soil science rather than soil mechanics. This research and associated literature has grown enormously in the last two three or four decades since the term allophane first found its way into geotechnical literature, and it shows a number of new and interesting findings. First, it shows that allophane seldom occurs by itself. Instead, it is almost invariably found with other clay minerals, especially a mineral called imogolite. Allophane seems to be almost inseparably linked to imogolite, and many papers on allophane are, in fact, on allophane and imogolite rather than on allophane alone. Second, it shows that allophane is not strictly amorphous, as early literature asserted. Both allophane and imogolite have some crystalline structure, albeit of a very different nature to other well-known clay minerals.
ALLOPHANE CLAYS
191
Allophane clays are formed by the in situ weathering of volcanic material, primarily ash, although they may be formed from other volcanic material. This parent material may be either basic or acidic in nature, but most allophane clays are derived from andesitic ash. This may only be because andesitic eruptions produce far more ash than other types of eruptions. Andesitic volcanoes erupt irregularly, at intervals of years, decades, or centuries, with each bout of activity continuing in bursts over a considerable period of time. Such eruptions produce very large quantities of ash, consisting of very fine particles, mainly in the silt or fine sand size range. The primary condition for allophane formation is that the parent material consist of noncrystalline (or of poorly ordered structure) composition. Volcanic ash meets this criteria; it is formed by the rapid cooling of relatively fine-grained pyroclastic material, the cooling process being too rapid for the formation of well-ordered crystalline structures. The fine particles consist of volcanic glass, that is, they are noncrystalline (or amorphous), similar to the shiny black rock known as obsidian. The parent volcanic ash from which allophane clays are formed is generally in the silt to sand particle size range. In addition to the above requirement of noncrystalline parent material, it appears that the weathering environment must be well drained, with water seeping vertically downward through the ash deposit. High temperatures also appear to favor or accelerate the formation of allophane clays. Generally, the allophane content of such clays in Indonesia is substantially higher than in similar clays in cooler, temperate climates, such as that of New Zealand. Allophane clays may be very deep; in Indonesia the writer has encountered cuts in these materials up to about 30 m deep, while site investigation drilling has shown depths of up to almost 40 m. This thickness results from successive eruptions and associated ash showers, with weathering progressing as the thickness grows. Examination of cut exposures in West Java, Indonesia, shows the individual layer thickness to vary generally between about 100 and 300 mm. Allophane is believed to undergo further chemical weathering through the following sequence: Volcanic ash −→ allophane −→ halloysite −→ kaolinite −→ sesquioxides −→ laterite This weathering process is essentially one of chemical conversion and leaching out of silica by seeping pore water. As the silica content decreases, the concentration of iron and aluminum increases, in the form of sesquioxides, that is, the hydrated forms of iron and aluminium oxide (geothite and gibbsite). These tend to act as cementing agents, which bring about the formation of the hard concretions that make up laterite. A requirement for the weathering process to progress from allophane to halloysite and on
192
VOLCANIC SOILS
to kaolinite (and the sesquioxides) appears to be a warm, wet climate of the kind found in the wet tropics. In Indonesia, for example, in the warm regions below an elevation of about 1000 m, red clays are found in which the predominant clay mineral is halloysite, with sufficient concentrations of iron compounds to produce the red color. Above 1000 m, where temperatures are lower, weathering does not appear to progress much beyond the formation of allophane; there is no red color in the soil, and only a small proportion of halloysite. In New Zealand, the volcanic ash weathers to allophanic clays but does not appear to progress much beyond this stage, regardless of altitude. The ratio silica/alumina is sometimes used by soil scientists as an indicator of weathering and a basis for the classification of these soils. 9.2.3
Structure of Allophane Clays
The precise structure of allophane clays is somewhat problematic. Their extraordinarily high natural water contents and void ratios clearly indicate an unusual material and call for an explanation in terms of either structure or chemical composition (or both). Various explanations have been offered over the years. As mentioned above, allophane has been described in the past as noncrystalline or amorphous, and gel-like. However, electron microscopy studies over the past 20 years (Wada 1989; Jacquet 1990) show that the material in its natural state does have an ordered structure—consisting of aggregations of spherical allophane particles with imogolite threads weaving among them, or forming bridges between them, as illustrated conceptually in Figure 9.1. This is based on an electron micrograph (Wada 1989) of the material in its undisturbed state. Also shown in Figure 9.1 is a form of halloysite clay minerals, which frequently occur in association with allophane and are a progression of the weathering sequence that forms allophane.
Allophane spheres
Imogolite threads
(a) Allophane and imogolite
(b) Halloysite
Figure 9.1 Conceptual representation of structure of allophane, imogolite, and halloysite clay minerals.
ALLOPHANE CLAYS
193
Remolding allophane clays appears to break up the aggregations of particles and threads spanning between them and turns the material into a homogeneous unstructured mass. This is generally accompanied by some loss of strength and an increase in compressibility, as well as a reduction in permeability. 9.2.4
Particle Size
Despite considerable uncertainty regarding their precise composition and structure, there is general agreement among soil scientists that allophane and imogolite are very fine-grained materials. On the basis of X-ray diffraction and electron microscopy, Wada (1989) describes allophane particles as nearly spherical with diameters in the range 3.5–5 nm (1000 nm = 1 μm), while imogolite appears as threads, smooth and curved, varying in diameter from 10 to 30 nm, and extending up to several micrometers in length. By way of comparison, the plate-like particles of kaolinite are in the range 0.3–3 μm across and one-tenth to one-third of this in thickness. Allophane particles are thus one to two orders of magnitude smaller than kaolinite particles. Measurement of the particle size of allophane clays by conventional sedimentation procedures using normal dispersing agents can be very difficult (Wesley 1973) because of a strong tendency of the particles to flocculate. At the start of such tests, sedimentation appears to take place in the normal manner, but after a short time (5–15 min) flocculation occurs and a sharp boundary appears between the flocculated material and clear water above it. None of the common dispersing agents prevent this happening and the author has not come across alternative more effective dispersing agents. Jacquet (1990) mentions the same problem with New Zealand allophane soils. Despite the above difficulty, it is worth quoting the results of some conventional particle size tests. Tests on New Zealand samples (Jacquet 1990) showed a fine sand fraction range from 4 to 18 percent, and a clay fraction from 21 to 56 percent, the latter figure being somewhat suspect because of the flocculation problem. Tests on Indonesian samples (Wesley 1973) showed a fine sand fraction of less than 5 percent, and an indicative clay fraction of greater than 75 percent. Lohnes and Tuncer (1977), from studies on five samples from Hawaii, obtained clay fractions ranging from only 3 to 29 percent. These low values could reflect flocculation during the sedimentation testing, since conventional test procedures were followed. 9.2.5
Natural Water Content, Void Ratio, and Atterberg Limits
The natural water content of allophane clay covers a wide range, from about 50 to 300 percent. This corresponds to void ratios from about 1.5 to 8. It appears that water content is a reasonable indication of allophane
194
VOLCANIC SOILS
content—the higher the water content the greater the allophane content. Atterberg limits similarly cover a wide range, and when plotted on the conventional plasticity chart invariably lie well below the A-line. This means that according to the Unified Soil Classification System they are silts. However, they do not display the characteristics normally associated with silt—the tendency to become “quick” when vibrated and to dilate when deformed. At the same time they are not highly plastic like true clays, so they do not fit comfortably into conventional classification systems. Figure 9.2 shows a plot of the Atterberg limits of allophane clays on the plasticity chart. For volcanic ash clays in Java, Indonesia, there is a reasonably clear correlation between the relative proportions of allophane and halloysite in the soil and the water content. Increases in natural water content and Atterberg limits accompany increases in allophane content and corresponding decreases in halloysite content. The relationship is illustrated in Figure 9.3.
9.2.6
Influence of Drying
Drying has a very important effect on allophane clays. Frost (1967) gave the first systematic account of this effect for both air and oven drying on tropical soils belonging to the allophane and halloysite group. He showed that clays from the mountainous districts of Papua New Guinea with values of plasticity index ranging from about 30 to 80 in their natural state become nonplastic when air or oven dried. Wesley (1973) describes similar effects from the allophane clays of Java, Indonesia. The properties of the clay described in this paper apply to the clay in its natural state, that is, without air or oven drying, unless otherwise stated.
100
Plasticity Index
80 60
e
lin
A-
40 20
0
40
80
120
160
200
240
Liquid Limit
Figure 9.2 Atterberg limits of allophane clays on the plasticity chart.
ALLOPHANE CLAYS
90 250
70
Halloysite (%) 50 30
195
10
Natural water content and Atterberg limits (%)
Liquid limit Natural water content Plastic limit 200
Note: The total percentage of allophane plus halloysite is about 90%. The remaining 10% is made up of coarser particles of varying composition.
150
100
50
0
20
40
60
80
100
Allophane (%)
Figure 9.3 Atterberg limits and natural moisture content related to allophane content (Wesley 1973).
9.2.7
Degree of Saturation, Liquidity Index, and Sensitivity
Data from two profiles investigated by Wesley (1973) are shown in Figure 9.4. This illustrates two important points. First, the soils are essentially fully saturated except within the top 1 or 2 m from the ground surface. Second, the natural water content fluctuates rather randomly with respect to the plastic limit and the liquid limit. This means that the liquidity index varies from about zero to greater than unity, with an accompanying variation in sensitivity. Measured sensitivity values in the upper profile in Figure 9.4 range from about 1 to 2.5. In the lower profile sensitivity was not measured directly, but the water content exceeds the liquid limit at a depth between 10 and 16 m, indicating very high sensitivity soil. The author has encountered moderate to high sensitivity at other sites in Indonesia. Jacquet (1990), from a study of New Zealand samples, obtained sensitivity values ranging from about 5 to 55, confirming the author’s experience that the sensitivity of allophane clays in New Zealand appears to be generally higher than in Indonesia.
196
VOLCANIC SOILS
Water content PL and LL 0
2
50
100
Su (kPa)
Clay fraction (%) 150 20 40 60 80 100 60
100
Degree of saturation (%)
Sensitivity 140
0
1
2
3 90
100
110
LL w PL
Depth (m)
4
6
8
10
Water content, PL and LL 50 100 150
0
5
200
60
Su (kPa) 100
140
LL w PL
Depth (m)
10
15
20
25
30
30
Figure 9.4 Typical profiles of basic properties of allophane clays.
Wallace (1973) and Moore and Styles (1988) report high liquidity index values for samples from Papua New Guinea, generally between 0.8 and 1. Wallace states that the measured degree of saturation of his undisturbed samples was generally less than one, which is perhaps not surprising, as all the samples were taken from close to the ground surface.
ALLOPHANE CLAYS
9.2.8
197
Identification of Allophane Clays
Various techniques are used by soil scientists to identify allophane: these are primarily X-ray diffraction and electron microscopy. Such methods are not normally readily available to geotechnical engineers. For engineering purposes, sufficient indicators of the presence of allophane are the following: • • • •
Volcanic parent material Very high water contents Very high liquid and plastic limits lying well below the A-line on the plasticity chart Irreversible changes on air or oven drying— from a plastic to a nonplastic material
If all of these are applicable then the soil is almost certainly allophanic. 9.2.9
Compressibility and Consolidation Characteristics
Typical results from oedometer tests on undisturbed samples from Indonesia and New Zealand are shown in Figure 9.5, plotted in both the conventional manner and also using a linear scale for pressure. As discussed in Chapter 4, the conventional e-log p plot presents a misleading picture of soil compressibility. The log curves in Figure 9.5a suggest that all the samples have similar compressibility characteristics with yield pressures of varying magnitude. However, when plotted using a linear pressure scale this is clearly no longer the case. Only samples 1 and 2 show a yield pressure. This yield pressure arises from the structure of the soil created by the weathering process, but why some samples show a yield pressure and others do not is unknown. It is likely to be related to the original denseness of the parent material, but this is uncertain. These graphs emphasize two important points made earlier in Chapter 4. First, to gain a clear picture of consolidation behavior it is necessary to plot the results using a linear scale. Second, the portion of the graph of interest in foundation design is often close to linear with respect to pressure, and favors the use of the linear parameter mv (or constrained modulus D) for settlement calculations rather than the log parameters Cc and Cs . Further examination of these clays shows that there does not appear to be a relationship between the initial void ratio and the compressibility of the samples, which seems rather surprising. Figure 9.6 shows data illustrating this point from a large number of oedometer tests on undisturbed samples of allophane clay. The constrained modulus D has been calculated for the load increment 0–200 kPa, and again between 1600 and 2000 kPa, and the values plotted against the initial
198
VOLCANIC SOILS
Pressure (kPa) 100 1000
10
5000
5
Void Ratio
4
3
2
1 (a) Log scale
Compression (Vertical strain %)
0
500
Pressure (kPa) 1000
1500
2000
10
20
30
40
Sample 1 2 3 4 5 6 (b) Linear scale
Figure 9.5 Oedometer tests showing compression versus pressure on log and linear scales.
void ratio. There is considerable scatter in the values, but there is no clear trend toward higher compressibility as the void ratio increases from 2 to nearly 6. The range decreases as the stress level increases. This trend is to be expected from the evidence in Figure 9.5b, which shows that the slope of the graphs is becoming fairly similar at the higher stress levels. We will now look at the consolidation rates encountered in these soils. In Table 4.2 of Chapter 4, the coefficient of consolidation of volcanic ash
ALLOPHANE CLAYS
199
20 (a) 0–200 kPa
Constrained modulus D (kPa)
15
10
5
0 15 (b) 1600–2000 kPa 10
5
1
2
3 4 Initial void ratio
5
6
Figure 9.6 Constrained modulus (D) versus initial void ratio.
clay is given as ranging from 0.01 to 200 m2 /day, which is an extremely wide range. However, this is the range measured in conventional oedometer tests, provided the test is taken to reasonably high pressures. The range at low stress levels, relevant to most foundation situations, is likely to be much smaller and near the higher end of this range. Figure 9.7 shows typical results for three loading increments from an oedometer test on an undisturbed sample of allophane clay, plotted using the Taylor square root of time procedure. The shape of the curves changes as the stress level increases. As discussed in Section 4.6 of Chapter 4, curves of the shape obtained at the lower stress levels indicate very rapid dissipation of pore pressure, and reliable estimates of t90 and the coefficient of consolidation are not possible. By the time the highest stress level is reached the soil is behaving in a conventional manner and the cv value can be determined. To overcome the difficulties associated with the very short drainage path length in a conventional oedometer test, pore pressure dissipation tests can be carried out in a triaxial cell. With small triaxial samples the length is normally 7.6 cm; with drainage from one end only, the drainage path length is also 7.6 cm, which means that the consolidation time will be increased by the square of the ratio 7.6/1 = 58, assuming a sample thickness of 2.0 cm in the oedometer test. Even with this increase, it can still be difficult to measure the pore pressures during the first one or two loading increments. A description of tests of this type is given by Wesley (2001).
200
VOLCANIC SOILS
min 4
Time 2
0
6
7
40
Pa 0k 56 –2 80 12
Total compression (%)
20
16
0–
60
32
0
kP
a
80 16–32 kP
a
100
Figure 9.7 Typical square root of time plots from oedometer tests.
A summary of the result of these tests is presented in Figure 9.8. The tests were carried out on four undisturbed samples, two each from Indonesia and New Zealand. On the New Zealand samples, the dissipation tests were repeated after thoroughly remolding the samples. Figure 9.8a shows the coefficient of consolidation values obtained from all the dissipation tests. It is seen that the cv value decreases by approximately four orders of magnitude as the stress increases from 50 to 1000 kPa. The cv value from the remolded samples is consistently low and close to the end (high stress level) value from the undisturbed samples. Remolding the soil apparently destroys the open structure of the undisturbed soil, which is believed to account for the high permeability. With the Indonesian samples, permeability measurements were also made between each consolidation stage; the results are shown in Figure 9.8b. The coefficient of permeability shows an identical trend to the cv values, as would be expected, confirming that it is the very high permeability of the soil that accounts for the high cv values. 9.2.10
Strength Characteristics
Typical results from CPT penetrometer tests are shown in Figure 9.9. One is from Indonesia and the other from New Zealand. These are fairly similar, and show that while the in situ strength is reasonably uniform, it does have small fluctuations over the full profile, and there are some zones with
201
ALLOPHANE CLAYS
50
50 Coefficient of permeability
1 Indonesian samples —undisturbed 0.1
0.01
0.001
100
100
1
10 Coefficient of consolidation
0.1
1
0.01
0.1 Indonesian samples only—undisturbed
0.001
New Zealand samples —remolded
0.0004 20
10
1000 2000
Coefficient of permeability (10−9 m/sec)
10
Coefficient of consolidation (cm2 /sec)
Coefficient of consolidation (cm2/sec)
New Zealand samples —undisturbed
0.01
0.0004 20
100
1000 2000
Pressure (kPa)
Pressure (kPa)
(a) Pore pressure dissipation tests
(b) Pore pressure dissipation and permeability tests
Figure 9.8 Summary of results from pore pressure dissipation tests.
0
5
Cone resistance (MPa) 10 15
20
0
20
10
Depth (m)
Depth (m)
10
Cone resistance (MPa) 5 10 15
20
30
20
30 Kamojang Geothermal Power Station, Indonesia
Omata oil storage tank, New Zealand
Figure 9.9 Cone penetrometer tests (CPT) from sites in Indonesia and New Zealand.
202
VOLCANIC SOILS
considerably higher values. These are believed to be zones of coarser material (lapilli) within the clay. The cone resistance varies between about 1 and 3 MPa. Using a correlation factor (Nk ) of 15, this corresponds to an undrained shear strength range of about 65–200 kPa. Values of undrained strength obtained from other methods at the Kamojang site (Figure 9.9a) ranged from about 50 to 170 kPa, confirming the range indicated by the CPT tests. The effective strength parameters c and φ are surprisingly high for a soil of such fine-grained composition. This is perhaps to be expected in view of the observations of field behavior described earlier, especially in terraced rice fields. Figure 9.10 summarizes results of measurements of peak and residual strength of allophane clays. Peak values are from triaxial tests and residual values from ring shear tests, using the Bromhead device. Both values are remarkably high and there is surprisingly little difference between them. Rouse et al. (1986) have obtained similar high values from allophane soils in Dominica. Figure 9.11 shows values of the residual angle φ r plotted against plasticity index. It is seen that there is no relationship between the two; φ r does not steadily decrease with increase in plasticity index as is the case with most sedimentary clays. With PI values above about 80, normal sedimentary soils would be expected to have φ r values of around 10◦ , whereas the allophane clay has values between 30◦ and 40◦ . There seems to be an
Residual strength from ring shear tests: Indonesian samples New Zealand samples
500
°
Shear stress (kPa)
400
a,
s
300 m
th
200
cÄ
=
40
kP
lt
ia
x ria
t es
20
= fÄ
t
fro
ng
k
a Pe
tre
s
100
0
100
200 300 400 Normal effective stress (kPa)
500
600
Figure 9.10 Peak and residual effective strength parameters.
ALLOPHANE CLAYS
203
Residual friction angle fÄr (degrees)
50 40 30 20 10
Indonesia New Zealand 0
20
40
60
80
Plasticity Index
Figure 9.11 Residual strength angle versus plasticity index.
intrinsic property of the minerals allophane and immogolite that accounts for the remarkably high values of both peak and residual shear strength. Figure 9.10 above shows a cohesion intercept (c ) for the undisturbed samples of 20 kPa. It is possible to obtain estimates of the actual mobilized in situ values of c from back-analysis of slopes with known slope angle by making appropriate assumptions about the value of φ and the seepage state. This is done in Section 5.3.5 of Chapter 5 for Indonesian rice fields. Belloni and Morris (1991) have done this for slopes in Ecuador. They have assumed water tables at the surface and seepage parallel to the surface and obtained the c by back-analysis. Their results are shown in Figure 9.12. It is seen that once the slope inclination reaches about 35◦ the c value is in the range of about 15–20 kPa, depending on the assumption regarding the value of φ , the seepage condition, and the depth to the potential failure plane. The value of c obtained by analyzing terraced rice fields (Section 5.3.5 of Chapter 5) was from 8 to 15 kPa. It appears, therefore, that mobilized field values of c are not very different from the value obtained from conventional triaxial testing. 9.2.11
Compaction Behavior
An account is given in Chapter 11 of the unusual compaction characteristics of allophane clays: 1. The compaction curve when the soil is dried from its natural water content is different from that after the soil has been dried and rewetted. This is illustrated in Figure 11.2 of Chapter 11. The natural curve does not show a distinct peak dry density and optimum water content. 2. Allophane clays tend to be sensitive to some extent. Many are of low to moderate sensitivity, but some are of high sensitivity. When these
204
VOLCANIC SOILS
60 H=1m fÄ = 30
H=4m fÄ = 35
fÄ = 25
H=2m fÄ = 35
fÄ = 35
Slope angle (degrees)
H=2m fÄ = 25, 30, 35
40
SF = 1 No earthquake Slopes below groundwater table Slopes above groundwater table 20
0
20 Cohesion (kPa)
40
Figure 9.12 Values of c from back-analysis of natural slopes in Ecuador (after Belloni and Morris 1991).
clays are compacted at, or near, their natural water content they tend to become softer the more they are compacted. An example of this behavior is illustrated in Figure 11.3 of Chapter 11. It is most important therefore that compaction tests be carried out using proper procedures. In general, this means not drying the soil any drier than the water content at which each point on the compaction curve is to be established. It is not uncommon that difficulties are experienced in the control of compaction during earthworks operations involving allophane clays because compaction tests have involved excessive drying of the soil prior to testing. Oven and air drying can both produce substantial and irreversible changes to the properties of the soil. If excessive dying is allowed, then values of water content and dry density may be specified that the contractor will have no hope of achieving. The values will apply to air-dried or oven-dried soil, while the contractor is dealing with soil that has not been dried beyond its current water content. As described in Chapter 11, because of the high water content and sensitivity of many volcanic clays, it may be necessary to adopt a rather different approach to compaction from that normally used. By means of field trials, an optimum compaction procedure can be established, which involves a
VOLCANIC ASH CLAYS DERIVED FROM RHYOLITIC PARENT MATERIAL
205
relatively low compactive effort and careful control to ensure this level of compaction is not exceeded. If it is exceeded, the soil will become unacceptably soft, and remedial action will be necessary. Such remedial action normally involves excavating and replacing the material, either the same material after it has been dried, or new material at the desired water content. However, with volcanic ash soils this may not be necessary. Experience shows that these materials tend to harden up if left alone for a day or two. This is sometimes thought to be thixotropic behavior due to chemical changes. In the author’s view this is unlikely to be the case; it is more likely to be due to pore pressure changes. The overcompaction that causes the problem remolds the soil and releases water from the intact structure of the soil, with the result that pore pressure at or near the surface may be essentially zero or even slightly positive a short distance below the surface. By leaving the soil for a short time, these pore pressures may become much more negative, with a consequent increase in strength. 9.2.12
Engineering Projects Involving Allophane Clays
Allophane clays have been involved in a large number of engineering projects, including earth dams and similar water retaining structures, highway embankments, geothermal power stations, and large storage tanks. Some of these projects, such as the Cipanunjang dam in Indonesia, have been carried out without any knowledge or understanding of the unusual properties of the soil. Others have been carried out with some understanding of the clays. Wesley and Matuschka (1988) describe a number of such projects in some detail. If difficulties have arisen, it has usually been in the area of compaction procedures and compaction monitoring. The reasons for such difficulties have already been referred to above and are considered further in Chapter 11. Specific cases of compaction difficulties are described by Moore and Styles (1988) and Parton and Olsen (1980). 9.3 VOLCANIC ASH CLAYS DERIVED FROM RHYOLITIC PARENT MATERIAL
The allophane clays described in Section 9.2 are believed to be derived from the weathering of andesitic parent material, which is the predominant type of volcanic ash found in Indonesia and New Zealand, and indeed throughout the world. The production of enormous quantities of ash is a characteristic of andesitic eruptions. However, volcanic ash can also be produced by rhyolitic eruptions, and there is little or no information on soils derived from such eruptions, and, in particular, whether their properties differ from those from andesitic parent material. For this reason, some very limited information from the author’s own records is included here. There are three main areas of volcanic ash clays in the North Island of New Zealand. Two
206
VOLCANIC SOILS
of them, Taranaki and the central highland of the North Island, appear to consist of typical andesitic brown ash clay, and behave accordingly. The third, the Bay of Plenty, is well known for slope instability problems, and its volcanic ash soils seem to behave in a significantly different manner to those of the other two areas. In May 2005, an intense, relatively short-duration rainstorm caused numerous slips in residential areas of its principal city, Tauranga. A large number of houses were damaged beyond repair, and a further large number were sufficiently threatened by their proximity to new slips that they had to be demolished. The area where most of the damage occurred is a relatively flat plateau area bounded by old cliffs. Many of the slips occurred in the cliff slopes, but there were also a substantial number (possibly the majority) that occurred in the slopes of valleys that run inland from the old cliff line. In general terms, the volcanic ash sequence in the area of the major slips, from the surface downward, is as follows: 1. At the top are pale brown younger ashes, including a very sandy layer known as Rotoehu ash. This ash is found at an average depth of about 2.5 m, and is usually between about 0.5 and 1 m thick. 2. Immediately beneath the Rotoehu ash there is a layer of stiff to hard brown clay about 1 m thick, known as Hamilton ash. This is believed to be of andesitic origin. 3. Below the Hamilton ash there is a series of layers that are believed to be of rhyolitic origin. These consist of clays and silty clays of medium to stiff strength, varying in color from yellowish brown to creamy white. Interspersed in these layers are a number of harder dark brown layers similar to the Hamilton brown ash already described. These dark layers are identified by geologists as old ground surfaces (paleosols). The dark brown layers are generally of low sensitivity, while the pale-colored layers are of high to very high sensitivity. Figure 9.13 shows a cross section of one particular slip from which a series of samples was obtained for basic laboratory tests. Also shown is the sequence of layers described earlier. Samples were not obtained from the younger layers above the Rotoehu layer because of access difficulties. The slip had the appearance of a typical circular arc failure. Hardly any of the material that slipped remained on the site; it disintegrated and disappeared into the vegetation-covered marshy area below the site. There was little indication from the appearance of the failure surface that its shape had been significantly influenced by any particular soil layer. It seemed to cut through all the layers in its path. Figure 9.13 gives descriptions of the layers, the approximate boundaries between them, and the locations from which disturbed samples were taken. Laboratory tests included natural water content and Atterberg limits (natural and oven dried), on all the samples, particle size on six samples, and
VOLCANIC ASH CLAYS DERIVED FROM RHYOLITIC PARENT MATERIAL
207
10 m
8m
S1
Probable ground surface prior to slip of May, 2005 S2
6m
Sandy silt, loose, nonplastic, pale yellowish brown (Rotoehu ash)
Clay, stiff to hard, dark brown (Hamilton andesitic ash-Paleosol)
S3 S4
Clay, firm, high sensitivity, pale yellowish brown
S5 4m
S6 S7 S8
Clay, stiff to hard, dark brown ( Paleosol)
Silty clay, firm, extremely sensitive, pale yellowish brown Clay, stiff to hard, dark brown (Paleosol)
S9
2m
Clay, firm, high sensitivity, pale yellowish brown
S10
0
2m
4m
6m
8m
10 m
12 m
Rhyolitic ash layers
Disturbed samples
14 m
Figure 9.13 Cross section of a slip in rhyolitic volcanic ash soils.
measurements of undrained shear strength using a hand vane. Both undisturbed and remolded strength were measured. The results of the tests are summarized in Figure 9.14. The most striking features of the tests are the rapid changes in properties over short distances, reflecting the changes from the dark brown paleosols to the light-colored material, and the extremely high sensitivity of some of the latter (rhyolitic) samples. Samples S7 and S9 had natural water contents significantly above the liquid limit, giving them sensitivities around 100 and 70, respectively. The samples occupy the normal position of volcanic ash soils on the plasticity chart, as shown in Figure 9.15 and all undergo some loss of plasticity after oven drying. In this respect, there is no significant difference in behavior between the brown samples and the light-colored samples. Loss of plasticity is generally a good indicator of allophane content, at least in the author’s view. These tests do not suggest that the allophane content is significantly different between the two soil types. Particle size measurements were made on six of the samples; the results are given in Figure 9.16. Two are from the dark brown samples, three from the light-colored samples, and one from the Rotoehu ash sample. There is surprising similarity between the curves from the two soil groups. None of the samples showed any tendency to flocculate during the hydrometer tests, suggesting that the allophane content was not particularly high.
208
VOLCANIC SOILS
Natural water content and Atterberg Limits (%) 40
60
80
0 100
0
Liquidity Index 1 1.5 2.0 Sensitivity 20 40 60 80
0.5
2.5
Undrained shear strength (kPa)
100
50 100 150 200
S1 Sand, nonplastic
8
Liquid Limit Plastic Limit Natural water content
Undisturbed Remolded
Liquidity Index Sensitivity
Elevation (m)
S2 6
S3
S4 S5 4
S6 S7 S8 S9
2
S10
Figure 9.14 Natural water content, Atterberg limits, undrained shear strength, and sensitivity.
Plasticity Index
60
Dark brown samples Pale yellow samples
S2
Natural Oven dried
40
S5
S4 S3 S8
S6
S10
S9
20
S7
0
20
40
60 Liquid Limit
80
100
Figure 9.15 Atterberg limits on the plasticity chart.
120
209
VOLCANIC ASH CLAYS DERIVED FROM RHYOLITIC PARENT MATERIAL
100
Stiff, dark brown clays, low sensitivity
Percent Passing
80 Firm to stiff, pale yellowish brown clays, high sensitivity
60
40
0.001
Sample 1 2 4 7 8 9
Silty sand, nonplastic (Rotoehu ash)
20
2
4
7 0.01
2
4
7
0.1
2
4
7
1
Particle Size (mm)
Figure 9.16 Particle size curves from six samples.
To summarize, the following observations are made: •
•
•
•
•
The most interesting finding from this small investigation is that the material believed to be of rhyolitic origin is made up of alternating layers of soil with very different properties. These layers are either dark brown stiff to hard clay (paleosols), or light-colored medium to stiff clay (or silty clay) of high to very high sensitivity. The layers vary in thickness from less than 0.5 m to about 1 m. The presence of the these highly sensitive layers make this material rather different from most volcanic soils of andesitic origin and presumably accounts for its much less satisfactory performance in slopes. The explanation for the formation of the two quite different materials is uncertain. If the dark brown layers are old ground surfaces and originate from the same parent layer as the underlying light-colored layers, then it is probable that they have undergone more intense weathering and can perhaps be expected to have different properties. In contrast with the above characteristics, volcanic clays of andesitic origin are generally more uniform; it is usually possible to detect distinct layers in these materials from visual inspection, but these are not accompanied by any systematic variation in properties. The different weathering pattern may be because the higher quartz content of rhyolitic ash makes it more resistant to weathering than andesitic ash. However, this does not provide a clear explanation for the high sensitivity, although perhaps there are still enough quartz remnants in the soil to form a fragile and unstable skeleton that collapses when disturbed.
210
9.4
VOLCANIC SOILS
OTHER UNUSUAL CLAYS OF VOLCANIC ORIGIN
There are no doubt many clays of volcanic origin that do not belong in the group described above. Examples are lake deposits of volcanic origin. If lakes exist in the proximity of volcanoes then it is to be expected that some volcanic material will be eroded and redeposited in lakes. This can give rise to extremely weak and compressible soils. Strictly speaking these are no longer residual soils, though they may contain some of the clay minerals commonly found in residual volcanic soils. Apart from material eroded and deposited in lakes in this way, some airborne material may be deposited directly into lakes (or into the sea). The clay underlying Mexico City, with its unusual characteristics and very high compressibility, is an example of a clay belonging to this group. Another unusual soil that does not belong in any tidy category is that made up of diatoms, often referred to as diatomaceous silt. Diatoms are tiny living organisms, somewhat like minute coral formations, that grow in silica-rich streams that flow into lakes. They come in a large range of intricate shapes and sizes that can only be seen under a microscope. The resulting soil is likely to have a very high void ratio and to be highly compressible. Verdugo (2008) provides a good overview of allophane and diatomaceous soils, pointing out that although they generally exist separately from each other, their formation processes are such that they may still occur together, and he describes a number of situations where they are found together. Verdugo points out that the form and size of diatoms are extremely variable, there being many thousands of different species of diatom existing worldwide, including Mexico City mentioned above. Figure 9.17 illustrates a situation where the soils are likely to be found together in a lake environment. Material entering the lake (or formed in the lake) may consist of any of the following: 1. Relatively coarse pyroclastic material deposited directly at the time of eruptions, or flowing into the lake as lahars. 2. Fresh airborne ash falling into the lake shortly after the eruption. 3. Soil eroded from surrounding hillsides and carried by streams or rivers into the lake. This soil is likely to be rich in the clay minerals allophane, immogolite, and halloysite. 4. Diatomaceous silts, since the volcanic environment may provide the special conditions needed for their formation, namely an adequate supply of water rich in dissolved silica. An example of a clay formed in this way is the plateau area in West Java, Indonesia, immediately south of the city of Bandung. This low-lying area was a freshwater lake during the Quarternary period, and eruption materials
OTHER UNUSUAL CLAYS OF VOLCANIC ORIGIN
211
Airborne ash clouds resulting in direct air fall into lake
Periodic volcanic activity
Su an rfac dr e ive ero r t sio ra ns n, po rt
Volcanic alluvial deposits
Lake surface
, t on r si spo o er ran ce er t a Volcanic slopes of rf riv Su nd allophane clays a
Silica-rich waters flowing into the lake, resulting in the growth of diatoms
Figure 9.17 Environment in which allophane and diatomaceous deposits can be found together.
gradually accumulated in the lake, forming a soft, highly compressible clay, which at its deepest part is about 30 m thick. The geotechinical properties of this clay were investigated in considerable detail during planning for a highway; they have been described in papers by Younger (1988) and Younger et al. (1989, 1990). The lake deposit is of highly variable composition with a complex structure. Infrared, electron microscope, and X-ray diffraction examinations have identified the presence of halloysite, albite, and crystabolite, with minor amounts of montmorillonite. Significant quantities of diatomaceous silica were also found. Younger et al. (1990) report that although allophane was not identified in these tests, the water content of the soil and its plasticity characteristics suggest that allophane was also present, which would not be surprising considering the origin of the material. Figure 9.18 shows water content, Atterberg limits, and undrained shear strength, taken from Younger et al. (1990). The most significant aspects of the soil properties appear to be the following: 1. There is a very large range of natural water contents and Atterberg limits. 2. Much of the soil is highly sensitive, since the natural water content is often close to, or greater than, the liquid limit. 3. The undrained shear strength is also highly variable, generally less than 50 kPa in the top 10 m, but increasing somewhat below 10 m. The undrained shear strength is significantly higher than that of a normally consolidated sedimentary soil. Figure 9.19 shows the Atterberg limits plotted on the conventional plasticity chart. The range of values is very great indeed, with liquid limits from about 50 to over 300. The majority of the points lie below the A-line, but
212
VOLCANIC SOILS
Water content (%) and Atterberg Limits 0 100 200 300
0
Undrained shear strength (kPa) 50 100
150
Laboratory Field
2 4 6 8 10
Depth (m)
12 14 16 18 20 22 General range for normally consolidated sedimentary soils
24 26 28 30
PL
wn
LL
32
Figure 9.18 Natural water content, Atterberg limits, and undrained shear strength for Bandung clay (from Younger et al. 1990).
there are still a significant number that plot above it, especially those with liquid limits below 100. As part of investigations and planning for the highway, trial embankments were constructed. The more significant findings from these embankments were the following: 1. Settlement was relatively small until the embankment height exceeded about 2.5–3 m. 2. At greater heights, the settlement was approximately 30 percent of the height of the embankment. 3. The rate of consolidation was such that the most of the primary consolidation was complete in a three- to four-month period, and for this reason sand or wick drains were not used to accelerate consolidation, except for extra high fills at overpasses or underpasses.
PUMICEOUS MATERIALS
213
250
Plasticity Index
200
150
100
50
0
50
100
150
200 250 Liquid Limit
300
350
400
Figure 9.19 Atterberg limits from Bandung clay plotted on the plasticity chart (from Younger et al. 1990). 9.5
PUMICEOUS MATERIALS
Pumice materials are not strictly residual soils. However, they are volcanic and have unusual properties and a brief description of some aspects of their behavior is included here. Pumiceous materials are a product of rhyolitic eruptions, and consist predominantly of quartz. They are characterized by the vesicular nature of their particles; each particle contains a dense network of fine holes, some of which are interconnected and open to the surface, while others appear to be entirely isolated inside the particles. The result is that the particles are lightweight, have very rough surfaces, and are easily crushed, especially when compared to more “normal” hard-grained sands, such as quartz sand. The behavior of pumiceous sands described in the next section should serve as a warning regarding the application of standard property correlations to volcanic soils. 9.5.1
Pumice Sands
Pumice sands are found in various parts of New Zealand and in other parts of the world. They are not infrequently encountered in engineering projects and their evaluation is a matter of considerable interest to geotechnical engineers. Some recent research by the geotechnical group at Auckland University has shown a particularly unusual and important aspect of their behavior. The principal objective of the research was to correlate CPT cone resistance with the relative density of the sand. Two large bulk samples of sand were used in this research. One was pumice sand and the other was a hard-grained sand consisting predominantly of quartz. The latter was included in the study for comparison
214
VOLCANIC SOILS
purposes. Both sands were river deposits, and were dredged from the river for commercial purposes. A series of standard laboratory tests was first carried out. Figure 9.20 shows particle size curves and results of conventional oedometer tests on samples in their loose and dense states. The oedometer tests highlight the large difference in compressibility of the two sands: the pumice sand is about five times more compressible than the quartz sand. A series of drained triaxial tests was carried out on both sands, the results of which are summarized in Figure 9.21. Figures 9.21a and b show graphs of deviator stress and volume change versus strain at low and high confining pressures, respectively. The quartz sand conforms to conventional behavior, with the dense sample showing peak values of deviator stress at both confining pressures. However, the pumice sand in the dense state shows a clear peak deviator stress at the low confining stress, but no such peak at the high confining stress. In the loose state, none of the samples show distinct peak values of deviator stress, although the quartz sand shows a very flat peak at the low confining stress. Figure 9.21c shows the Mohr-Coulomb failure lines for both materials. It is seen that the friction angle (φ ) of the pumice sand is about 41.5◦ , whether the sample is loose or dense. The friction angle of the quartz sand is 36.5◦ and 41◦ in the loose and dense states, respectively. To investigate behavior during CPT tests, a large calibration chamber was used. The chamber was approximately 0.76 m in diameter and 0.85 m in height. Tests were carried out on both sands in their dense and loose
Stress (kPa) 1000
san
d
100
Sieve size mm mm 75 150 300 600 1.18 2.36
ice Pum
Dense
Compression (%)
40
MI
CE
Loose
Dense
10
Loose
20
0 0.06 0.1
2000
QUARTZITIC
PU
Quartz sa
Percentage finer
nd
80
60
0
20 1 Particle size mm (a) Particle size curves
6 (b) Standard oedometer tests
Figure 9.20 Particle size curves and standard oedometer test results.
PUMICEOUS MATERIALS
300
1500 Confining pressure = 300 kPa
Deviator stress (kPa)
Deviator stress (kPa)
Confining pressure = 50 kPa
200
100
1000
500
Quartzitic (dense) ” (loose) Pumice (dense) ” (loose) 0
10
Quartzitic (dense) ” (loose) Pumice (dense) ” (loose)
20
30
0
10
Strain (%) 10
20
5
Volume change (%)
10
30 Volume change (%)
0
20
30
20
30
Strain (%)
Strain (%) 7
215
0
0
10
0
−10 −3 (a)Tests at low confining pressure (50 kPa)
(b) Tests at high confining pressure (300 kPa) 1200
1200 Dense Loose
Dense Loose .5°
41
°
800 = φ′
2
σ′1 − σ′3
= φ′
2
σ′1 − σ′3
800
φ′
41
°
6.5
=3
400
400
Pumice 0
400
800
1200
1600
Quartzic 0
σ′1 + σ′3 2
400
800
1200
1600
σ′1 + σ′3 (c) Mohr Coulomb failure lines
2
Figure 9.21 Summary of results of drained triaxial tests on both sands.
states, using applied vertical stresses of 50, 100, and 200 kPa. The results are summarized in Figures 9.22 and 9.23. Figure 9.22 shows typical results for the two sands, in their loose and dense states, at a vertical confining stress of 200 kPa. The results from the pumice sand are startling as they show no significant difference in behavior between loose and dense samples. Similar results were obtained at other confining pressures. The quartzitic sand behaved more or less as expected, with a large difference between the loose and dense states. In Figure 9.23a the cone resistance curves from
216
VOLCANIC SOILS
Cone resistance (MPa)
Cone resistance (MPa) 0
2
4
8 0
6
5
10
15
20
25
Dr = 91.5% sÄv = 200 kPa Dr = 96.5% sÄv = 200 kPa
Depth (m)
0.2
0.4 Dr = 6.8% sÄv = 200 kPa
Dr = 4.2% sÄv = 200 kPa
0.6
Pumice sand
Quartzitic sand
0.8
Figure 9.22 Typical results of cone penetrometer tests at confining pressure = 200 kPa.
Figure 9.22 are shown on the same graph so that a direct comparison can be made between the behavior of the two sands. Figure 9.23b summarizes all the test results in the form of the familiar graph of cone resistance (qc ) versus vertical stress. Cone resistance (MPa) 0
5
10
15
20
25 Cone resistance (MPa) 0
Pumice, dense Quartzitic, loose
0.6
Effective vertical stress (kPa)
Depth (m)
Quartzitic, dense
0.4
10
15
20
Pumice, dense Pumice, loose
Pumice, loose
0.2
5
Quartzitic, dense Quartzitic, loose
50
100
150
200 0.8 (a) Results from both sands at 200 kPa confining stress
(b) Summary of results of all tests
Figure 9.23 Summary of all the test results.
PUMICEOUS MATERIALS
217
These figures all illustrate the surprising and dramatic difference in behavior between the two sands. The quartz sand behaves as expected, showing large differences in cone resistance between the loose and dense states, and steadily increasing values with confining stress. The pumice sand, on the other hand, shows the following rather extraordinary characteristics: 1. There is very little change in cone resistance between its loose and dense states, and the increase in resistance with confining stress is less pronounced than with the quartz sand. 2. The penetration resistance of the pumice sand is a little higher than that of the quartz sand in the loose state. This dramatic difference in behavior can only be attributed to the difference in particle strength of the two sands, and the consequent difference in behavior during shearing demonstrated by the drained triaxial tests (Figure 9.21). These show that with the pumice sand, in both the dense and loose state, the deviator stress slowly climbs toward a peak value, but there is no post peak decline in strength. The strain to reach the peak value is large, generally between 20 and 30 percent. This peak value is essentially the same regardless of whether the sand is initially in the dense or loose state. It may therefore be the case that in the cone test, the state of the sand when failure occurs at the cone point is the same regardless of whether its initial condition was loose or dense. For a more detailed account of the behavior of this pumice sand see Pender et al. (2006) and Wesley (2006). It will be apparent to the reader that this research did not achieve its intended objective, namely a correlation between its relative density and cone resistance. Despite this, it was certainly informative as a warning about the limitations of the cone test in such materials. It appears that to develop a correlation for the pumice sand, the parameter measured will need to relate to the compressibility of the sand rather than its strength, because while there is no significant difference in the ultimate strength in the loose and dense state, there is a clear difference in their compressibility (Figure 9.20). 9.5.2
Pumiceous Silts and Gravels
Mixed deposits of pumiceous material cover a large part of the central North Island volcanic plateau and are frequently encountered and utilized during road construction. This may be, in general, earthworks operations, or as the subgrade or base course of the road pavement itself. Particular characteristics of these materials during such works include the following: 1. Slopes cut into pumiceous soils remain stable at remarkably steep angles. In very coarse pumicious gravels cuts can be made to considerable height with almost vertical faces. There are two explanations for this. The first is that the pumice particles have high frictional
218
VOLCANIC SOILS
characteristics and may be of very angular shape, in which case the particles tend to interlock, creating a small but significant cohesive component of shear strength. 2. Pumiceous materials tend to be of high permeability, so that only in very intense rainstorms will there be any possibility of positive pore pressures. With the finer materials, which will be partially saturated, there is probably a component of strength coming from the negative pore pressure above the water table. 3. Pumiceous gravels generally make good base course material for pavement construction. Many forest roads made by timber companies are made entirely of pumice gravel, without sealed surfaces. When compacted, either by systematic rolling or the passage of timber vehicles, the particles undergo a certain amount of crushing and become more tightly packed with the passing of time. 4. A characteristic of some of these pumice materials is a tendency to become very soft and display substantial weaving during compaction, especially in nondrying weather conditions (which are not infrequent in New Zealand). However, the material stiffens up or hardens quite rapidly if left untouched overnight or for a day or two. This last characteristic is sometimes thought to be a thixotropic effect, although this seems unlikely in view of the inert nature of pumice particles. To investigate whether this effect was thixotropic or simply a pore pressure change, the following investigation was carried out. Two pumice materials were tested, one a silt and the other a gravel. Particle size and compaction test results are shown in Figure 9.24. A number of compacted samples of each material were then prepared by compacting them at identical water contents. The water content was about 2 percent wet of optimum in each case, this value being chosen so as to produce samples of relatively low strength, comparable with those often produced during compaction in the field. The samples were compacted in brass tubes. Strength measurements were made immediately on two samples of each material to determine the initial, or “as-compacted” strength. The strength measurements were made using unconfined compression tests for the gravel, and undrained triaxial tests (without a confining stress) for the silt. One set of samples was then sealed by carefully waxing the tubes, and stored for testing later at specific time intervals. In this way, any changes in strength in these samples could result only from chemical effects. The second set of samples was used to investigate changes in pore pressure. With the silt samples this was done by setting up the samples in a traxial cell and connecting the porous stone at the base of the sample to a suction pump via a water/air interface. This induced negative pore water pressure (pore water tension) within the samples. Tests were done at pore water tensions from 25 to nearly 100 kPa, this being the maximum value feasible with this
219
PUMICEOUS MATERIALS
1.4 ro
Ze
100
Dry density (gm/cm3)
el rav ice g
45 2.
Pum
= G. S.
Pum iceo us s ilt
e lin
Percent finer
ids
40
Pumiceous silt
vo
60
1.2
air
80
Ze
1.0
ro
air
vo
ids
lin
e
S.
G.
0.8
20
=
1.
48
Pumiceous gravel 0.001
5 0.01
5 0.1 5 1 Particle size (mm)
5 10 0.6 10
30
50
70
Water content (%)
Figure 9.24 Particle size and compaction curves of the two pumiceous materials.
technique. Ignoring a small correction because of incomplete saturation, this produced effective stresses in the samples equal to the applied suction. This procedure was used as it was considered to most closely replicate the situation in the field. With a low water table, as is often the case in well-drained pumiceous areas, the material compacted at the surface will be subjected to a negative pore water pressure governed by the depth of the water table. To continue the investigation to higher stress levels, several more tests were carried out using cell pressure in the usual way to consolidate the samples. The compacted gravel samples were very coarse, open materials, containing considerable air, and the technique used with the silt samples was not considered appropriate for the gravel samples. Instead, these samples were simply extruded and left open to the air, so that evaporation took place and presumably induced increasing pore water tension within the gravel. Unconfined strength measurements were then made when weight measurements indicated moisture content reductions at 2 percent intervals. The two sets of sealed samples of both materials were opened at specific time intervals and tested immediately after opening to ensure no loss of moisture. The results of the tests are shown in Figure 9.25 in the form of graphs of strength versus time and strength versus either pore water tension (for the silt) or water content reduction (for the gravel). The results clearly show that strength changes occur only as a result of change in pore water tension or water content. The sealed samples tested at time intervals of up to 60 days show no significant or consistent change in strength with time. Hence, the hardening characteristic is clearly a purely
VOLCANIC SOILS
Unconfined compressive strength (kPa)
220
400
Water content decrease (%) 5 10
0
15
nt)
Pumice gravel
ing
200
en
nte
co
w
g an
ch
s(
le mp
r ate
sa
Op
Sealed samples t) (constant water conten
0
40
20
60
70
Time (days)
Unconfined compressive strength (kPa)
Effective stress (kPa) 0 400
60
120
Pumiceous silt ive
ng
ngi
ect
ubj
s les
ha to c
ct effe
180 ess
str
mp
200
Sa
Sealed samples (constant water content) 0
20
40 Time (days)
60
70
Figure 9.25 Strength changes versus time and pore water tension in the pumiceous gravel and silt.
physical one resulting from changes in effective stress and not from chemical processes. In practice, it is likely that quite high pore water tension develops in both the silt and the gravel when they are left alone after compaction. The silt is frequently a subgrade material lying above a deep water table, so that the pore water tension will be approximately equal to the depth of the water table. The gravel, on the other hand, is often used on forest roads as the main pavement material, and is thus open to the atmosphere. Hence, evaporation can occur, which would be expected to create high pore water tension within the material.
REFERENCES
221
REFERENCES Belloni, L., and D. Morris. 1991. Earthquake induced shallow slides in volcanic debris soils. Geotechnique 41(4): 539–551. Frost, R. J. 1967. Importance of correct pre-testing preparation of some tropical soils. Proc. First Southeast Asian Regional Conf. on Soil Engineering, Bangkok, 44–53. Gidigasu, M. D., and S. K. Bani. 1972. Geotechnical characteristics of troublesome lateritic materials. Proc. 8th Int. Conf. on Soil Mechanics and Foundation Engineering, Moscow, Vol. 4, 89–96. Jacquet, D. 1990. Sensitivity to remoulding of some volcanic ash soils in New Zealand. Engineering Geology 28(1): 1–25. Lohnes, R. A. and Tuncer, E. R. (1977) Engineering characteristics of andosols. Proc. 5th Southeast Asian Conf. on Soil Engineering, Bangkok, 305–313. Moore, P. J., and J. R. Styles. 1988. Some characteristics of volcanic ash soil. Proc. 2nd Int. Conf. on Geomechanics in Tropical Soils, Singapore, 161–166. Morin, W. J., and P. C. Todor. 1975. Laterite and Lateritic Soils and Other Problem Soils of the Tropics. Baltimore, MD: USAID Lyon Associates. Parton, I. M., and A. J. Olsen, 1980. Properties of Bay of Plenty volcanic soils. Proc. 3rd Australia New Zealand Conference on Geomechanics, Wellington, Vol. 1, 165–169. Pender, M. J., L. D. Wesley, T. J. Larkin, and S. Pranjoto. 2006. Geotechnical properties of a pumice sand. Soils and Foundations 46(1): 69–81. Rouse, W. C., A. J. Reading, and R. P. D. Walsh. 1986. Volcanic soil properties in Dominica, West Indies. Engineering Geology 23: 1–18. Terzaghi, K. 1958. Design and performance of the Sasamua Dam. Proc. Inst. of Civil Engineers 9: 369–393. Uehara, G. 1982. Soil science for the tropics. ASCE Geotech. Eng. Div. Specialty Conf. on Engineering and Construction in Tropical and Residual Soils, Hawaii, 13–29. Verdugo, R. 2008. Singularities of geotechnical properties of complex soils in seismic regions. ASCE, Journal of Geotechnical and Geoenvironmental Engineering 134(7): 982–991. Wada, K. 1989. Allophane and imogolite. Chapter 21 of Minerals in Soil Environments, 2nd ed. SSSA Book Series No 1: 1051–1087. Wallace, K. B. 1973. Structural behaviour of residual soils of the continually wet highlands of Papua New Guinea. Geotechnique 23(2): 203–218. Wesley, L. D. 1973. Some basic engineering properties of halloysite and allophane clays in Java, Indonesia. Geotechnique 23(4): 471–494. Wesley, L. D. 1974. Tjipanundjang Dam in West Java, Indonesia. Journal of the Geotechnical Division ASCE 100/GT5: 503–522. Wesley, L. D. 1977. Shear strength properties of halloysite and allophane clays in Java, Indonesia. Geotechnique 27(2): 125–136. Wesley, L. D. 1998. Some lessons from geotechnical engineering in volcanic soils. Proc. Int. Symposium on Problematic Soils, October 1998, Vol. 1, 851–863. Sendai, Japan: Balkema.
222
VOLCANIC SOILS
Wesley, L. D. 2001. Consolidation behaviour of allophane clays. Geotechnique 51(10): 901–904. Wesley, L. D. 2006. Geotechnical characteristics of a pumice sand. Proceedings of the 2nd International Workshop on Characterisation and Engineering Properties of a Pumice Sand, Singapore, Vol. 4, 2449–2473. Wesley, L. D., and T. Matuschka. 1988. Geotechnical engineering in volcanic ash soils. Proc. 2nd Int. Conf. on Geomechanics in Tropical Soils, Singapore, December 1988, Vol. 1, 333–340. Younger, J. S. 1988. Natural and lime stabilised properties of Bandung clay. Proceedings 2nd International Conference on Geomechanics in Tropical Soils, Singapore, December 1988. Younger, J. S., J. Rijanto, and C. Setjadiningrat. 1989. Bandung clay: characteristics and response under trial embankment loadings. Proceedings 12th International Conference on Soil Mechanics and Foundation Engineering, Vol. 3, 1773–1777. Younger, J. S., A. Jayaputra, and A. Rachlan. 1990. Characteristics of Bandung clay and performance under embankment loading: a review. Proceedings, Konperensi Geoteknik, Indonesia.
CHAPTER 10
RESIDUAL SOILS NOT DERIVED FROM VOLCANIC MATERIAL
10.1
INTRODUCTION
As indicated earlier, a comprehensive description of specific soil types is not the aim of this book. However, some general comments on soils derived from specific parent materials and, in particular, weathering environments are made in the hope that they will be a useful starting point for understanding these materials. Figure 10.1 is an attempt to group residual soils according to their parent material and weathering environment, and to indicate the expected characteristics of each group. This grouping is a broad-brush approach, as also are the comments and descriptions given in the following sections for each group. Before making these comments, we should remember the comment made earlier that apart from the last group (group 6) in Figure 10.1, the geotechnical properties of residual soils are generally good. Apart from this last group, the principal concern of geotechnical engineers is usually the stability of slopes. Foundation design and performance is relatively trouble free. In contrast, the major concern with the group 6 materials is foundation design and performance. Slope stability can also be a challenge in these materials but is normally of secondary concern to foundation performance. 10.2
WEATHERED GRANITE (GROUP 1 IN FIGURE 10.1)
Granite rock contains a high proportion of silica, which makes it rather resistant to weathering. For this reason most residual soils derived from 223
224
RESIDUAL SOILS NOT DERIVED FROM VOLCANIC MATERIAL
Parent Rock
Nature of Weathering Process
Characteristics of the Resulting Soil
(1) Granite
Breaks up the rock and converts part of it to clay minerals of low activity.
Deep weathering profile (“Little”), with a gradual transition zone from rock to soil. Normally a low clay fraction and considerable silt and sand-sized quartz particles.
(2) Basalt or andesite
Almost complete conversion of the rock into clay or silt. Clay minerals of low activity, normally halloysite and kaolinite. Often involves solution of cementing material and release of existing clay minerals or other particles, rather than conversion of parent material into new clay minerals.
Normally a sharp transition from fresh rock to clay. Unlike weathered granite the clay contains little evidence of its parent rock.
(3) Soft sedimentary rocks
(4) Hard sedimentary rocks
Boundary between soil and unweathered rock likely to be abrupt without a gradual transition. Weathering process likely to be similar to that in igneous rocks, with a deep transition zone and gradually less weathering.
Likely to depend on the nature of the clay minerals (if already present) released by the weathering process. If parent material is shale or argillite, the resulting soil is likely to have undesirable engineering properties because of presence of montmorillinite-type clay minerals. Soil less likely to reflect properties of the original sedimentary soil than in the case of soft rocks.
(5) Rocks found in well drained areas in a hot wet tropical environment
Weathering process tends to be laterization — involving solution of silica and gradual increase in aluminum and iron compounds. Increasing iron concentration results in the red color.
Initially, the soil may be simply a red clay. With age, the plasticity tends to decrease, and the iron aluminum compounds act as cementing agents. Ultimately the cemented material takes the form of hard “concretions,” and the soil becomes a clayey gravel.
(6) Rocks found in poorly drained areas in warm climates having distinct wet and dry seasons
Weathering tends to produce highly active clay minerals of the smectite group, especially montmorillinite.
Soil is likely to belong to the “black cotton” group — dark grey to black with highly undesirable properties. Source of expansive clays, and many shrink and swell problems.
Figure 10.1 An overview of some dominant residual soil groups.
granite tend to be rather coarse and of low to moderate plasticity. They are likely to contain a significant proportion of unweathered quartz particles in the silt and sand range. Some characteristics of the weathered granite soils of Hong Kong have already been described in Chapter 9 (Slope Stability), in particular their coarseness and the likelihood that some, if not most, of
225
WEATHERED GRANITE (GROUP 1 IN FIGURE 10.1)
Degree of saturation 0 100
0.2
0.4
0.6
0.8
1.0
2
Malaysian soils Hong Kong soils
40
rse
fine
ec oa
me
re Ext
Depth (m)
4 60 ne
ge fi
ra Ave
an
Me
20
Av er ag
Percent finer
80
8
e
oars me c
Extre
0
Water table
6
0.002 0.006 0.02 0.06
0.2
0.6
Particle size (mm) (a) Particle size distribution
10 2
6
20
No water table
12 (b) Degree of saturation
Figure 10.2 Some properties of Hong Kong and Malaysian soils derived from granite (after Lumb 1962 and Shukri et al. 2004).
them are partially saturated above the water table. Figure 10.2 illustrates measurements of particle size and degree of saturation from various sites in Hong Kong, and additional particle size curves are shown for residual granite soils in Malaysia. The data on the Hong Kong soils are taken from Lumb (1962) and that on the Malaysian soils from Shukri et al. (2004). The particle size curves in Figure 10.2a illustrate an important point about the properties of weathered granite soils. The average clay fraction of the Hong Kong soils is about 6 percent, while that of the Malaysian soils is about 40 percent. If the data in Figure 10.2 are truly representative of the soils in these two places, then it shows that while the weathered granites of Malaysia can be expected to behave as clays, those in Hong Kong are more likely to behave as silty or clayey sands. Observed field behavior suggests the Hong Kong soils behave rather like weakly cemented sands. Lumb (1962), in describing the weathering process in Hong Kong, notes that the more intense weathering close to the surface “produces a dense surface layer of lower permeability than the underlying soil, thus limiting the infiltration of rain water.” He goes on to state that the underlying soil is “free draining due to its open structure and coarse grading and the limitations on infiltration results in the soil being unsaturated where no permanent water table exists, such as on steep slopes.” Figure 10.2b shows measurements of degrees of saturation at two sites, one with a very deep water table and the other with a water table at 6 m. At the deep water table the degree of saturation varies between about 40 and 60 percent, while at the 6-m-deep water table the degree of saturation is between 80 and 100 percent. The observations of Lumb and the data he presents on degrees of saturation suggest that the response of Hong Kong weathered granites to rainfall is likely to be as partially saturated, rigid materials that undergo changes in
226
RESIDUAL SOILS NOT DERIVED FROM VOLCANIC MATERIAL
degree of saturation with negligible volume change. In contrast, the clays of Malaysia are likely to remain fully saturated except for a relatively shallow zone at the surface, and respond to rainfall as clays, with volume changes occurring as the pore pressure changes. With respect to the selection of design slopes in weathered granites, it is often the case, at least in deep cuts, that the most economical slope will be one that is steepest near the toe and becomes significantly flatter over the upper part, as illustrated in Figure 8.23 of Chapter 8. Such a profile takes account of the decreasing degree of weathering with depth, and accompanying increase in strength. One property of weathered granites that does not appear to have received the attention it deserves in the literature is their susceptibility to surface erosion. Cut slopes in weathered granites in Malaysia suffer severe erosion during the intense rainstorms to which they are frequently subjected. On the highway between the capital Kuala Lumpur and a town called Karak, the erosion channels on some of the slopes a few years after completion of construction were so deep a human being could not climb into or out of them without the aid or a ladder! The factors controlling susceptibility to erosion and the means of controlling erosion are not well established and could well benefit from some detailed research programs. 10.3
WEATHERED SEDIMENTARY ROCKS
10.3.1 Soft Rocks—Sandstones, Mudstones, and Shale (Group 3 in Figure 10.1)
The clays produced from soft sedimentary rocks are a rather mixed group, quite different from group 1 or group 2 clays. Soft sedimentary rocks tend to lie across the boundary between hard clay and rock, and their weathering process reflects this. In many cases the weathering process is one that releases existing clay minerals rather than creates new ones. In other cases, especially in sandstones that are made up of interbedded layers of sandstone and mudstone, the sandstone may possibly undergo weathering that produces new clay minerals, while the mudstone releases existing clay minerals. The weathering of shale appears to be a case of release of existing clay minerals. It is debatable whether clays or this sort should really be classed as residual soils, since their properties are more likely to reflect their former lives as sedimentary soil than the weathering process that forms them. Weathered shales are an example of this; they are likely to be have undesirable characteristics inherited from their previous existence, namely the presence of a highly active clay minerals belonging to the montmorillonite group. In addition to the nature of the parent rock, climate also appears to have a significant influence on soils derived from sandstones. This is evident from the differences in properties of these soils in New Zealand and
WEATHERED SEDIMENTARY ROCKS
227
Malaysia. The weathered Waitamata sandstones of Auckland, New Zealand (see Sections 4.3.3 and 5.3.5) clearly retain much of the structure and composition of their parent rock. The soil consists of distinct layers, some of which are very high-plasticity clays and some are sandy silts. The depth of weathering is not particularly deep, generally less than about 15 m, with a reasonably sharp transition from rock to soil. Cliffs that make up much of the coastline in the area consist of near vertical faces for heights of up to 50 m or more with only 5 m to about 15 m of soil at the top. Auckland has a temperate climate with moderate temperatures so weathering is not very intense. In contrast to this, weathering in the wet tropics is much more intense and soils are likely to retain fewer of the characteristics of their parent material. An example of this behavior is the soil found along a new highway built in the 1970s to connect the towns of Kuala Krai and Gua Musang in Malaysia (with which the author had some involvement). The highway runs through steep hilly terrain and involved a large number of deep cuttings. Over much of the route, the soil exposed in these cuttings was of sandstone origin, and the thickness and orientation of the original layers was still clearly evident, especially from changes in color. Despite this, there was not a great variation in the characteristics of soil, which was generally of moderate to high plasticity. Very few of the cuttings reached the parent rock. This illustrates an important difference between highway construction in temperate and tropical climates when deep cuttings are involved. In temperate climates, with shallow weathering, deep cuts are likely to extend well into unweathered rock, whereas in the wet tropics, with deep weathering, it may well be that few cuts reach the parent rock.
10.3.2
Hard Sedimentary Rocks (Group 4 in Figure 10.1)
The term hard sedimentary rocks is used here to denote sedimentary deposits that have undergone induration (or lithification) to the extent that they have essentially lost all trace of the clay or sand from which they were formed. They can therefore be expected to weather in a similar pattern to hard igneous rocks such as granite, and show a gradual transition pattern from fresh rock to soil. The greywacke rock found widely in the North Island of New Zealand is an example of a rock belonging to this category. The rock itself is a hard durable rock that is widely used locally as base course for highway pavement construction, and as concrete aggregate. As mentioned in Section 2.2 (Chapter 2), Pender (1971) proposed a slightly amended version of the Little classification profile for use with this weathered greywacke. Pender (1980) also showed that the shear strength parameters c and φ steadily decline with increasing void ratio (reflecting increasing degree of weathering), as illustrated in Figure 10.3. The Mohr-Coulomb failure lines are best-fit lines from a series of triaxial tests. Each line represents results
228
RESIDUAL SOILS NOT DERIVED FROM VOLCANIC MATERIAL
Void ratio 35
3
0.3
Fr
5 0.4 .55
Increased weathering, increase in void ratio
ict
0 5 0.6 5 0.7 .85
200
0
100
0
100
200
300
ion
an
gle
Cohesion intercept (kPa)
Shear stress (kPa)
300
400
30 60
Coh
esio
50
n in
terc
Friction angle (degrees)
400
25
ept
40 30 0.3
0.4
Normal stress (kPa)
0.5
0.6
0.7
0.8
0.9
Void ratio
Figure 10.3 Influence of void ratio on shear strength parameters in weathered greywacke (data from Pender 1980).
of a large number of tests, ranging from 7 to 60 over the six categories. The increasing void ratio presumably reflects greater weathering. The graphs of cohesion intercept (c ) and friction angle (φ ) against the void ratio show a well-defined decline in the friction angle, but a much less well-defined decline in the cohesion intercept. This illustrates the difficulty of accurately defining c values compared to φ values. As stated above, hard sedimentary rocks can be expected to weather in a manner similar to igneous rocks, provided weathering takes place in well-drained topography. Clays produced by the weathering process can be expected to be generally unremarkable and well behaved. A word of warning, however, is necessary here, especially in the light of the greywacke rock in New Zealand. Within the greywacke formation it is not uncommon to find zones of argillite. These presumably originate from the clay layers within the sandstone from which the greywacke is formed. Argillite is a much softer rock than the surrounding greywacke, and appears to still contains active clay minerals. It is not a suitable material for pavement construction or concrete aggregate, and weathers to clays with much less favorable properties than those coming from the greywacke. 10.4 LATERITES AND TROPICAL RED CLAYS (GROUP 5 IN FIGURE 10.1)
The terms laterite or lateritic soil are used rather loosely in soil mechanics and geotechnical literature. Lateritization or lateritic refers to a weathering process whereby silica-rich material is slowly removed by solution leaving behind compounds of iron and aluminum known as sesquioxides or
LATERITES AND TROPICAL RED CLAYS (GROUP 5 IN FIGURE 10.1)
229
hydrated oxides. Although it is more commonly associated with residual soils, lateritization also occurs in sedimentary soils, especially shales and soft sandstones or mudstones. Because it involves a steadily increasing concentration of hydrated iron oxide, the process turns soils a red or reddish brown color. Many such soils are loosely termed lateritic soils, and may even be called laterites. This terminology is not very accurate. The term laterite refers to the end product, when the sesquioxides become concentrated and act as cementing agents that bind the material together to form “nodules” the size of coarse sand or fine gravel. The term lateritic is best used to denote material that is starting to show signs of this cementing action. Many red clays are called lateritic clays,
Reddish brown clay
Red clay
2
6
Reddish brown clay
Depth (m)
4
8 w PL LL
10 0
40
80
120
160 40
Water content and Atterberg Limits
60
80 100
Vane Undrained triaxial 60
Clay fraction %
80 100 120 140 90 Undrained shear strength (kPa)
100
110
Degree of saturation %
300
sÄ1 − sÄ3 (kPa) 2
Consolidated undrained triaxial tests
200
100
fÄ = 37° cÄ = 14 kPa 0
100
200
300
400
sÄ1 + sÄ3 (kPa) 2
Figure 10.4 Some typical properties of tropical red clays of volcanic origin (Java, Indonesia).
230
RESIDUAL SOILS NOT DERIVED FROM VOLCANIC MATERIAL
which is correct in the sense that the weathering process to which they are subject is a lateritization process, but the soils may not have any properties that identify them as lateritic. Such clays are best called red clays rather than lateritic clays. It is probably only in the top several meters that the weathering process actually produces true laterites. The tropical red clays of Java, Indonesia, are a good example of red clays that have not progressed along the weathering sequence to the stage that their properties are significantly affected by the lateritization process. These clays have been derived from volcanic materials and so should perhaps have been included in the previous chapter on volcanic soils. They are included here because they belong in the lateritization group, which includes both volcanic and other parent materials. Typical properties of the Java red clays are illustrated in Figure 10.4. These clays are almost ideal materials for geotechnical construction purposes despite the fact that they are of medium to high plasticity, and plot rather close to the A-line on the plasticity chart. They are of low compressibility and high shear strength and not subject to large volume changes on drying or wetting. Also, their natural water content is close to the plastic limit, so that earthworks can be undertaken without concerns about drying the soil. They do have one property that is less than ideal, which is that they are rather sticky materials. Trimming samples for laboratory testing can be difficult because the soil clings to knives or trimming saws. Samples are best formed by taking samples in the field with the diameter required for laboratory tests. 10.5
BLACK OR BLACK COTTON CLAYS
These are a distinctive group of residual soils that display highly undesirable engineering properties. As indicated in Chapter 1, they are generally formed in relatively flat, low-lying areas, with a wet and dry seasonal climate. They contain active clay minerals, such as montmorillonite, that cause large volume changes in the soil as a result of changes in water content. This gives rise to their expansive (swelling) or contractive (shrinking) behavior, which is a source of major damage to buildings, especially light buildings on shallow foundations. The term expansive clays is widely used in soil mechanics literature to describe soils that swell, and the term shrinking clays is used to a lesser extent for soils that shrink. Contractive clays would be a more logical term for the latter, as they are the reverse of expansive clays, but this term is not often used. With respect to both expansive and shrinking clays, it is important to recognize that such clays are not so much a type of soil as a climatic condition in which a particular type of soil is found. The actual soil involved in swelling and shrinking phenomena may be identical; it just happens to exist in different climates. An appropriate definition of an expansive clay is as follows:
BLACK OR BLACK COTTON CLAYS
231
An expansive clay is one containing a high proportion of active clay minerals and that normally exists in a compressed state due to the presence of high negative pore water pressure (soil suction).
Similarly, a contractive clay can be defined as follows: A contractive clay is one containing a high proportion of active clay minerals and that exists in a wet or damp climate, which from time to time suffers prolonged hot dry seasons.
It is therefore not the case that expansive clays can be identified simply by conducting tests on the clay alone. It is not uncommonly implied in the literature that by conducting index tests, especially Atterberg limits and/or clay fraction, we can identify expansive clays. All we can identify by such tests are soils that have the potential to behave as expansive clays or shrinking (contractive) clays, depending on the climate in which they exist. Clays with high liquid limits that plot well above the A-line on the plasticity chart normally have the potential to be either expansive or contractive. In relatively dry climates, such as Australia, the Middle East, South Africa, and parts of the United States, they are likely to be expansive clays, while in temperate or wet climates they are likely to be contractive clays. The problem of expansion occurs when the existing pore pressure state is disturbed in some way. This most frequently occurs as a result of human activity. Irrigation or leaking water pipes make water available to the soil, which soaks it up and swells. It can also occur simply as a result of covering the surface. By building a house or road, or paving or sealing the surface, the normal evaporation at the surface is cut off, with the result that the soil takes in water and expands. The problem of shrinkage occurs when soils that normally exist in a relatively wet or damp climate are subject to unusually hot, dry weather. The soil at the perimeter of buildings suffers greater moisture loss than the soil beneath the center, causing differential settlement between the perimeter and internal foundations, and consequent damage to the building. This occurs from time to time with the weathered Waitemata clays of Auckland, New Zealand (described in Section 4.3.3 of Chapter 3), which are generally of moderate to high plasticity. In the United Kingdom, damage from soil shrinkage has become a major concern since the unusually dry summers of 1975– 76, and in the years following 1989 (Doorkamp 1993). The cost of repairs over two decades has exceeded one billion English pounds. The problem is most acute in Southeast England, especially with buildings founded on London clay, and to a lesser extent on Lias clay. Foundation design in these soils has little to do with bearing capacity estimates; it has much more to do with devising foundation systems that will protect the structure from the adverse affects of soil swelling (or shrinkage). The measures used to do this are outside the scope of this book.
232
RESIDUAL SOILS NOT DERIVED FROM VOLCANIC MATERIAL
REFERENCES Doornkamp, J. C. 1993. Clay shrinkage and induced subsidence. The Geographic Journal 159(2): 196–202. Lumb, P. 1962. The properties of weathered granite. Geotechnique 12(3): 226–243. Pender, M. J. 1971. Some properties of weathered greywacke. First Australia-New Zealand Conference on Geomechanics, Melbourne, Vol. 1, 423–429. Pender, M. J. 1980. Friction and cohesion parameters for highly and completely weathered Wellington Greywacke. 3rd Australia-New Zealand Conference on Geomechanics, Wellington, Vol.1, 171–175. Shukri, M., B. B. K. Huat, and S. Jamaludin. 2004. Index, engineering properties and classification of tropical residual soils. Tropical Residual Soils Engineering, Bujung B. K. Huat, Gue See Sew, and Faisal Haji Ali, eds. Leiden: A. A. Balkema.
CHAPTER 11
COMPACTION OF RESIDUAL SOILS
11.1
INTRODUCTION
Readers are no doubt familiar with the conventional understanding of compaction behavior of soils and the method developed by Proctor for measuring compaction characteristics and determining appropriate parameters for the control of compaction in the field. This conventional approach can be applied to many residual soils in the same way as it can to sedimentary soils; at the same time the characteristics of some residual soils make its application very difficult. Some rethinking of conventional wisdom is called for, with the possible adoption of alternative approaches. Among these characteristics are the following: 1. Residual soils are often much more variable than sedimentary soils, so that there is a continuous and random variation of optimum water content and maximum dry density throughout the soil. Figure 11.1 shows the results of compaction tests on samples from two sites involving residual soils. Both sites are relatively small, and the soil involved in each case is of the same geological origin. At the first site, the soil consists of relatively recent soils of Pleistocene origin, while the second site consists of much older soils weathered from a range of volcanic deposits, including basaltic lava flows and ash layers. Despite the common origin of the materials and the limited size of the sites, there is a large variation in the type of soil, as reflected in the compaction curves shown in the figure.
233
234
COMPACTION OF RESIDUAL SOILS
Industrial site: Pleistocene deposits
Dry Density (gm/cm3)
1.8
1.3
Steel mill site: Weathered basalt and ashes
1.2 Ze
1.4
ro
Ze
ro
ai
rv
oi
ds
oi
ds
lin
e
lin
e
1.0
ai
rv
1.1 1.0 0.9
0.6 20
40 60 Water content (%)
80
40
50 60 Water content (%)
70
Figure 11.1 Results of conventional compaction tests on samples from two sites near Auckland, New Zealand (after Pickens 1980).
2. The natural water content of some residual soils, especially those of volcanic origin, is often substantially higher than the optimum water content, and climatic conditions are not such that drying the soil is a practical possibility. At the same time it is true that many residual soils have natural water contents close to, or even below, their optimum water content and can be satisfactorily compacted without significant drying. 3. The highly structured nature of some residual soils tends to be destroyed by normal compaction methods so that the soil becomes progressively softer during the compaction process. 4. Some residual soils do not show clear peaks of dry density during conventional compaction tests, and thus do not have clearly defined optimum water contents. Figure 11.2 shows the results of standard Proctor compaction tests on two samples of clay derived from the weathering of volcanic ash. They are believed to consist predominantly of the clay mineral allophane. The tests have been carried out by progressively drying the soil from its natural water content, which was 195 and 166 percent for samples (a) and (b), respectively. Fresh soil was used for each point on the compaction curve. This procedure is essential, as repeated compaction may progressively soften the soil, and excessive drying before testing can cause irreversible changes to its properties (see Wesley 2002). Sample (a) does not show an optimum water content at all, while sample (b) shows an optimum value at around 135 percent, although it is not well defined. Tests were also carried out after various degrees of drying of the soil followed by rewetting; this clearly illustrates the irreversible changes that take place during drying
SOME REFLECTIONS ON COMPACTION BEHAVIOR
235
Dry density g/cm3
1.2 Natural Air dried Oven dried
1.0
Ze
ro
0.8
air
vo
ids
0.6
0.4
Sample (a)
20
40
60
80
100 120 140 Water content (%)
160
180
200
180
200
Dry density g/cm3
1.2 Natural Air dried Oven dried Air dried to 65 %
1.0
Ze
ro
0.8
air
vo
ids
0.6 Sample (b) 0.4 20
40
60
80
100 120 140 Water content (%)
160
Figure 11.2 Compaction tests on two allophane clays.
of the soil. The highest dry density is achieved by progressively drying the soil to a very low water content, but such a procedure is unlikely to be feasible in practice. 11.2 SOME REFLECTIONS ON COMPACTION BEHAVIOR OF SOILS AND QUALITY CONTROL METHODS
The Proctor compaction test has been widely used throughout the world since it was first introduced by Proctor (1933), and, in general, it has served the geotechnical profession very well. However, it has shortcomings on practical grounds, as some of the comments above already indicate. It can also be criticized on conceptual grounds, as it focuses on soil properties that are not of direct relevance to the performance of the compacted fill, namely dry density and water content. This focus is perfectly reasonable for most fills, as strength and compressibility are normally the properties of most concern in a compacted fill, and both are likely to be directly related to density. However, there are other properties that are sometimes of greater importance, such as permeability and ductility. These are not directly related
236
COMPACTION OF RESIDUAL SOILS
to dry density, and the use of other properties for controlling compaction may be more appropriate to ensure target values of such properties are achieved. The simplicity and widespread use of the Proctor test has possibly tended to blind the profession to the possibility of using parameters other than dry density and water content. It is important to recognize that the compaction of soil using mechanical methods (whether in the laboratory or the field) is likely to have two effects, one of which is not an intended effect and is not compaction at all. These effects are 1. “Densifying” the soil, that is, pushing the particles closer together and squeezing out air trapped between the particles. 2. Remolding the soil, causing it to soften. This involves the destruction of structure and is usually accompanied by the release of water trapped within or between the particles, adding to the softening process. The softening effect is likely to be present in all soils that exhibit some sensitivity. Much of the destructuring may occur during excavation, transport, and spreading of the soil prior to compaction, but it is also probable that many intact fragments of the soil will still exist and will retain their structure prior to the start of the compaction operation. This softening effect during compaction has been observed by many earthworks supervisors and is sometimes referred to as overcompaction. Figure 11.3 illustrates the softening effect that compaction has on volcanic soils in Japan. Tests have been done on a range of volcanic ash soils at their natural water content using varying compactive effort. The variation has been achieved by changing the number of blows of the compaction hammer on each layer. The strength of the soil after compaction has been measured using a cone penetrometer test, which gives a measure of the undrained shear strength of the soil. It is seen that nearly all of the samples become steadily softer as the blow count increases. Only sample A shows a consistent strength increase until the blow count reaches about 70, beyond which a small decrease in strength occurs. 11.3 OPTIMUM COMPACTIVE EFFORT AS WELL AS OPTIMUM WATER CONTENT
The behavior illustrated in Figure 11.3 shows that sensitive soils have an optimum compactive effort with respect to their maximum strength after compaction. This will be the case at their natural water content, and is likely to still be the case after the soil has undergone some drying. This provides us with an approach for compacting soils that are significantly wetter than optimum water content and cannot be dried back because of climatic conditions. Ideas of optimum water content and maximum dry density can be put aside in favor of investigating the compaction procedure
ALTERNATIVE COMPACTION CONTROL
237
16 Arrows indicate “optimum compactive effort”
w = 110%
A
Cone Index qc
12
Solid lines are various Kanto loams Kanuma soil w = 220%
w
=1
21
%
8
B w=
11
7%
4
C
w = 108%
Volcanic ash soil w = 59% 0
20
w = 109%
40 60 80 Number of Rammer Blows
100
D E
120
Figure 11.3 Influence of repeated compaction on the strength of volcanic ash soils (after Kuno et al. 1978).
that will produce the maximum undrained shear strength. This does not guarantee that such soils can, in fact, be successfully compacted regardless of weather conditions, but it does at least provide an approach that may be successful. Shear strength has been increasingly used in recent years for monitoring compaction; the following section describes a method of compaction control using undrained shear strength plus another parameter, air voids. 11.4 ALTERNATIVE COMPACTION CONTROL BASED ON UNDRAINED SHEAR STRENGTH AND AIR VOIDS
This method was developed in New Zealand to cope with the rapid variations in properties that occur in many local residual soils, such as those illustrated in Figure 11.1. It is also useful in overcoming some of the other difficulties described above that may be encountered when compacting residual soils. The method is described in detail by Pickens (1980),
COMPACTION OF RESIDUAL SOILS
Vane tests Unconfined comp. tests
300
1.6
200
1.5
100
1.4
1.3 20
Optimum water content
Dry density gm/cm3
1.7
0
25
30 Water content (%)
35
Undrained shear strength (kPa)
238
40
Figure 11.4 Standard Proctor compaction test on clay, including measurements of undrained shear strength.
and only an outline of the method is given here. Figure 11.4 illustrates the basis on which undrained shear strength can be used as one of the control parameters; it shows the results of a standard Proctor compaction test on clay, during which measurements of undrained strength have been made, in addition to density and water content. The measurements were made using both a hand shear vane and unconfined compressive tests. The two strength measurements give significantly different results. It is seen that at the optimum water content the undrained shear strength is about 150 kPa from the unconfined tests and about 230 kPa from the vane tests. We can note in passing that these values are to be expected; the optimum water content from a standard Proctor test on clay is normally close to the plastic limit, and the undrained shear strength at the plastic limit is normally assumed to be in the range of about 170–200 kPa. Conventional compaction specifications may allow water contents 2 or 3 percent greater than optimum, in which case the comparable shear strength values would be about 120 and 180 kPa. Thus, to obtain a fill with comparable properties to those obtained with conventional control methods, specifying a minimum undrained shear strength in the range of about 150–200 kPa would be appropriate. This would put an upper limit on the water content at which the soil could be compacted. Since the undrained shear strength steadily rises with decreasing water content, the required shear strength could be achieved by compacting the soil in a very dry state, which would generally be undesirable, as dry fills may soften and swell excessively when exposed to rainfall. To prevent the soil from being too dry a second parameter is specified, namely the air voids in the soil.
ar en
str gth
Limits from shear strength and air voids criteria
Ze
ro
air
vo
ids
Dry density
Dry density limit
Air Water content limits from compaction test
239
Shear strength
e Sh
Limits from water content and dry density criteria
Water content limit from shear strength criteria
ALTERNATIVE COMPACTION CONTROL
voi
ds
Shear strength limit
lim
it
Water content
Figure 11.5 Compaction control using alternative specification parameters.
Conventional Proctor compaction tests show that at optimum water content and maximum dry density the air voids in the soil are generally about 5 percent. If the soil is compacted 2–3 percent drier than the optimum water content corresponding to the compaction effort being used, the air voids may be as much as 8 or 10 percent. Thus, to prevent the soil from being compacted too dry, an upper limit is placed on the air voids, normally in the range of 8–10 percent. Figure 11.5 illustrates how this method of controlling compaction relates to the traditional method. The zero air voids line is always the upper limit of the dry density for any particular water content, and thus applies to both methods. The traditional method involves an upper and lower limit on water content, and a lower limit on dry density, and thus encloses the area shown in the figure. The alternative method in effect places an upper limit on water content, corresponding to the minimum shear strength, and a lower limit on dry density, corresponding to the line parallel to the zero air voids line representing the upper limit of air voids. There is no specific lower limit of water content, but the air voids limit prevents the soil from being too dry, as normal compaction equipment will be unable to achieve the required air voids limit. Experience has shown that suitable limits for the two control parameters are as follows: Undrained shear strength (hand vane values): Not less than 150 kPa (average of 10 tests) Minimum single value: 120 kPa Air voids (for “normal” soils): Not greater than 8 percent
240
COMPACTION OF RESIDUAL SOILS
These values have been found to be very satisfactory in producing firm, high-quality fills. The undrained shear strength can be measured in situ by hand shear vane, or by taking samples for unconfined compression tests. The hand shear vane is the much simpler of the two methods. The air voids can be determined only by measuring the density, water content, and specific gravity, in the usual way. The author’s experience has been mainly in temperate or wet tropical climates, where it is often the case that the soil is too wet and the undrained shear strength criterion is difficult to meet, while the air voids requirement is easily achieved. This means that the quality control consists essentially of checking the shear strength. With the hand shear vane this checking can be done easily as the compaction operation proceeds. It also makes it very easy for the contractor to do his (or her) own checking without having to wait for results from the quality control agency. While the criteria above are suitable for a wide range of compaction operations, there are some situations where other properties may be important, and the criteria can be adjusted accordingly. For example, the core of an earth dam built on compressible foundations, or in a seismic zone, may need to be plastic, or ductile, to allow for possible deformations in the dam. This can be achieved by adopting a lower undrained shear strength; a value between about 70 and 90 kPa would produce a firm, reasonably plastic material, assuming the clay is of moderate to high plasticity. For a clay embankment being built for a new highway, it may be desirable that the layers closest to the surface (on which the pavement itself will be constructed) have a higher strength than those deeper down. This could be achieved by increasing the required undrained shear strength to, say, 200 kPa. It will be evident from the account given above that this method of compaction control does not actually require compaction tests at all. However, it is still useful to carry out compaction tests to determine the degree of drying or wetting needed to bring the soil to a state appropriate for compaction. To summarize, the advantages of the shear strength and air voids control method are as follows: 1. Large variations in soil properties present no difficulty in applying the method. The same specification limits apply regardless of the variations. 2. Field control is more direct, as the value of the undrained shear strength is known immediately the measurements are made. 3. The specification is easily varied to produce fills with particular properties needed in special situations.
THE USE OF SHEAR STRENGTH TO OVERCOME DIFFICULTIES
241
11.5 THE USE OF SHEAR STRENGTH TO OVERCOME DIFFICULTIES IN COMPACTING RESIDUAL SOILS
We will now reconsider some of the difficulties mentioned at the start of this chapter that are encountered from time to time with residual soils. 11.5.1
Soils That Contain Wide and Random Variations in Properties
The advantages of using undrained shear strength and air voids in this situation have already been adequately covered above. The only additional comment to be made here is that the author has been involved with many projects where this method of control has been used, and its advantages cannot be overemphasized. These projects have included residential developments and earth dams in New Zealand, geothermal projects in Indonesia, and major highway projects in Malaysia. The method effectively overcomes all the uncertainties associated with variations in soil properties. The author’s experience has normally been with soils that are wet of optimum. This means that it is only the shear strength specification that needs to be focused on. The air voids are easily kept below the maximum permitted, and it is not essential to carry out tests to check the air voids. Because the shear strength is easily measured using a hand shear vane, a large number of tests can easily be made over a wide area, and the results are known immediately. This is a big advantage over the conventional method when samples have to be dried in an oven and the results are known only after a considerable time delay. 11.5.2 Nonsensitive Soils Considerably Wetter than Optimum Water Content
Residual soils of this type are rarely encountered, but are mentioned here to complete the picture. The situation with these soils is no different to that with sedimentary soils and involves the same challenge of how to dry the soil. If the soil is of low plasticity, drying is not particularly difficult, but drying soils of high plasticity can be very difficult. There are no easy ways to overcome this problem, but the following are essential requirements: • •
Adequate spells of fine sunny weather. This can be a very uncertain expectation in some countries. Plenty of wide open space to spread out the soil for drying. Such a space should be created or obtained so that it is exposed to maximum direct sunlight, and also to maximum wind.
242 •
COMPACTION OF RESIDUAL SOILS
Good site management. This means organizing the whole operation to maximize spells of fine weather for drying, and also being ready to “seal” the surface of any uncompacted material or stockpiles if rain is approaching. This “sealing” can be done by shaping all exposed surfaces so that rainfall cannot pond on them, and rolling the surface with a smooth-wheeled roller to create a tight impermeable surface layer.
11.5.3
Sensitive, Highly Structured Soils
These are the soils of the type shown in Figure 11.3, which in their undisturbed state are firm to stiff materials, but which become very soft when remolded or compacted. When dealing with soils of this type it is therefore important to understand their properties, and plan the compaction criteria accordingly. Laboratory tests of the type illustrated in Figure 11.3 will help in developing this understanding, but field tests will be considerably more informative. Depending on the climate and the time available for drying the soil, it may be necessary to choose between two options: 1. Drying the soil to its optimum water content, and using normal compactive effort to produce a high quality fill 2. Accepting that substantial drying is not feasible because of weather conditions, and adopting a much lower compactive effort so that the soil can be effectively compacted at (or close to) its natural water content To determine the feasibility of the second option, and especially to determine the optimum compaction method, it is desirable to conduct field trials involving the excavation, transport, and compaction of the soil. The excavation, transport, and spreading should be carried out in such a way that disturbance and remolding of the soil is kept to a minimum. In other words, the natural structure and strength of the soil should be retained as much as is practical. The compaction operation should similarly be conducted so that remolding the soil is minimized. Light, tracked equipment is likely to be most appropriate for this purpose, and the compaction process consists essentially of “squeezing” intact fragments of soil together to form a uniform fill. For this purpose, only a few passes of the compaction equipment are likely to be preferable to a large number of passes, as is clearly illustrated by the behavior in Figure 11.3. A hand field vane (or possibly a light hand-operated penetrometer) is the ideal tool for monitoring such trials. Careful consideration needs to be given to the lower limit of undrained shear strength to be accepted in this situation. For the construction of highway embankments or platforms that will not support high loads, quite low values of undrained shear strength may be acceptable, and indeed
REFERENCES
243
unavoidable. The author was involved in a geothermal project in West Java, Indonesia, which required the construction of a level platform to take transformers and other switchyard equipment. The soil at the site was allophane clay (weathered volcanic ash) of low to moderate sensitivity. The site was at a high altitude where the weather was seldom dry. During the rainy season there was no possibility of drying the soil and the “dry” season was of uncertain arrival and duration. A conventional specification had been adopted requiring an undrained shear strength of 150 kPa, but the maximum that could be achieved on site was less than 100 kPa, and this could be achieved only with difficulty. The specification was altered and the required shear strength lowered to 70 kPa, which could be achieved without too much difficulty. A fill of this strength is still a firm material and in this case was well capable of supporting the equipment it was intended for. It is also capable of taking light vehicular traffic. The required platform was completed without further difficulties and has performed perfectly satisfactorily ever since. 11.6
HARD, PARTIALLY WEATHERED, RESIDUAL SOILS
Some residual soils contain too much hard material for the above control method to be applied. In this case, a purely method-based specification may be appropriate. This would specify the type of compaction equipment to be used, the layer thickness, and the number of passes. Such a specification is satisfactory if there will be close supervision to ensure the method is adhered to. A more reliable method may be to use a hand-operated dynamic penetrometer; this can be calibrated by field trials to establish suitable criteria for particular materials. REFERENCES Kuno, G., R. Shinoki, T. Kondo, and C. Tsuchiya. 1978. On the construction methods of a motorway embankment by a sensitive volcanic clay. Proc. Conf. on Clay Fills, London, 149–156. Pickens, G. A. 1980. Alternative compaction specifications for non-uniform fill materials. Proc. Third Australia-New Zealand Conf. on Geomechanics, Wellington, 1.231–1.235. Proctor, R. R. 1933. Fundamental principles of soil compaction. Engineering News-Record, 111 (9, 10, 12, and 13). Wesley, L. D. 2002. Geotechnical characterization and behaviour of allophane clays. Proc. International Workshop on Characterisation and Engineering Properties of Natural Soils, Singapore, 2002, Vol. 2. Leiden, The Netherlands: Balkema, 1379–1399.
INDEX aging of soils, 3, 5 air voids, 238, 239, 241, 242 allophone clays, see volcanic soils bearing capacity, 135–139 allowable differential settlement, 98–99 andosols, 22–24 angle of shearing resistance (“friction angle”), 103–107 antecedent rainfall. 177 Atterberg limits, 17, 19, 105, 106, 126, 194, 195, 211–213, 330 plasticity chart, 18–20, 194, 207–208 use in soil classification, 19, 20, 194 back-analysis methods, 180–184 bearing capacity of clay, 135–139 black cotton soil (black clay), 9, 22, 151 block sampling, 122–124 boreholes, 119–121 classification methods, 14, 21–33 based on weathering profile, 13, 14
based on pedalogical classification, 22–24 for local use in specific soil types, 24 for residual soils, 21–33 system based on mineralogy and structure, 24–33 Unified soil classification system (USCS), 17–18 clay minerals, 7–9, 16, 25, 26, 190–193 coarse-grained soils, 37 coefficient of consolidation, 67, 69–72, 138 determination from odometer test, 67–72 typical values of in residual soils, 71 coefficient of one dimensional compressibility (mv ), 64, 67–69, 155 coefficient of permeability, 47, 166, 199–201 cohesion intercept, 151, 155, 202–204
245
246
INDEX
compaction, 233–243 control method using shear strength and air voids, 237–243 difficulties in compaction control, 233–235 of allophone clays, 203, 204, 235 standard Proctor compaction tests, 239 compactness indexes, 20, 21 composition, 24, 25 compressibility, 53–69 compression indexes, 54–57, 96, 195 computer processing of laboratory tests, 128–129 consolidation, 53 average degree of, 74, 75 magnitude, 76–99 oedometer test, 54–72, 85, 197–200 Terzaghi one-dimensional theory, 73 three dimensional, 73–76 time rate, 67–75, 82, 95, 96, 198–200 cone penetrometer test (CPT) test, 121, 122, 130–133, 138, 200–203, 214–217 critical circle, 163, 165 degree of consolidation, 74, 75 degree of saturation, 2, 195, 196, 225, 229, 230 design, 9–11 diatomaceous silt, 210–213 differential settlement, 81, 82, 98, 100 drilling, 119–121 hand auger, 119 machine, 120, 121 dry density (dry unit weight), 234, 238, 239
earth pressure, 139–146 earth retaining walls, see retaining walls effective stress strength parameters, 104–114, 181–184 equipotential lines, 161–165 excavated slopes, 184, 186 excess pore pressure, 40–43 expansive soils, 230–231 factor of safety, 160–163, 167–169 failure modes, 152, 153 flow lines, 161–163 flow nets, 162–164 formation processes, 2–5, 13–16 foundations, 135–139 friction angle (angle of shearing resistance), 103–107, 202, 203 correlations with index parameters, 105, 106 residual value, 112, 113 fully saturated soils, 2, 195, 196, 225, 229, 230 geogrid reinforcement, 146–149 granite, 14, 223–225 greywacke, 24, 227, 228 groundwater, 36–40 Halloysite, 7, 8, 23, 26, 27, 191, 192 hand augur, 119 hardening (or aging) of soils, 3, 5 Hong Kong, 172–180, 225 pore pressures, 173–176 hydrostatic pressure, 36, 38–40 igneous rocks, 6, 13, 14, 223–227 index tests, 17–21 influence of climate and weather, 2, 156–172, 174, 175, 177 in situ testing, 121, 122, 124–126 intrinsic properties, 18
INDEX
Judgment, 10 Kaolinite, 191 laterite, 191, 228, 229 lateritic soils, 22, 228, 229 liquidity index, 20, 21, 64–65, 103 liquid limit, 103 maximum dry density, 233, 235 Mohr-Coulomb failure criterion, 1 moisture content, see water content montmorillonite 9, 16, 19, 23 normally consolidated soil, 4, 5 observation, 10, 11, 16, 17 oedometer test, 54–71, 127 on residual soils, 57–65 determination of coefficient of consolidation, 67–72 determination of compressibility parameters, 54, 56–57 one-dimensional consolidation, 73 optimum compactive effort, 236, 237 optimum water content, 234–236 over-compaction, 236, 237 over-consolidated soil, 4, 5 over-consolidation ratio (OCR), 3, 33, 58–60 partially saturated soil, 36–38 particle size, 18,19, 193, 209, 214, 219, 225 peak shear strength, 107, 202 pedalogical classification, 22 penetrometer testing, 121, 122, 130–133, 138, 215–217 (Dutch) cone penetrometer test (CPT), 121, 122 standard penetration test (SPT), 121, 122
247
permeability, 37–40 coefficient of, 47, 166, 199–201 phreatic surface, 48 Piedmont soil, 58–60 Piezometers, 173–174 plastic limit, 103 plasticity chart, 18, 19 plasticity index, 18, 19, 194, 208, 213 pore pressure, 35–51, 201 dissipation tests, 200, 201 excess pore pressure, 67–72 pore water tension (suction), 36, 38, 39, 152, 166, 167, 169–172, 175, 176 pre-consolidation pressure, 3, 33, 58–60 Proctor compaction test, 238 rainfall influence on slope stability, 37, 40–44, 157–159 reinforced earth retaining walls, 146–150 drainage measures, 149 pore pressures and drainage measures, 148–249 use of residual soil, 146–150 relative density (density index), 20, 21, 132, 133 remolded soil, 65–67 residual soils consolidation behavior, 57–71 formation, 2–5 shear strength, 101–114, 124–126 undrained shear strength, 102, 103, 130–132 residual strength, 112–124 residual angle of shearing resistance (friction angle), 103–107 retaining walls, 139–150 reinforced earth, 146–150
248
INDEX
rhyolitic clays, 205–209 root time method, 69–72, 200 safety factor, 160–163, 167–169 sample disturbance, 122, 123 sandstone, 14–16, 60 saprolite, 15 seasonal effects, 37, 40–44, 157–177 sesquioxides, 191, 192 sedimentary rocks, 103, 226–228 sedimentary soils, 4, 224, 226–228 seepage and pore pressures in hill slopes, 44–47, 55 influence on stability, 146–169 transient, 164–172 worst case, 159–164 sensitivity, 81–82 settlement, 49, 50, 53–100 estimation in clays, 76–96 primary, 70, 71 secondary, 70, 71 time rate, 96 uncertainties in estimates, 82–96 shear strength, 101–114, 124–126 effective strength parameters, 103–107 measurement of, 103–110, 124–126 of sedimentary clays, 102 of residual soils, 107–112 residual value, 112–114 undrained, 102, 103, 130–132 site investigations, 115–133 block sampling, 122–124 continuous coring, 120, 121 hand auger boreholes, 119 in situ shear tests, 121, 122, 124–126 organization and administration, 116, 117 sample disturbance, 122, 123 slope stability, 151–188 analytical methods, 152, 154
determination of safety factor, 164–169 effective stress analysis, 162, 163 failure types/mechanisms, 152, 153 infinite slope, 140, 141 influence of weather and climate, 157–177 long term stability, 159, 160 short term stability, 159, 160 soft rocks, 226, 227 soil formation, 2–5, 13–16 standard penetration test (SPT), 121, 122 steady state flow, 159–163 storm events, 157–159, 169–172 strain softening, 64, 65 strain hardening, 64, 65 stress history, 3, 4, 58 structure of soils, 5–7, 20, 24, 67 macro-structure, 5, 24 micro-structure, 5, 24 swell (rebound) index, 56, 57 time rate of consolidation, 67–75, 82, 95, 96, 198–200 Terzaghi theory of consolidation, 40–43, 158 time factor, 73–76 topography influence, 8, 16 translational slides, 153, 140, 141 transient seepage state, 164–172 triaxial tests consolidated undrained, 104, 105, 107–110 tension tests, 111, 112 tropical red clay, 164–169, 228, 229 unconfined compression strength, 15 undisturbed sampling, 122–124 undrained shear strength, 102, 103, 130–132, 201–202
INDEX
Unified Soil Classification System (USCS), 194 vane test, 238 vegetation on slopes, 187 virgin consolidation line, 3, 5, 53, 65 volcanic soils, 1, 6, 8, 15, 22, 189–222 diatomaceous silt, 210–213 halloysite in soils, 7–9, 16, 25, 26, 190–193 pumiceous materials, 213–220 pumice sands, 213–217 volcanic ash clays (allophone clays), 17, 62, 63, 189–205 compressibility, 197–200 coefficient of consolidation, 198–200 compaction of, 203, 204
249
effective stress strength parameters, 202–204 rhyolitic clays, 205–209 undrained shear strength, 201, 202 Waitemata residual clay, 60–63, 111–112 water content, 8, 20 water table, 35–50, 165, 166, 169–172 weathering, 2, 3, 13, 14 chemical, 2 physical, 2 profiles, 10, 13–15 yield pressure (vertical yield stress), 21, 58, 64, 138 Young’s modulus, 67–69 zero air voids line, 234, 235
TECHNOLOGY/ENGINEERING/CIVIL
Geotechnical Engineering in Residual Soils digs deep to help enrich the reader’s knowledge on the subject of soils—in particular, residual soils—as they pertain to engineering. Appearing mostly in underdeveloped parts of the United States and tropical countries, these soils are playing an increasingly important role in building designs as construction encroaches into these areas. In recognition of this fact, this guide equips geotechnical engineers with essentials for learning the concepts and principles of residual soil behavior—and serves as a starting point to assist them in pursuing innovative engineering strategies for working effectively with residual soils. Geotechnical Engineering in Residual Soils: • Introduces geotechnical engineers to those aspects of residual soil behavior that they ought to be aware of when undertaking projects in these soils • Highlights the mistaken interpretations of soil behavior that can result from the application to residual soils of traditional concepts derived from sedimentary soils • Includes numerous illustrations throughout, specifically addressing the unique properties of residual soils • Includes coverage of special topics, such as the role of negative pore pressure above the water table, the influence of weather conditions on soil behavior, the properties of volcanic soils, and compaction of residual soils • Is written by an author with more than thirty years of firsthand experience analyzing and designing for construction on residual soils Thorough and insightful, Geotechnical Engineering in Residual Soils delivers a fresh overview on understanding the structural and mechanical properties of soils from an engineering perspective—and informs readers how to solidify design approaches to set their projects on a sure footing. LAURENCE D. WESLEY worked as a practicing geotechnical engineer for more than thirty years, with experience in New Zealand, Australia, Indonesia, Malaysia, and Bahrain. He is a Lifetime Member of the American Society of Civil Engineers, and a retired senior lecturer in geotechnical engineering at the University of Auckland. Cover Art © Istockphoto.com/Alejandro Raymond | Cover Design: Holly Wittenberg
Geotechnical Engineering in Residual Soils
The pioneering guide that breaks ground on the unique engineering properties of residual soils
Wesley
Geotechnical Engineering in Residual Soils
Laurence D. Wesley