AN INTRODUETION To Geotechnical Engineering ROBGET DeHloliz Willams Bo Kovacs FUCA Wand raga ly pare We OCAn Introduction to Geotechnical Engineering PRENTICE HALL, Englewood Cliffs, New Jersey 07632Library of Congress Catcloging in Publication Data Hourz, Roamer D. 2 ‘Aa introduetion to geotechnical enginedring Tocludes index. Soll properties. 2. Soil mechanics. 1. Kovacs. William D., joint author. IL. Tite TATIONS" 6241'513. 8023292 ISBN 0-13-48429¢0 DEDICATION: To Our Teachers, Past and Present Editorial [production supereision ‘and interior design by Karen Skrable Manufacturing buyers: Anthony Caruso and Joyce Levatino Cover design by Edsal Enterprises EB © 1981 by Prentice Hall, Inc. A Paramount Communications Company Englewood Cliffs, New Jersey 07632 All rights reserved. No part of this book may bbe reproduced, in any form or by any means, without permission in writing from the publisher. 90000 Printed in the United States of America 30 29 28 27 26 ISBN O-13-484394-0 9978013484394 Prentice-Hall International (UK) Limited, London Prentice-Hall of Australia Pty. Limited, Sydney Prentice-Hall Canada Inc., Toronto Prentice-Hall Hispanoamericana, S.A., Mexico Prentice-Hall of India Private Limited, New Delhi Prentice-Hall of Japan, Inc., Tokyo Simon & Schuster Asia Pre. Lid., Singapore Editora Prentice-Hall do Brasil, Lida., Rio de JaneiroContents PREFACE xi 4 INTRCDUCTION TO GEOTECHNICAL ENGINEERING 1 1.1 Geotechnical Engineering 1 1.2. The Unique Nature of Soil and Rock Materials 3 1.3 Suggested Approach to the Study of Geotechnical Engineering 4 14 Scope of this Book 5 15 Soil Formation and the Nature of Soil Constituents 6 1.6 Historical Development of Geotechnical Engineering 7 1.7 Notes on Symbols and Units 9 2 INDEX AND CLASSIFICATION PROPERTIES OF SOILS 10 2.1 Introduction 10 22 Basic Definitions and Phase Relations 1 23 Solution of Phase Problems 16 24 Soil Texture 25 2.5 Grain Size and Grain Size Distribution 26 2.6 Particle Shape 32 2.7 Atterberg Limits and Consistency Indices 34 28 Activity 41 Problems 4/ 3 SOIL CLASSIFICATION 47 3.1 Introduction 47 3.2 The Unified Soil Classification System (USCS) 4933 34 ‘The AASHTO Soil Classification System 64 ‘Comparison of the USCS and AASHTO Classification Systems 72 Problems 72 4 CLAY MINERALS AND SOIL STRUCTURE 77 41 42 43 44 45 46 47 48 49 Introduction 77 Clay Minerals 78 Identification of Clay Minerals 88 Specific Surface 89 Interaction Between Water and Clay Minerals Interaction of Clay Particles 95 a Soil Structure and Fabric 96 Cobesive Soil Fabrics 96, Cobesionless Soil Fabrics 102 Problems 107 5 COMPACTION 109 Sa 52 53 54 35 56 $a 6 WATER IN SOILS, I: Introduction 109 Compaction 110 Theory of Compaction 171 Properties and Structure of Compacted Cohesive Soils 117 Field Compaction Equipment and Procedures Contents 90 Im, Field Compaction Coritrd¥and Specifications 141 Estimating Performance of Compacted Soils 153 Problems 161 CAPILLARITY, SHRINKAGE, SWELLING, FROST ACTION 166 61 62 63 64 65 Introduction 166 Capillarity 167 Shrinkage Phenomena in Soils 178 Engineering Significance of Shrinkage and ‘Swelling 186 Frost Action 190 Problems 195Contents, wt 7 WATER IN SOILS, Il PERMEABILITY, SEEPAGE, EFFECTIVE STRESS 199 7.1 Introduction 199 72 Dynamics of Fluid Flow 200 7.3. Darcy's Law for Flow Through Porous Media 203 14 Measurement of Permeability 206 7.5 Intergranular or Effective Stress 213 7.6 Relationship between Horizontal and Vertical Stresses 225 7.7 Heads and One-Dimensional Flow 227 78 Seepage Forces, Quicksand, and Liquefaction 232 7.9 Seepage and Flow Nets: Two-Dimensional Flow 246 7.10 The Method of Fragments 258 7.11 Control of Seepage and Filters 270 Problems 273 8 CONSOLIDATION AND CONSOLIDATION SETTLEMENTS 283 8.1 Introduction 283 8.2 Components of Settlement 284 83 Compressibility of Soils 285 84 The Oedometer and Consolidation Testing 289 8.5 Preconsolidation Pressure; Normal, Overconsolidation, Underconsolidation 294 8.6 Consolidation Behavior of Natural Soils 299 8.7 Settlement Calculations 309 8.8 Factors Affecting the Determination of of 326 89 Prediction of Field Consolidation Curves 328 8.10 Soil Profiles 335 8.11 Approximate Methods and Typical Values of Compression Indices 341 8.12 Stress Distribution 342 Problems 367 9 TIME RATE OF CONSOLIDATION 376 9.1 Introduction 376 9.2 The Consolidation Process 377 9.3. Terzaghi’s One-Dimensional Consolidation Theory 380 9.4 Determination of the Coefficient of Consolidation cy 395 9.5 Determination of the Coefficient of Permeability 402 9.6 Typical Values of cy 405wu 97 98 Evaluation of Secondary Settlement . 405 Comprehensive Example of a Time Rate of Settlement Problem 4/4 Problems 423 - 10 THE MOHR CIRCLE, FAILURE THEORIES, AND STRESS PATHS 437 " 10.1 10.2 103 104 105 106 Introduction 431 Stress at a Point 432 ‘Stress-Strain Relationships and Failure Criteria’ 446 ‘The Mobr-Coulomb Failure Criterion 449 Tests for the Shear Strength of Soils 458 Stress Paths 473 Problems 485 SHEAR STRENGTH OF SANDS AND CLAYS 490 MW 2 u3 M4 ns us na 18 ng Hot 192 193 194 19s 11.96 11.97 119.8 11.99 1.9.10 nan 1.9.12 1.9.13 Introduction 490 Angle of Repose of Sands 492 Behavior of Saturated Sands During Drained Shear 493 Effect of Void Ratio and Confining Pressure on Volume Change 496 Behavior of Saturated Sand During Undrained Shear 504 Factors that Affect the Shear Strength of Sands 5/4 ‘The Coefficient of Earth Pressure at Rest for Sands 519 Liquefaction and Cyclic Mobility Behavior of Saturated Sands 521 ‘Stress-Deformation and Strength Characteristics of Saturated Cobesive Soils 536 Consolidated-Drained (CD) Test Behavior 538 Typical Values of Drained Strength Parameters 543 Use of CD Strength in Engineering Practice 545 Consolidated-Undrained (CU) Test Behavior 545 Typical Values of the Undrained Strength Parameters 553 Use of CU Strength in Enginecring Practice 556 Unconsolidated-Undrained (UU) Test Behavior 559 Typical Values of UU Strengths 566 Unconfined Compression Test 566 Other Ways to Determine the Undrained Shear Strength 570 Sensitivity 585 Use of Undrained (UU) Shear Strength in Engineering Practice 586 Special Problems of the Shear Strength of Cohesive Soils 59511.10 Pore Pressure Parameters 599 M111 The Coefficient of Earth Pressure at Rest for Clays 605 11.12. Stress Paths During Undrained Loading —Normally Consolidated Clays 610 11.13 Stress Paths During Undrained Loading — Overconsolidated Clays 630 11.14 Applications of Stress Paths to Engineering Practice 634 Problems 640 APPENDIX A APPLICATION OF THE “SI” SYSTEM OF UNITS TO GEOTECHNICAL ENGINEERING 665 Al Introduction 665 A2 The SI Metric System 666 A3_ Basic and Derived SI Metric Units 667 A4_ SI Units of Interest to Geotechnical Engineers and Their Conversion Factors 669 APPENDIX B-1 DERIVATION OF LAPLACE’S EQUATION 687 APPENDIX B-2 DERIVATION AND SOLUTION OF TERZAGHI'S ONE-DIMENSIONAL CONSOLIDATION THEORY 683 B21 Assumptions 683 B22 Derivation 683 B-2.3 Mathematical Solution 686 APPENDIX 8-3 PORE PRESSURE PARAMETERS 691 B-3.1 Derivation of Skempton’s Pore Pressure Equation 691 B-3.2 Definition of Ao, and Ao, for Rotation of Principal Stresses 693 B-3.3 Formulas for Pore Pressure Parameters for Different Stress Path Tests 694 B-3.4 Proof that 4j.= Ay, and Age Aye 696 B-3.5 Derivation of the Henkel Pore Pressure Equation and Coefficients 696 REFERENCES 701 INDEX 719Preface ‘An Introduction to Geotechnical Engineering is intended for use in the first of a two-course sequence in geotechnical engineering usually taught to third- and fourth-year undergraduate Civil Engineering students. We as- sume the students have a working knowledge of undergraduate mechanics, especially statics and mechanics of materials (including fluids). A know!- edge of basic geology, although helpful, is not essential. We introduce the “language” of geotechnical engineering in the first course, that is, the classification and engineering properties of soils. Once the student has a working knowledge of how soil behaves as an engineering material, he/she can begin to predict soil behavior and, in the second course, to carry out the design of simple foundations and earthwork systems. We feel that there is a need for more detailed and modern coverage of the engineering properties of soils than is found in most undergraduate texts. This applies to both the students “majoring” in geotechnical en- gineering and the general civil engineering undergraduate student. We find that our students are involved in increasingly more complex projects, especially in transportation, structural, and construction engineering. En- vironmental, economic, and political constraints demand innovative solu- tions to civil engineering problems. The availability of modern analytical techniques and the digital computer has had an almost revolutionary effect ‘on engineering design practice. This development demands a better knowl- edge of site conditions as well as better defined geotechnical engineering design parameters. ‘We have tried to make the text easily readable by the average undergraduate. To this end, An Introduction to Geotechnical Engineering is written at a simple rather than sophisticated level, although the material covered may be rather sophisticated at times. Involved derivations, unreadxn Pretace by the typical student, are relegated to appendices where they are available to the interested student. The emphasis throughout is on the practical, and admittedly empiri- cal, knowledge of soil behavior required by the geotechnical engineer for the design and construction of foundations and embankments. Most of the material in the text is descriptive, since most of the engineering design applications are usually left to the second course in Foundation Engineer- ing. Consequently, in order to interest the student, we have tried to indicate wherever possible the engineering significance and specific design applications of the soil property being discussed. We have tried to em- Phasize why such a property is needed, how it is determined or measured, and, to some extent, how it is actually used in practice. The only “design” type problem we cover in a one-semester course (15 weeks) is estimating the settlement of shallow foundations on saturated clays., The text is sufficiently flexible that innovative instructors can add additional design examples should they so desire. It seems that units are always a problem with geotechnical engineers. In line with the trend towards the use of S. units encouraged by the American Society of Civil Engineers and Ame can Society for Testing and Materials, we have used this system in the text. The uninitiated may find the discussion of S.J. in Appendix A helpful. In addition to the almost exclusive use of S.1. units in examples and problems, we have been careful to use the latest definition of density (mass/unit volume) in phase relationships as well as in geostatic and hydrostatic pressure computations. We consider the laboratory portion of the first course to be an essential part of the neophyte engineer's experience with soils as a unique engineering material. How else is the young engineer to begin to develop a “feel” for soils and soil behavior, so essential for the successful practice of geotechnical engineering? Thus, an emphasis on laboratory and field testing is found throughout the text. The order of the laboratory portion of our first course has dictated the organization and development of the material in the text, We begin with phase relations, visual classification of soils, and simple classification tests. Thus, the early chapters introduce the discipline of Geotechnical Engineering, Phase Relationships and Index Properties, Soil Classification, and Clay Minerals and Soil Structure. This material provides the background and terminology for the later -chapters. Following a very practical discussion of Compaction in Chapter 5, Chapters 6 and 7 describe how water influences and affects soil behavior. Topics Presented include capillarity, shrinkage, swelling, and frost action as well as permeability, seepage, and effective stress. These two chapters again serve as background for the next four chapters which deal with consolida- tion and shear strength.Pretece tt The treatment of these latter topics is quite modern and up-to-date. The Schmertmann procedure for determining field compressibility is in- cluded as is a modern treatment of secondary compression developed by Prof. Mesri and his co-workers. Prof. Lambe’s stress path method is introduced in Chapter 10 and used to advantage in Chapter 11, especially when practical engineering applications of shear strength theory are dis- cussed. The pioneering work of Profs. Seed and Lee on the drained and undrained strength of sands is presented in Chapter 11. Also in this chapter we discuss the stress-deformation and strength characteristics of cohesive soils. Although the treatment is modern, because this is primarily an undergraduate textbook, considerations of strength anisotropy, critical state concepts, the Jiirgenson-Rutledge hypothesis, and Hvorslev’s strength parameters have been left to more advanced texts. Even though the book is primarily for the beginning student in geotechnical engineering, advanced students in other disciplines and en- gineers desiring a refresher in engineering properties may find the book helpful. Because of the many fully-worked example problems, the book is almost “self-teaching.” This aspect of the text also frees the instructor in a formal course from the necessity of working example problems during lectures. It allows the instructor to concentrate on explaining basic princi ples and illustrating specific engineering applications of the points in question. The third group we hope will find this book useful are practicing geotechnical engineers. Typical values are given for all classification ard engineering properties for a wide variety of soils; we have found such a compendium very useful in our own engineering practice. To acknowledge all who have contributed to this book is a formida- ble task. We have tried whenever possible to indicate by references or quotations, concepts and ideas originating in the literature or with our former teachers, especially Profs. A. Casagrande and H. B. Seed. We apologize for any omissions. We must also mention the students in our beginning geotechnical engineering course at Purdue who have graciously suffered through several versions of An Introduction t0 Geotechnical En- gineering in note form. Their criticism and helpful comments on the text have been very valuable. The authors have greatly benefited from discus- sions with Prof. MoE. Harr of Purdue University regarding the section on the method of fragments in Chapter 7. We hope to bring further attention to the profession of this powerful design method. Dr. E. Simiu of the U.S. National Bureau of Standards critically read a recent version of the manuscript and provided many helpful comments. It should be noted that An Introduction to Geotechnical Engineering was written while William D. Kovacs was on the faculty of Purdue University. The text has no connec- tion with Kovacs’ present affiliation with the National Bureau of Stan-aw Pretece dards. Our faithful secretaries Mrs. Janice Wait Bollinger, Miss Cathy Minth and Mrs. Edith Vanderwerp deserve special thanks for typing and correcting the several drafts. The first author also wishes to gratefully acknowledge the interest and encouragement of his wife Cricket Morgan. Her work with the proofreading and corrections is especially appreciated. We of course will appreciate any comments and criticism of readers. R.D. Holtz, W. D. Kovacs West Lafayette, indianaone Introduction to Geotechnical Engineering 1.1 GEOTECHNICAL ENGINEERING Geotechnical engineering, as the name implies, concerns the applica- tion of civil engineering technology to some aspect of the earth. Usually, the geotechnical engineer is concerned only with the natural materials found at or near the surface of the earth. Civil engineers call these earthen materials soil and rock. Soil, in an engineering sense, is the relatively loose agglomerate of mineral and organic materials and sediments found above the bedrock. Soils can be relatively easily broken down into their con- stituent mineral or organic particles. Rocks, on the other hand, have very strong internal cohesive and molecular forces which hold the constituent ‘mineral grains together. This is true whether the rock is massive bedrock or a piece of gravel found in a clay soil. The dividing line between soil and rock is arbitrary, and many natural materials encountered in engineering practice cannot be easily classified. They may be either a “very soft rock” or a “very hard soil.” Other scientific disciplines have different meanings for the terms soil and rock. In geology, for example, rock means all the materials found in the earth’s crust, independently of how much the mineral particles are bound together. Soils to a geologist are just decom- posed and disintegrated rocks generally found in the very thin upper part of the crust and capable of supporting plant life. Similarly, pedology (soil science) and agronomy are concerned with only the very uppermost layers of soil, that is, those materials relating to agriculture and forestry. Geotech- nical engineers can learn much from both geology and pedology. Both sciences, especially engineering geology, are important adjuncts to geotech- nical engineering and there is considerable overlap between these fields. But differences in terminology, approach, and objectives may cause some confusion, especially for the beginner.2 Introduction to Geotechnical Engineering Geotechnical engineering has several different aspects or emphases. Soil mechanics is the branch of geotechnical engineering concerned with the engineering mechanics and properties of soil, whereas rock mechanics is concerned with the engineering mechanics and properties of rock, usually but not necessarily the bedrock. Soil mechanics applies the basic principles of mechanics including kinematics, dynamics, fluid mechanics, and the mechanics of materials to soils. In other words, soil rather than water or steel or concrete, for example, now becomes the engineering material whose properties and behavior we must understand in order to build with it or upon it. A similar comment could also be made for rock mechanics. It should be noted, however, that there are significant differences between the behavior of soil masses and rock masses, and in principle there is not much overlap between the two disciplines. Foundation engineering applies geology, soil mechanics, rock mecha- nics, and structural engineering to the design and construction of founda- tions for civil engineering and other structures. The foundation engineer must be able to predict the performance or response of the foundation soil or rock to the loads imposed by the structure. Some examples of the kinds of problems faced by the foundation engineer include foundations for industrial, commercial, and residential buildings, and other types of sup- port structures for radar towers, as well as foundations for oil and other kinds of tanks and offshore structures. Even ships must have a dry dock during construction or repairs, and the dry dock must have a foundation. The support of rockets and appurtenant structures during construction and launch have led to very interesting and challenging foundation engineering problems. Related geotechnical engineering problems facing the founda- tion engineer are the stability of natural and excavated slopes, the stability of permanent and temporary carth-retaining structures, problems of con- struction, controlling water movement and pressures, and even the mainte- nance and rehabilitation of old buildings. Not only must the foundation safely support the static structural and construction loads, but it must also adequately resist dynamic loads due to blasting, earthquakes, etc. Hf you think about it, it is impossible to design or construct any civil engineering structure without ultimately considering the foundation soils and rocks to some extent, and this is true whether the structure is built on the earth or is extraterrestrial. The performance, economy, and safety of any civil engineering structure ultimately is affected or may even be controlled by its foundation. Earth materials are often used as a construction material because they are the cheapest possible building material. However, its engineering Properties such as strength and compressibility are often naturally poor, and measures must be taken to densify, strengthen, or otherwise stabilize and reinforce soils so that they will perform satisfactorily in service.1.2 The Unique Nature of Soll and Rock Materials 3 Highway and railway embankments, airfields, earth and rock dams, levees, and aqueducts are examples of earth structures, and the geotechnical engineer is responsible for their design and construction. Dam safety and rehabilitation of old dams are important aspects of this phase of geotechni- cal engineering. Also related, especially for highway and airfield engineers, is the design of the final surface layer on the earth structure, the pavement. Here the overlap between the transportation and geotechnical disciplines is apparent. Rock engineering, analogous to foundation engineering for soils, is concerned with rock as a foundation and construction material. Because ‘most of the earth's surface is covered with soil (or water), rock engineering, usually occurs underground (tunnels, underground power houses, petro- eum storage rooms, mines, etc.). But sometimes rock engineering occurs at the surface, such as in the case of building and dam foundations carried to bedrock, deep excavations to bedrock, stability of rock slopes, etc. In presenting some of the typical problems facing the geotechnical engineer, we wanted you to see, first, how broad the field is and, second, how important it is to the design and construction of civil engineering structures. In a very real sense, geotechnical engineering combines the basic physical sciences, geology and pedology, with hydraulic, structural, transportation, construction, and mining engineering, 1.2 THE UNIQUE NATURE OF SOIL. AND ROCK MATERIALS: Geotechnical engineering is highly empirical and is perhaps much more of an “art” than the other disciplines within civil engineering because of the basic nature of soil and rock materials. They are often highly variable, even within a distance of a few millimetres. Another way of saying this is that soils are heterogeneous rather than homogeneous materials. That is, their material or engineering properties may vary widely from point to point within a soil mass. Furthermore, soils in general are nonlinear materials; their stress-strain curves are not straight lines. To further complicate things (as well as to make them interesting!) soils are nonconservative materials; that is, they have a fantastic memory—they remember almost everything that ever happened to them, and this fact strongly affects their engineering behavior. Instead of being isotropic, soils are typically anisotropic, which means that their material or engineering properties are not the same in all directions. Most of the theories we have for the mechanical behavior of engineering materials assume that the materials are homogeneous and isotropic, and that they obey linear stress- strain laws. Common engineering materials such as steel and concrete do4 Introduction to Geotechnical Engineering not deviate too significantly from these ideals, and consequently we can use, with discretion, simple linear theories to predict their response under engineering loads. With soils and rock, we are not so fortunate. As you shall see in your study of geotechnical engineering, we may assume a linear stress-strain response, but then we must apply large empirical correction or “safety” factors to our designs to account for the real material behavior. Furthermore, the behavior of soil and rock materials in situ is often governed or controlled by joints, fractures, weak layers and zones, and other “defects” in the material; yet our laboratory tests and simplified methods of analysis often do not take into account such real characteristics of the soil and rock. That is why geotechnical engineering is really an “art” rather than an engineering science. Successful geotechnical engineering depends on the good judgment and practical experience of the designer, constructor, or consultant. Put another way, the successful geotechnical engineer must develop a “feel” for soil and rock behavior before a safe and economic foundation design can be made or an engineering structure can be safely built. 1.3 SUGGESTED APPROACH TO THE STUDY OF GEOTECHNICAL ENGINEERING Because of the nature of soil and rock materials, both laboratory and field testing are very important in geotechnical engineering. One way that student engineers can begin to develop a feel for soil and rock behavior is to get some experience in the laboratory by performing the standard tests for classification and engineering properties on many different types of soils and rocks. In this way the novice begins building up a “mental data bank” of how certain soils and rocks actually look, how they might behave should, for example, the amount of water present change, how they might behave under different kinds of engineering loads, and what the range of probable numerical values is for the different tests. This is sort of a self-calibration process, so that when you are faced with a new sail deposit or rock type, you will in advance have some idea as to the engineering problems you will encounter at that site. You can also begin to judge, at least qualitatively, the validity of laboratory and field test results for the materials at that site. So laboratory as well as field experience is important for you to help develop a “feel” for soil and rock behavior. Of course, just as with any other subject, this exposure in the laboratory to soil and rock properties and behavior must be complemented by a diligent study of the theoretical, empirical, and design components of geotechnical engineering practice.1.4 SCOPE OF THIS BOOK Rather than attempt an all-inclusive approach to geotechnical en- gineering, the primary emphasis in this text will be on the engineering behavior of soil materials. Soil mechanics and the analysis and design of foundations and earth structures is generally a fairly straightforward, but creative, application of mechanics, strength of materials, and elementary structural engineering, Often the key in the successful practice and applica- tion of geotechnical engineering lies in a sound knowledge and understand- ing of the engineering properties and behavior of soils in situ, when they are subjected to their engincering loads and environmental conditions. Therefore we feel that the beginning student must first develop an appreci- ation for the engineering properties of soils as distinct from other common. civil engineering materials before proceeding to instruction in the analysis and design phases of foundation and earthwork engineering, ‘This is an elementary text, and the approach we have tried to follow is to emphasize the fundamentals, with an eye toward the practical applications that you as a practicing civil engineer are likely to encounter in your engineering practice. Finally, we hope you will know enough about soils and soil deposits to avoid serious mistakes or blunders in those aspects of your professional career that involve soil and soil materials, In the first part of the book, we introduce some of the basic definitions and index properties of soil that are used throughout the book. Then some common soil classification schemes are presented. Classification of soils is important because it is the “language” engineers use to com- municate certain general knowledge about the engineering behavior of the soils at a particular site. The rest of the book is concerned with the ‘engineering properties of soil, properties that are necessary for the design of foundations and earth structures. Topics covered include how water affects soil behavior, their shrinkage and swelling characteristics, and their per- meability (how water flows through soils). Then we get into the compressi- bility of soil, which is the important engineering property one needs to predict the settlement of engineering structures constructed on soil masses. Finally, we describe some of the elementary strength characteristics of both granular and cohesive soils. Soil strength is important, for example, for the design of foundations, retaining walls, and slopes. Much of the practice of geotechnical engineering depends on topics that include geology and the nature of landforms and soil deposits. You are strongly encouraged to take a physical- geology or an engineering geology course in connection with your studies of geotechnical engineer- ing.6 Introduction to Geotechnical Engineering It is hoped that with the background of this text, you will be prepared for a follow-up course in foundation and earthwork engineering; you should know how to obtain the soil properties required for most designs, and you should have a pretty good idea as to the probable range of values for a given property if you know the general classification of the soil. Finally, you should have a fairly good idea of what to look for at a site, how to avoid costly and dangerous mistakes, and be aware of your own limitations and knowledge of soils as an engineering material. 1.5 SOIL FORMATION AND THE NATURE. OF SOIL CONSTITUENTS We mentioned earlier that soil from a civil engineering point of view is the relatively loose agglomeration of mineral and organic materials found above the bedrock. In a broader sense, of course, even shallow bedrock is of interest to geotechnical engineers and some of these applica- tions have already been mentioned. You may remember from your basic science courses that the earth has a crust of granitic and basaltic rocks 10 to 40 km thick. Overlying this more or less solid rock is a relatively thin layer of variable thickness of what geologists call unconsolidated materials. These materials can vary in size from sub-microscopic mineral particles to huge boulders. Weathering and other geologic processes act on the rocks at or near the earth’s surface to form these unconsolidated materials, or soil. Weathering, which usually results from atmospheric processes, alters the composition and structure of these rocks by chemical and physical means. Physical or mechanical weathering causes disintegration of the rocks into smaller particle sizes. Physical weathering agents include freezing and thawing, temperature changes, erosion, and the activity of plants and animals including man. Chemical weathering decomposes the minerals in the rocks by oxidation, reduction, carbonation, and other chemical processes. Generally, chemical weathering is much more important than physical weathering in soil formation. In short then, soils are the products of the weathering of rocks. Soils at a particular site can be residual (that is, weathered in place) or transported (moved by water, wind, glaciers, etc.), and the geologic history of a particular deposit significantly affects its engineering behavior. ‘The nature of soil constituents is discussed in greater detail throughout this text. For now, we want to make a few points just to set the stage for what we are about to study. You already have a layman’s idea about soil. At least you know in general what sand and gravel are, and perhaps you even have a general idea about fine-grained soils such as silts and clays. These terms have quite precise engineering definitions, as we shall laterHistorical Development of Geotechnical Engineering 7 see, but for now the general concept that soils are particles will suffice. Particles of what? Well, usually particles of mineral matter or, more simply, broken up pieces of rock that result from the weathering processes we spoke of previously. If we just talk for now about the size of the particles, gravels are small pieces of rock that typically contain several minerals, whereas sands are even smaller and each grain usually contains only a single mineral. If you cannot see each grain of a soil, then the soil is either a silt or a clay or a mixture of each. In fact, natural soils generally are a mixture of several different particle sizes and may even contain organic matter. Some soils such as peat may be almost entirely organic. Futhermore, because soils are a particulate material, they have voids, and the voids are usually filled with water and air. It is the physical and chemical interaction of the water and air in the voids with the particles of soil, as well as the interaction of the particles themselves, that makes soil behavior so complicated and leads to the nonlinear, nonconservative, and anisotropic mechanical behavior we mentioned previously. Now, if you add the variability and heterogeneity of natural soil deposits due to the capriciousness of nature, you probably can begin to see that soils are indeed complex engineering and construction materials. Helping you put some order into this potentially chaotic situation is our primary objective in this book. 1.6 HISTORICAL DEVELOPMENT OF GEOTECHNICAL ENGINEERING As long as people have been building things, they have used soils as a foundation or construction material. The ancient Egyptians, Babylonians, Chinese, and Indians knew about constructing dikes and levees out of the soils found in river flood plains. Ancient temples and monuments built all around the world involved soil and rock in some way. The Aztecs con- structed temples and cities on the very poor soils in the Valley of Mexico tong before the Spaniards arrived in the New World. European architects and builders during the Middle Ages learned about the problems of settlements of cathedrals and large buildings. The most noteworthy exam- ple is, of course, the Leaning Tower of Pisa. Scandinavians used timber piles to support houses and wharf structures on their soft clays. The “design” of foundations and other constructions involving soil and rock was by rule of thumb, and very little theory as such was developed until the mid-1700's. Coulomb is the most famous name of that era. He was interested in the problems of earth pressures against retaining walls, and some of his calculation procedures are still in use today. The most common theory for8 Introduction to Geotechnical Engineering the shear strength of soils is named after him. During the next century, the French engineers Collin and Darcy (D'Arcy) and the Scotsman Rankine made important discoveries. Collin was the first engineer to be concerned with failures in clay slopes as well as the measurement of the shear strength of clays. Darcy established his law for the flow of water through sands. Rankine developed a method for estimating the earth pressure against retaining walls. In England, Gregory utilized horizontal subdrains: and compacted earth-fill buttresses to stabilize railroad cut slopes. By the turn of the century, important developments in the field took place in Scandinavia, primarily in Sweden. Atterberg defined the con- sistency limits for clays that are still in use today. During the period 1914-1922, in connection with investigations of some important failures in harbors and railroads, the Geotechnical Commission of the Swedish State Railways developed many important concepts and apparatuses in geotech- nical engineering. Methods for calculating the stability of slopes were developed. They developed subsurface investigation techniques such as weight sounding and piston and other types of samplers. They understood important concepts such as sensitivity of clays and consolidation, which is the squeezing of water out of the pores of the clay. At that time, clays were thought to be absolutely impervious, but the Swedes made field measure- ‘ments to show that they weren't. The Commission was the first to use the word geotechnical (Swedish: geotekniska) in the sense that we know it today: the combination of geology and civil engineering technology. Even with these early developments in Sweden, the father of soil mechanics is really an Austrian, Prof. Karl Terzaghi. He published in 1925 the first modern textbook on soil mechanics, and in fact the name “soil mechanics” is a direct translation of the German word erdbaumechanik, which was part of the title of that book. Terzaghi was an outstanding and very creative engineer. He wrote several important books and over 250 technical papers and articles, and his name will appear many times in this book. He was a professor at Robert College in Istanbul, Technische Hochschule in Vienna, M. I. T., and at Harvard University from 1938 until his retirement in 1956. He continued to be active as a consultant until his death in 1963 at the age of 80. Another important contributor to the advancement of modern soil mechanics is Prof. Arthur Casagrande, who was at Harvard University from 1932 until 1969. You will see his name often in this book because he made many important contributions to the art and science of soil mecha- nics and foundation engineering. Other important contributors to the field include Taylor, Peck, Tschebotarioff, Skempton, and Bjerrum. Since the 1950's the field has grown substantially and the names of those responsible for its rapid advancement are too numerous to mention.1.7, Notes on Symbols and Units s Both Terzaghi and Casagrande began the teaching of soil mechanics and engineering geology in the United States. Before the Second World War, the subject was offered only as a graduate course in very few universities, After the war, it became common for at least one course in the subject to be required in most schools of civil engineering. In recent years graduate programs in all phases of geotechnical engineering have been implemented at many universities, and there has been a real information explosion in the number of conferences, technical journals, and textbooks published during the past two decades. Important recent developments you should know about include de- velopments in earthquake engineering and soil dynamics, the use of digital computers for the solution of complex engineering problems, and the introduction of probability and statistics into geotechnical engineering analysis and design. 1.7 NOTES ON SYMBOLS AND UNITS ‘At the beginning of each chapter, we list the pertinent symbols introduced in the chapter. As with most disciplines, a standard notation is not universal in geotechnical engineering, so we have tried to adopt the symbols most commonly used. For example, the American Society for Testing and Materials (ASTM, 1979) has a list of Standard Definitions of Terms and Symbols Relating to Soil and Rock Mechanics, Designation D 653, which was prepared jointly some years ago with the American Society of Civil Engineers (ASCE) and the International Society of Rock Mecha- nics (ISRM). Recently the International Society for Soil-Mechanics and Foundation Engineering (ISSMFE, 1977) published an extensive list of symbols. Although there are some deviations from this list because of our personal preference, we have generally tried to follow these recommenda- tions Units used in geotechnical engineering can be politely called a mess and, less politely, several worse things. There has developed in practice a jumbled mixture of cgs-metric, Imperial or British Engineering units and hybrid European metric units. With the introduction of the universal and consistent system of units, “Le Systéme International d'Unités” (SI) in the United States and Canada, we believe it is important that you learn to use those units in geotechnical engineering practice. However since British Engineering units are still commonly used, it is important that you become familiar with the typical values of both sets of units. To assist you with unit conversion where necessary, we have included a brief explanation of SI units as applied to geotechnical engineering in Appendix A.Index and Classification Properties of Soils 2.1 INTRODUCTION In this chapter we introduce the basic terms and definitions used by geotechnical engineers to index and classify soils. The following notation is used in this chapter. Symbol Dimension Unit Definition 4 = ~ ‘Activity (Eq. 2-23) G = = Coefficient of'curvature (Eq. 2-20) w ale ef 2)> 2 Be 0.893 V, 16 ™ Fig. Ex. 2.26 Index and Classification Properties of Solle can solve for V, directly, or Since V, = V, + V, + V,. we can solve for V,, since we know the other terms. V, = V,— V, — V, = 1.0 — 0.593 ~ 0.160 = 0.247 m? Once the phase diagram has been filled in, solution of the rest of the problem involves just plugging in the respective numbers into the ap- propriate definition equations. We recommend that when you make the computations, you write out the equations in symbol form and then insert the numbers in the same order as written in the equation. Also, it is a good idea to have the units accompany the calculations. Solving for the remainder of the required items is easy. From Eq. 2-9, = $ - 1éMe = 1.6Mg/m* pea or aoe = 0.686 ¥, oh 109 = 227+ 010 199 = 49.7% aa - ote = 39.3% ‘The saturated density p,q, is the density when all the voids are filled with water, that is, when S = 100% (Eq. 2-10). Therefore, if the volume of air V, were filled with water, it would weigh 0.247 m’ x | Mg/m? or 0.247 Mg. Then (0.247 Mg + 0:16 Mg) + 1-6 MB _ 2 1 Mg/m? Another, and perhaps even easier way to solve this example problem, is to assume V, is a unit volume, 1 m?. Then, by definition, M, = p, = 2.7 (when p, is assumed to be equal to 2.70 Mg/m*). The completed phase diagram is shown in Fig. Ex. 2.2c. Since w = M,,/M, = 0.10, M, = 0.27 Mg and M, = M, + M, = 2.97 ‘Mg. Also V, = M,, since p, = 1 Mg/m’; that is, 0.27 Mg of water occupies2.3 Solution of Phase Problems ” a volume of 0.27 m’. Two unknowns remain to be solved before we can proceed: they are V, and V,. To obtain these values, we must use the given information that p = 1.76 Mg/m’. From the definition of total density (Eq. 2-6), M, _ 2.97Mg p= 1.76 Mg/m? Solving for ¥,, 2ST ME 1 688 a? 1.76 Mg/m? Therefore V, = V,— Vy ~ V, = 1.688 ~ 0.27 — 1.0 = 0.418 You can use Fig. Ex. 2.2c to verify that the remainder of the solution is identical to the one using the data of Fig. Ex. 2.2b. Volume (m?) Mass (Ma) Va = 027 ™, 10 vy 270 | ™, Fig. Ex 226 EXAMPLE 2.3 Required: Express the porosity 1 in terms of the void ratio ¢ (Eq. 2-3a) and the void ratio in terms of the porosity (Eq. 2-3b). : Solution: Draw a phase diagram (Fig. Ex. 23a). For this problem, assume V,= 1 (units arbitrary). From Eq. 2-1, V, =e since V, = 1. Therefore V, = 1 + e. From Eq. 2-2, the definition of20 Index and Classification Properties of Solte A w s Fig. Ex. 238 A w s Fig. Ex 2.36 nis V,/V,, or n=—* (2-3a) l+e Equation 2-3b can be derived algebraically or from the phase dia- gram (Fig. Ex. 2.3b). For this case, assume V, = 1 From Eq. 2-2, V, = n since V, = 1. Therefore V, 2-1, the definition of e = V,/¥,. So —n. From Eq, ee (2-30) EXAMPLE 2.4 €=0.62, w=1s% p, 2.65 Mg/m? ¢. w for S = 100% 4. pay for S = 100% Solution: Draw phase diagram (Fig. Ex. 2.4). & Since no volumes are specified, assume V,= 1 mi. Just as in2.8 Solution of Phase Problems a Example 2.3, this makes the V, = e = 0,62 m’ and V,= 1 +e = 1.62 m, From Eq. 2-9, M, ny and M, = ,V, (from Eq. 2-7). So pv, _ Ps een oom ot = i since ¥, = Um in Fig, Ex24 2.65 ° ag 7 1686 Ma/m! Note: The relationship Pe if Pa Tae (2-12) is often very useful in phase problems. Volume (n°) ‘Mass (Mab | x Bis w Ble : i o al. Fig. Ex 24 — j_+ We know that M, = wM, (from Eq. 2-5) and M, = p,¥, _ vt wed, _ esl +) ~ ¥, I¥e since V, = 1m? Plug in the numbers. _ 2.65(1 + 0.15) Pr T+ 02 ‘This relationship is often useful to know. . (1 + w) (+e) = 1.88 Mg/m? (2-13)Index and Classification Properties of Solte Check: (2-14) 1.636 Mg/m? “15 You should verify that p, = p/(1 + w), which is another very useful relationship to remember. c. Water content for $= 100%. From Eq, 2-4, we know that V, = V, = 0.62 m’. From Eq. 2-8, M, = Vp, = 0.62 m? (1 Mg/m*) = 0.62 Mg. Therefore w for $= 100% must be My _ 0.62 Miscuway” Fr 7 355 7 02H 0F 23.4% 4d. Py, From Eq. 2-10, we know py, = (M, + M,)/V,, or 2.65 + 0.62 i 162 Check, by Eq. 2-13: _ PAL + w) _ 2.65(1 + 0.234) ite Te ‘= 2.019 or 2.02 Mg/m? 2.02 Mg/m? EXAMPLE 2.5 Required: Derive a relationship between S, e, w, and p,. Solution: Look at the phase diagram with V, = 1 (Fig. Ex. 2.5). From Eq. 2-4 and Fig. 2.5, we know that V, definitions of water content (Eq. 2-5) and p, (Eq. SV, = Se. Frora the -7), we can place the Volume Wass r Toe Ww 7 a w Me Fath w,¥, t “ft Y s M= 2M Foex26 of 42.3. Solution of Phase Problems as} equivalents for M, and M, on the phase diagram. Since from Eq, 2-8, M, = pV.» we now can write the following equation: My = PV, = WM, = wet, 7 p.Se = wp,v, Since V,= 1m’, 0, Se = wp, (2-15) Equation 2-15 is one of the most useful of all equations for phase problems. You can also verify its validity from the fundamental definitions of p,, S, e, w, and p, ‘Note that using Eq. 2-15 we can write Eq. 2-13 another way: (2-16) (2-17) EXAMPLE 2.6 Given: A silty clay soil with p, = 2700 kg/m’, S = 100%, and the water content = 46%, Required: Compute the void ratio e, the saturated density, and the buoyant or submerged density in kg/m’. Solution: Place given information on a phase diagram (Fig. Ex. 2.6) Assume V,= 1 m’; therefore M, = V,p, = 2700 kg. From Eq. 2-15, we can solve for e directly: _ 9.46 x 2700 pS 1000 1.0 But ¢ also equals V, since V, = 1.0; likewise M, = 1242 kg since M, is ee = 1.24224 Index and Classification Properties of Solis Volume (m?) te z/> aya w als + + $ Fig. Ex. 26 ‘S= 100% numerically equal to V, because p, = 1000 kg/m’. Now that all the unknowns have been found, we may readily calculate the saturated density (Eq. 2-10). (1242 + 2700) kg (also mama We.could also use Eq. 2-17 directly. p, + pee _ 2700 + 1000 (1.242) aca T+ 1.242 ‘When a soil is submerged, the actual unit weight is reduced by the buoyant effect of the water. The buoyancy effect is equal to the weight of the water displaced. Thus, in terms of densities, (Eqs. 2-11 and 2-17): 0" = Paar — Py = 1758 kg/m? — 1000 kg/m? = 758 kg/m? 1758 kg/m? or Ba + Put Tee Psa Pw Ite p’ = 758 kg/m? In this example, p’ is less than the density of water. Go back and look at Table 2-1 for typical values of p’. The submerged or buoyant density of soil will be found to be very important later on in our discussion of consolidation, settlement, and strength properties of soils. (2-18) In summary, for the easy solution of phase problems, you don’t have to memorize lots of complicated formulas. Most of them can easily be2.4 Soll Texture a derived from the phase diagram as was illustrated in the preceding exam- ples. Just remember the following simple rules: 1, Remember the basic definitions of w, ¢, p,, S, ete. 2. Draw a phase diagram. 3. Assume either V, = 1 or ¥, = 1, if not given. 4, Often use p,Se = »p,. 2.4 SOIL TEXTURE So far we haven't said very much about what makes up the “solids” part of the soil mass. In Chapter 1 we gave the usual definition of soil from an engineering point of view: the relatively loose agglomeration of mineral and organic materials found above the bedrock. We briefly described how weathering and other geologic processes act on the rocks at or near the earth's surface to form soil. Thus the solid part of the soil mass consists primarily of particles of mineral and organic matter in various sizes and amounts. The texture of a soil is its appearance or “feel,” and it depends on the relative sizes and shapes of the particles as well as the range or distribution of those sizes. Thus coarse-grained soils such as sands or gravels obviously appear coarse textured, while a fine-textured soil might be composed of predominantly very tiny mineral grains which are invisible to the naked eye. Silts and clay soils are good examples of fine-textured soils. The soil texture, especially of coarse-grained soils, has some relation to their engineering behavior. In fact, soil texture has been the basis for certain soil classification schemes which are, however, more common in agronomy than in soils engineering. Still, textural classification terms (gravels, sands, silts, and clays) are useful in a general sense in geotechnical engineering practice. For fine-grained soils, the presence of water greatly affects their engineering response—much more so than grain size or texture alone, Water affects the interaction between the mineral grains, and this may affect their plasticity and their cohesiveness. Texturally, soils may be divided into coarse-grained versus fine- grained soils. A convenient dividing line is the smallest grain that is visible to the naked eye. Soils with particles larger than this size (about 0.05 mm) are called coarse-grained, while soils finer than the size are (obviously) called fine-grained. Sands and gravels are coarse grained while silts and clays are fine grained. Another convenient way to separate or classify soils is according to their plasticity and cohesion (physics: cohesion—sticking28 Index and Classification Properties of Solts TABLE 2-2 Textural and Other.Characteristics of Soils Gravels, Sands Sits ‘Coarse grained Fine grained | Fine grained Can see individ Cannot see Cannot see ‘ual grains individual individual by eye grains ‘grains Characteristics: | Cobesionless Cohesionless | Cohesive ‘Nonplastic ‘Nonplastic Plastic Granular Granular = Effect of water Relatively unimportant | Important ‘Very important ‘on engineering (exception: loose sat- behavior: lurated granular mater- ils and dynamic loadings) Effect of grain size | Important Relatively Relatively distribution on ‘unimportant | unimportant casinering be together of like materials). For example, sands are nonplastic and non- cohesive (cohesionless) whereas clays are both plastic and cohesive. Silts fall between clays and sands: they are at the same time fine-grained yet nonplastic and cohesionless. These relationships as well as some general engineering characteristics are presented Table 2-2. You will need to obtain some practice, best done in the laboratory, in identifying soils according to texture and some of these other general characteristics such as plasticity and cohesiveness. Also you should note that the term clay refers both to specific minerals called clay minerals (discussed in Chapter 4) and to soils which contain clay minerals. The behavior of some soils is strongly affected by the presence of clay minerals. In geotechnical engineering, for simplicity such soils are usually called clays, but we really mean soils in which the presence of certain clay minerals affects their behavior. 2.5 GRAIN SIZE AND GRAIN SIZE DISTRIBUTION ‘As suggested in the preceding section, the size of the soil particle, especially for granular soils, has some effect on engineering behavior. Thus, for classification purposes, we are often interested in the particle or grain sizes present in a particular soil as well as the distribution of those sizes.2.5 Grain Size and Grain Size Distribution ar 7 i Jussi just Fig. 2.3 Grain size ranges according to several engineering soit classification systems (modified after Al-Hussaini, 1977), The range of possible particle sizes in soils is tremendous. Soils can range from boulders or cobbles of several centimetres in diameter down to ultrafine-grained colloidal materials. The maximum possible range is on the order of 10%, so usually we plot grain size distributions versus the logarithm of average grain diameter. Figure 2.3 indicates the divisions between the various textural sizes according to several common engineer ing classification schemes. It should be noted that traditionally in the28 Index and Classification Properties of Sots United States the units for the various sizes depend on the grain size. For materials greater than about $ mm (about 1/4 in.), inches are commonly used, although millimetres could be used just as well. Grain sizes between 5 mm and 0.074 mm are classified according to U.S. Standard sieve number, which of course can be related to a specific grain size as shown in Fig, 2.3. Soils finer than the No. 200 sieve are usually dimensioned in millimetres or, for the very fine-grained colloidal particles, in micrometres. How is the particle size distribution obtained? The process is called ‘mechanical analysis or the gradation test. For coarse-grained soils, a sieve analysis is performed in which a sample of dry soil is shaken mechanically through a series of woven-wire square-mesh sieves with successively smaller ‘openings. Since the total mass of sample is known, the percentage retained or passing each size sieve can be determined by weighing the amount of soil retained on each sieve after shaking. Detailed procedures for this test are specified by ASTM (1980), Designations C 136 and D 422. The corresponding AASHTO (1978) test standards are T 27 and T 88. The U.S. Standard sieve numbers commonly employed for the particle size analysis of soils are shown in Table 2-3. Since soil particles are rarely perfect spheres, when we speak of particle diameters, we really mean an equivalent particle diameter as determined by the sieve analysis. TABLE 2.3 U.S. Standard Sieve Sizes and ‘Their Corresponding Open Dimension US. Standard Sieve Opening Sieve No. (mm) 4 475 10 200 20 oss 40 0.425 60 025 100 ous 140 0.106 200 0.075 It turns out that the sieve analysis is impractical for sieve openings less than about 0.05 to 0.075 mm (No. 200 U.S. Standard sieve). Thus for the fine-grained soils, silts, and clays, the hydrometer analysis is commonly used. The basis for this test is Stoke’s law for falling spheres in a viscous fluid in which the terminal velocity of fall depends on the grain diameter and the densities of the grains in suspension and of the fluid. The grain diameter thus can be calculated from a knowledge of the distance and time of fall. The hydrometer also determines the specific gravity (or density) of the suspension, and this enables the percentage of particles of a certain equivalent particle diameter to be calculated. As with the sieve analysis,2.5 Grain Size and Grain Size Distribution 2 the percentage of total sample still in suspension (or already out of suspension) can therefore readily be determined. Detailed procedures for the hydrometer test are given by ASTM (1980), Designation D 422, and AASHTO (1978) Standard Method T 88. The US.BR. (1974) and U.S. Army Corps of Engineers (1970) also have similar standardized procedures for this test. The distribution of the percentage of the total sample less than a certain sieve size or computed grain diameter can be plotted in either a histogram or, more commonly, in a cumulative frequency diagram. The equivalent grain sizes are plotted to a logarithmic scale on the abscissa, whereas the percentage by weight (or mass) of the total sample either passing (finer than) or retained (coarser than) is plotted arithmetically on the ordinate (Fig. 2.4). Note that this figure could just as well be plotted with the smaller grain sizes going towards the right. Some typical grain size distributions are shown in Fig. 2.4. The well-graded soil has a good representation of particle sizes over a wide range, and its gradation curve is smooth and generally concave upward. On the other hand, a poorly graded soil would be one where there is either an excess or deficiency of certain sizes or if most of the particles are about the same size. The uniform gradation shown in Fig. 2.4 is an example of a poorly graded soil. The ‘gap-graded ot skip-graded soil in that figure is also poorly graded; in this case, the proportion of grain sizes between 0.5 and 0.1 mm is relatively low. We could, of course, obtain the usual statistical parameters (mean, median, standard deviation, etc.) for the grain size distributions, but this is more commonly done in sedimentary petrology than in soil mechanics. Of ‘course the range of particle diameters found in the sample is of interest. Besides that, we use certain grain diameters D which correspond to an equivalent “percent passing” on the grain size distribution curve. For example, Dio is the grain size that corresponds to 10% of the sample passing by weight. In other words, 10% of the particles are smaller than the diameter Dio. This parameter locates the grain size distribution curve (GSD) along the grain size axis, and it is sometimes called the effective size. ‘The coefficient of uniformity C, is a crude shape parameter, and it is defined as Deo ‘ c= 2-19) Do 2-19) where Day =grain diameter (in mm) corresponding to 60% passing, and Djo =grain diameter (in mm) corresponding to 10% passing, by weight (or mass). Actually, the uniformity coefficient is misnamed since the smaller the number, the more uniform the gradation. So it is really a coefficient of” Sieve analysis (US. Stendard sieve) 3in (sew 0) 2y6i3m Aq (weys Je8e00) paureies waaay g 8 8 1 Well graded 0.001 100 8 3 g (ssews 40) 2ybiom Aq (ueur souy) Bulssed ruadieg 20 ° Grain diameter (mm) Fig. 24 Typical grain size distributions.2.5 Grain Size and Grain Size Distribution ” “disuniformity.” For example, a C, = 1 would be a soil with only one grain size. Very poorly graded soils, for example, beach sands, have C,’s of 2 or 3, whereas very well-graded soils may have a C, of 15 or greater. Occasionally, the C, can range up to 1000 or so. As an example, the clay core material for Oroville Dam in California has a C, of between 400 and 500; the sizes range from large boulders down to very fine-grained clay particles. Another shape parameter that is sometimes used for soil classifica- tion is the coefficient of curvature defined as c= ol “(Do Peo) where Dyy = grain diameter (in mm) corresponding to 30% passing by weight (or mass). The other terms were defined previously. A soil with a coefficient of curvature between | and 3 is considered to be well graded as long as the C, is also greater than 4 for gravels and 6 for sands. (2-20) EXAMPLE 2.7 Given: The grain size distribution shown in Fig. 2.4. Required: Determine Dio, C,, and C, for each distribution. Solution: For Eqs. 2-19 and 2-20 we need Dig, Dyg, and Dep for each gradation curve in Fig. 2.4. a. Well-graded soil; simply pick off the diameters corresponding to 10%, 30%, and 60% passing. Dy = 0.02 mm, Dy = 0.6 mm, Dg = 9 mm From Eq. 2-19,32 Index and Classification Properties of Soll From Eq. 2-20, c= (Pw _ 067 “~ (Dio)(Deo) ~ (0.02)8) Since C, > 15 and C, is between | and 3, this soil is indeed well graded. 9. Gap-graded soil; use same procedure as in (a): : Dy = 0.022, Dy = 0.052, Dy = 1.2 From Eq. 2-19, Dw 12 Ge De 7 doa = 5 From Eq. 2-20, ete) (0.052)? © Dia Deo) (0.022)(1.2) ~ Even though by the uniformity coefficient criterion, this soil is well graded, it fails the coefficient of curvature criterion. Therefore it is indeed poorly graded. €. Uniform soil; use same procedure as in (a): Diy =03, Dy = 043, Dey = 0.55 From Eq. 2-19, Dey _ 0-55 eras (Dio) (0.437 (DioMDwo) — (0:3)(0.55) This soil is still poorly graded even though the C, is slightly greater than unity; the C, is very small. 2.6 PARTICLE SHAPE ‘The shape of the individual particles is at least as important as the grain size distribution in affecting the engineering response of granular soils. It is possible to quantify shape according to rules developed by sedimentary petrologists, but for geotechnical engineering purposes such\ 2.6 Particle Shape 3 refinements are rarely warranted. Only a qualitative shape determination is usually made as part of the visual classification of soils. Coarse-grained soils are commonly classified according to the shapes shown in Fig. 2.5. Rounded Ge Subangular Angular Fig. 2.5 Typical shapes of coarse-grained bulky particles. (Photograph by M. Surendra) ‘A distinction can also be made between particles that are bulky and those which are needlelike or flaky. Mica flakes are an excellent example of the latter, and Ottawa sand is an example of the former. Cylinders of each differ drastically in behavior when compressed by a piston. The bulky grains hardly compress at all, even when in a very loose state, but the mica flakes will compress, even under small pressures, up to about one-half of their original volume. When we discuss the shear strength of sands, you will learn that grain shape is very significant in determining the frictional characteristics of granular soils.2.7 ATTERBERG LIMITS AND CONSISTENCY INDICES We mentioned (Table 2-2) that the presence of water in the voids of a soil can especially affect the engineering behavior of fine-grained soils. Not only is it important to know how much water is present in, for example, a natural soil deposit (the water content), but we need to compare or scale this water content against some standard of engineering behavior. This is what the Atterberg limits do—they are important limits of engineering behavior. If we know where the water content of our sample is relative to the Atterberg limits, then we already know a great deal about the engineer- ing response of our sample. The Atterberg limits, then, are water contents at certain limiting or critical stages in soil behavior. They, along with the natural water content, are the most important items in the description of fine-grained soils. They are used in classification of such soils, and they are useful because they correlate with the engineering properties and engineer- ing behavior of fine-grained soils. The Atterberg limits were developed in the early 1900's by a Swedish soil scientist, A. Atterberg (1911). He was working in the ceramics in- dustry, and at that time they had several simple tests to describe the plasticity of a clay, which was important both in molding clay into bricks, for example, and to avoid shrinkage and cracking when fired. After many experiments, Atterberg came to the realization that at least two parameters were required to define plasticity of clays—the upper and lower limits of plasticity. In fact, he was able to define several limits of consistency or behavior and he developed simple laboratory tests to define these limits. ‘They are: 1, Upper limit of viscous flow. 2. Liquid limit—lower limit of viscous flow. 3. Sticky limit—clay loses its adhesion to a metal blade. 4. Cohesion limit—grains cease to cohere to each other. 5. Plastic limit—lower limit of the plastic state. 6. Shrinkage limit—lower limit of volume change. He also defined the plasticity index, which is range of water content where the soil is plastic, and he was the first to suggest that it could be used for soil classification. Later on, in the late 1920's K. Terzaghi and A. ‘Casagrande (1932b), working for the U.S. Bureau of Public Roads, stan- dardized the Atterberg limits so that they could be readily used for soils classification purposes. In present geotechnical engineering practice we usually use the liquid limit (LL or w,), the plastic limit (PL or wp), and sometimes the shrinkage limit (SL or w,). The sticky and the cohesion limits are more useful in ceramics and agriculture.‘oeuodses Ulens-seens pezi1e/0U06 ou) Se OM Ee 11S B Jo SeIBIS SNOUIBA OLN BUYKOYS WNnUAUOD JUeIUOD JAE OZ “Bid 2 ‘ ‘ m>0 on o>n (1m ! “piney | ioe sare36 Index and Classification Properties of Solis Since the Atterberg limits are water contents where the soil behavior changes, we can show these limits on a water content continuum as in Fig 2.6. Also shown are the types of soil behavior for the given ranges of water contents. As the water content increases, the state of the soil changes from a brittle solid to a plastic solid and then to a viscous liquid. We can also show on the same water content continuum the generalized material response (stress-strain curves) corresponding to those states. You may recall the curves shown in Fig. 2.7 from fluid mechanics, where the shear velocity gradient is plotted versus the shear stress. Depend- ing on the water content, it is possible for soils to have a response represented by all of those curves (except possibly the ideal Newtonian liquid). Note, too, how different this response is from the stress-strain behavior of other engineering materials such as steel, concrete, or wood. (Shear velocity gradient) ° Fits G Fig. 2.7 Behavior of several materials including soils over a range of water contents. Atterberg’s original consistency limit tests were rather arbitrary and not easily reproducible, especially by inexperienced operators. As men- tioned, Casagrande (1932b, 1958) worked to standardize the tests, and he developed the liquid limit device (Fig. 2.8) so that the test became more|__ 10 men gage for checking height of fall of cup Casagrande grooving too! Brass cup Remalded soil sample Height of fail of cup Oy Fig. 2.8 (a) Schematic diagram of the Casagrande liquid limit device and grooving tool; dimensions in millimetres. (b) Cut groove prior to turning the crank. (c) After turning the crank to apply sulficient blows of the cup to close the groove 13 mm. (4) Plastic limit threads. Parts (2) through (c) after Hansbo (1975). (Photographs by M. Surendra.)38 Index and Classification Properties of Soils operator-independent. He defined the LL as that water content at which a standard groove cut in the remolded soil sample by a grooving tool (Figs. 2.8a, b) will close over a distance of 13 mm (} in.) at 25 blows of the LL cup falling 10 mm on a hard rubber or micarta plastic base (Fig. 2.8¢). In practice, it is difficult to mix the soil so that the groove closure occurs at exactly 25 blows, but Casagrande found that if you plot the water contents of tests where you get closure at other blow counts versus the logarithm of the number of blows, you get a straight line called the flow curve. Where the flow curve crosses 25 blows, that water content is defined as the liquid limit. The plastic limit test is somewhat more arbitrary, and it requires some practice to get consistent and reproducible results. The PL is defined ‘as the water content at which a thread of soil just crumbles when it is carefully rolled out to a diameter of 3 mm (} in.). It should break up into segments about 3 to 10 mm (i in. to } in.) long. If the threads can be rolled to a smaller diameter, then the soil is too wet (above the PL); if it crumbles before you reach 3 mm (! in.) in diameter, then you are past the PL. Properly rolled out PL threads should look like those shown in Fig. 28d. Even though the liquid limit and plastic limit tests appear simple, both tests do take some practice to get consistent results. In Sweden, the fall-cone test is used to determine the liquid limit (Hansbo, 1957). It seems to give more consistent results than the Casagrande device, especially for Swedish clays, and it is somewhat simpler and faster to use. Karlsson (1977) presents an excellent discussion of the reliability of both procedures. Sometimes a one-point liquid limit test can be used because, for soils of similar geologic origin, the slopes of the flow curves are similar. Then all you have to do is obtain the water content w, of the sample with closure of the groove at any blow count 7, and use the following relationship nant LL= w( 35) (2-21) where tan B is the slope of the flow curve. For best results the blow count should be between about 10 and 40. Lambe (1951), U.S. Army Corps of Engineers (1970), and Karlsson (1977) provide good discussions of the one-point liquid limit test. You may have noticed that we have not mentioned the ASTM procedures for the Atterberg limits tests. We do not recommend the ASTM procedures because, for one thing, they require that the limits be con- ducted on air-dried specimens. For some soils, such a procedure will give very different results than if the limits are conducted at the natural water content (Karlsson, 1977). The other problem with ASTM is the grooving tool for the liquid limit test, It does not allow for any control of the height2.7 Atterberg Limits and Conslatency indices ae of the groove, and therefore it will give inconsistent results. For this reason, we recommend the Casagrande grooving tool (Fig. 2.8) be used. ‘The range of liquid limits can be from zero to 1000, but most soils have LL’s less than 100. The plastic limit can range from zero to 100 or more, with most being less than 40. Even though the Atterberg limits are really water contents, they are also boundaries between different engineer- ing behaviors, and Casagrande (1948) recommends that the values be reported without the percent sign. They are numbers to be used to classify fine-grained soils, and they index soil behavior. You will, however, see the limits reported both ways and using both symbols: LL and PL, and w, and wp with a percent sign. The other Atterberg limit sometimes used in geotechnical engineering practice, the shrinkage limit, is discussed in some detail in Chapter 6. ‘We mentioned earlier that Atterberg also defined an index called the plasticity index to describe the range of water content over which a soil was plastic. The plasticity index, PI or J,, therefore is numerically equal to the difference between the LL and the PL, or PI = LL—- PL (2-22) ‘The PI is useful in engineering classification of fine-grained soils, and many engineering properties have been found to empirically correlate with the PI. When we first started the discussion on the Atterberg limits, we said that we wanted to be able to compare or scale our water content with some defined limits or boundaries or engineering response. In this way, we would know if our sample was likely to behave as a plastic, a brittle solid, or even possibly a liquid. The index for scaling the natural water content of a soil sample is the liquidity index, LI or J, is defined as w, — PL LI = (2-23) where w, is the natural water content of the sample in question. If the LI is less than zero then, from the water content continuum of Fig. 2.6, you would know that the soil will have a brittle fracture if sheared. If the LI is between zero and one, then the soil will behave like a plastic. If LI is greater than one, the soil will be essentially a very viscous liquid when sheared. Such soils can be extremely sensitive to breakdown of the soil structure. As long as they are not disturbed in any way, they can be relatively strong, but if for some reason they are sheared and the structure of the soil breaks down, then they literally can flow like a liquid. There are deposits of ultra sensitive (quick) clays in Eastern Canada and Scandinavia, Figure 2.9 shows a sample of Leda clay from Ottawa, Ontario, in both the40 Ilcation Properties of Solls Fig. 2.9 (a) Undisturbed and (b) thoroughly remolded sample of clay from Ottawa, Ontario. (Photograph courtesy of the Wn of Building Research, National Research Council of ‘Canada. Hand by D. C. MacMillan.) undisturbed and remolded states at the same water content. The undis- turbed sample can carry a vertical stress of more than 100 kPa; when thoroughly remolded, it behaves like a liquid. It wasn't emphasized previously, but the limits are conducted on thoroughly remolded soils, and when we discuss the structure of clays in Chapter 4, we will see that the natural structure of a soil very strongly governs its engineering behavior. So then how come the Atterberg limits work? They work empirically; that is, they correlate with engineering properties and behavior because both the Atterberg limits and the engineer- ing properties are affected by the same things. Some of these “things” include the clay minerals, the ions in the pore water, the stress history of the soil deposit, etc. And these factors are discussed in detail in the chapter on soil structure (Chapter 4). For now, just accept that these very simple, arbi- trary, and empirical Atterberg limits are most useful in classifying soils for engineering purposes and that they correlate quite well with the engineer- ing behavior of soils.2.8 ACTIVITY In 1953, Skempton defined the activity A of a clay as PL clay fraction (2-24) where the clay fraction is usually taken as the percentage by weight of the soil less than 2 um. Clays which have an activity around 1 (0.75 1.25 are active clays. Activity has been useful for certain classification and engineering property correlations, especially for inactive and active clays Also, there is fair/good correlation of the activity and the type of clay mineral (Chapter 4). However, the Atterberg limits alone are usually sufficient for these purposes, and the activity provides no really new information. PROBLEMS 2-1. A water content test was made on a sample of silty clay. The weight of the wet soil plus container was 17.53 g, and the weight of the dry soil plus container was 14.84 g, Weight of the empty container was 7.84 g, Calculate the water content of the sample. 2-2. During a plastic limit test, the following data were obtained for one of the samples: Wet weight + container = 22.12 g Dry weight + container = 20.42 g Weight of container = = 1.508 Whaat is the PL of the soil? 2-3. A sample of fully saturated clay weighs 1350 g in its natural state and 975 g after drying, What is the natural water content of the soil? 2-4, For the soil sample of Problem 2-3, compute (a) void ratio and (b) Porosity. 2-5. For the soil sample of Problem 2-3, compute (a) the total or wet density and (b) the dry density. Give your answers in Mg/m’, kg/m’, and Ibf/ft?, a26. 24, 2.8, 29, 2.13. 2.14, 215. blanks in the phase Index and Classification Properties of Sol A 1 m® sample of moist soil weighs 2000 kg. The water content is 10%. Assume p, is 2.70 Mg/m?. With this information, fill in all jagram of Fig. P2-6. 7 ft Air Water Ales Soil - Fig. P26 For the information given in Problem 2-6, calculate (a) the void ratio, (b) the porosity, and (c) the dry density. ‘The dry density of a compacted sand is 1.82 Mg/m? and density of the solids is 2.67 Mg/m?. What is the water content of the material when saturated? A 100% saturated soil has a total density of 2050 kg/m? and a water content of 25%. What is the density of the solids? What is the dry density of the soil? What is the water content of a fully saturated soil with a dry density of 1.70 Mg/m?? Assume p, = 2.71 Mg/m’. A dry quartz sand has a density of 1.68 Mg/m?. Determine its density when the degree of saturation is 75%. The density of solids for quartz is 2.65 Mg/m’. The dry density of a soil is 1.65 Mg/m? and the solids have a density of 2.68 Mg/m’. Find the (a) water content, (b) void ratio, and (c) total density when the soil is saturated. A natural deposit of soil was found to have a water content of 20% and to be 90% saturated. What is the void ratio of this soil? The void ratio of clay soil is 0.5 and the degree of saturation is 70%. Assuming the density of the solids is 2750 kg/m’, compute (a) the water content and (b) dry and wet densities in both SI and British Engineering units. The volume of water in a sample of moist soil is 0.056 m’. The volume of solids V, is 0.28 m°. Given that the density of soil solids p, is 2590 kg/m’, find the water content.Probleme 43 2-16, Verify from first principles that: 9) p= o,(F*) = ares T+e @) pm ad) © wi ase Mens @n=7h 2-17, Derive an expression for p, in terms of the porosity 1 and the water content w for (a) a fully saturated soil and (b) a partially saturated soil. 2-18. Derive an expression for (a) dry density, (b) void ratio, and (c) degree of saturation in terms of p, p,, p,, and w. 2-19. Develop a formula for (a) the wet density and (b) the buoyant density in terms of the water content, the density of the soil solids, and the density of water. 2-20. From Archimedes’ principle show that Eq. 2-11, p’ = pay ~ Py iS the same as (p, ~ p,)/(I + €)- 2-21. The “chunk density” method is often used to determine the unit weight (and other necessary information) of a specimen of irregular shape, especially of friable samples. The specimen at its natural water content is (1) weighed, (2) painted with a thin coat of wax or paraffin (to prevent water from entering the pores), (3) weighed again (W, + W,,,), and (4) weighed in water (to get the volume of the sample + wax coating—remember Archimides?). Finally, the natural water content of the specimen is determined. A specimen of silty sand is treated in this way to obtain the “chunk density.” From the information given below, determine the (a) wet density, (b) dry density, (c) void ratio, and (d) degree of saturation of the sample. Given: Weight of specimen at natural water content = 181.8 g Weight of specimen + wax coating 215.9 Weight of specimen + wax in water = 589 Natural water content 25% Soil solid density. p, =2700 kg/m* Wax solid density. py. = 940 ke /m' Hint: Use a phase diagram. 2-22. The total volume of a soil specimen is 80000 mm? and it weighs 145 g. The dry weight of the specimen is 128 g, and the density of the soil solids is 2.68 Mg/m’. Find the (a) water content, (b) void ratio, (©) porosity, (a) degree of saturation, (e) wet density, and (f) dry density. Give the answers to parts (e) and (f) in both SI and British Engineering units.2.23, 2-24. 2.25. 2-26. 2.27. 2-28. 2.29. 2-30. 231. Index and Classification Properties of Soils The values of minimum ¢ and maximum e for a pure silica sand were found to be 0.46 and 0.66, respectively. What is the corre- sponding range in the saturated density in kg/m? ‘A 588 cm® volume of moist sand weighs 1010 g. Its dry weight is 918 g and the density of solids is 2670 kg/m*. Compute the void ratio, the porosity, water content, degree of saturation, and the total density in kg/m’. ‘A sample of saturated glacial clay has a water content of 47%. On the assumption that p, = 2.70 Mg/m?, compute the void ratio, porosity, and saturated density. A sensitive volcanic clay soil was tested in the laboratory and found to have the following properties: (@) p = 1.28 Mg/m* (b) e= 9.0 (©) S = 95% (@) p, = 2.75 Me/m? (©) w = 311% In rechecking the above values one was found to be inconsistent with the rest. Find the inconsistent value and report it correctly. The saturated density ¥.,, of a soil is 1351bf/ft°. Find the buoyant density of this soil in both Ibf/ft? and kg/m’. A sand is composed of solid constituents having a density of 2.68 Mg/m?. The void ratio is 0.58. Compute the density of the sand when dry and when saturated and compare it with the density when submerged. ‘A sample of natural glacial till was taken from below the ground water table. The water content was found to be 55%. Estimate the wet density, dry density, buoyant density, porosity, and void ratio. Clearly state any necessary assumptions. Calculate the maximum possible porosity and void ratio for a collection of (a) ping pong balls (assume they are 30 mm in diameter) and (b) tiny ball bearings 0.3 mm in diameter. A cylinder contains 500 cm? of loose dry sand which weighs 750 g, and under a static load of 200 kPa the volume is reduced 1%, and then by vibration it is reduced 10% of the original volume. Assume the solid density of the sand grains is 2.65 Mg/m*. Compute the void ratio, porosity, dry density, and total density corresponding to each of the following cases: (a) Loose sand. (b) Under static load (©) Vibrated and loaded sand.Probleme “ 2:32. 2:33. ‘The natural water content of a sample taken from a soil deposit was found to be 11.5%. It has been calculated that the maximum density for the soil will be obtained when the water content reaches 21.5%. Compute how many grams of water must be added to each 1000 g of soil (in its natural state) in order to increase the water content to 21.5%. On five-cycle semilogarithmic paper, plot the grain size distribution curves from the following mechanical analysis data on the six soils, A through F. Determine the effective size as well as the uniformity coefficient and the coefficient of curvature for each soil. Determine also the percentages of gravel, sand, silt, and clay according to (a) ASTM, (b) AASHTO, (c) USCS, and (4) the British Standard. US Standard Sieve No. Percent Passing by Weight or Particle Sie SoA SoilB Soll€ SollD Soll Soll F 75mm @ in) 100 100 38a) 0 = » 6 10 9 95 @) 6 No.4 fe eee 100 No. 10 2 82 7% 10 89 No. 20 — © — 8 = No. 40 se 8 8 8 No. 60 =— Bb 5 om No. 100 Ss = 0 = No. 140, = 6 3 4 © No. 200 45 2 = 51 100 nm Oe sla ai a 9 204m 2 9 BD crc? 10nm 1 BB 2% 8 Sam illest een ta) 8 7 2am - = 4 - 2 tam eater: - » Note: Missing data is indicated by a dash in the column. (a) Explain briefly why it is preferable, in plotting GSD curves, to plot the grain diameter on a logarithmic rather than an arithmetic scale. (b) Are the shapes of GSD curves comparable (for example, do they have the same C, and C,) when plotted arithmetically? Ex- plain.2.35. 2:36. 2.37. Index and Classiication Properties of Sols The soils in Problem 2-33 have the following Atterberg limits and natural water contents. Determine the PI and LI for each soil and comment on their general activity. Property SollA SollB Sol Sol SolE SollF Wai 2 4 4 1 nm LL 1B 35 35 = © PL 8 v 18 NP a Comment on the validity of the results of Atterberg limits on soils Gand H. SolG Sol iL 3 38 PL 20 a SL 25 = The following data were obtained from a liquid limit test on a silty clay. No.of Blows Water Content, % 35 a 29 418 21 BS 15 49 Two plastic limit determinations had water contents of 23.1 and 23.6%. Determine the LL, PI, the flow index, and the toughness index. The flow index is the slope of the water content versus log of, number of blows in the liquid limit test, and the toughness index is the PI divided by the flow index.three Soil Classification 3.1 INTRODUCTION From the discussion in Chapter 2 on soil texture and grain size distributions, you should have at least a general idea about how soils are classified. For example, in Sec. 2.4 we described sands and gravels as coarse-grained soils, whereas silts and clays were fine grained. In Sec. 2.5, we showed the specific size ranges on a grain size scale (Fig. 2.3) for these soils according to the standards of ASTM, AASHTO, etc. Usually, how- ever, general terms such as sand or clay include such a wide range of engineering characteristics that additional subdivisions or modifiers are required to make the terms more useful in engineering practice. These terms are collected into soil classification systems, usually with some specific engineering purpose in mind. A soil classification system represents, in effect, a language of com- munication between engineers. It provides a systematic method of cate- Zorizing soils according to their probable engineering behavior, and allows engineers access to the accumulated experience of other engineers. A classification system does not eliminate the need for detailed soils investi- gations or for testing for engineering properties. However, the engineering Properties have been found to correlate quite well with the index and classification properties of a given soil deposit. Thus, by knowing the soil classification, the engineer already has a fairly good general idea of the way the soil will behave in the engineering situation, during construction, under structural loads, etc. Figure 3.1 illustrates the role of the classifica- tion system in geotechnical engineering practice. Many soil classification systems have been proposed during the past 50 years or so. As Casagrande (1948) pointed out, most systems used in a“ ‘oll Ctassiticatton Classification and Index Properties (w.e, p, §, GSO, LL, Pi, ete.) | Classification System (Language) Engineering Properties (permeability, compressibility, shrink swell shear strength, etc.) Engineering Purpose (highways, airfields, foundation, dams, etc.) Fig. 3.1 Role of classification system in geotechnical engineering practice, civil engineering had their roots in agricultural soil science. This is why the first systems used by civil engineers classified soil by grain size or soil texture. Atterberg (1905) apparently was the first to suggest that something other than grain size could be used for soil classification. To this end, in 1911 he developed his consistency limits for the behavior of fine-grained soils (Sec. 2.7), although at that time for agricultural purposes. Later the U.S. Bureau of Public Roads based the classification of fine-grained soils almost entirely on the Atterberg limits and other simple tests. Casagrande (1948) describes several other systems that have been used in highway engineering, airfield construction, agriculture, geology, and soil science. Today, only the Unified Soil Classification System (USCS) and the American Association of State Highway and Transportation Officials (AASHTO) system are commonly used in civil engineering practice. The Unified Soil Classification System is used mostly by engineering agencies of the U.S. Government (U.S. Army Corps of Engineers and U.S. Depart- ment of the Interior, Bureau of Reclamation) and many geotechnical engineering consulting firms and soil testing laboratories. With slight modification this system is also in fairly common use in Great Britain and elsewhere outside the United States. Nearly all of the state Departments of Transportation and Highways in the United States use the AASHTO system, which is based upon the observed behavior of soils3.2 The Unified Soll Classification System (USCS) “0 under highway pavements. The Federal Aviation Administration (FAA) of the U.S. Department of Transportation had its own soil classification system for the design of airport pavements, but it now uses the Unified Soil Classification System. Once you become familiar with the details, both the USCS and AASHTO systems are easy to use in engineering practice. 3.2 THE UNIFIED SOIL CLASSIFICATION SYSTEM (USCS) This system was originally developed by Professor A. Casagrande (1948) for use in airfield construction during World War II. It was modified in 1952 by Professor Casagrande, the U.S. Bureau of Reclama- tion, and the U.S. Army Corps of Engineers to make the system also applicable to dams, foundations, and other construction (U.S. Army En- gineer Waterways Experiment Station, 1960). The basis for the USCS is that coarse-grained soils can be classified according to their grain size distributions, whereas the engineering behavior of fine-grained soils is primarily related to their plasticity. In other words, soils in which “fines” (silts and clays) do not affect the engineering performance are classified according to their grain size characteristics, and soils in which fines do control the engineering behavior are classified according to their plasticity characteristics, Therefore, only a sieve analysis and the Atterberg limits are necessary to completely classify a soil in this system. The four major divisions in the USCS are indicated in Table 3-1. They are (1) coarse-grained, (2) fine-grained, (3) organic soils, and (4) peat. Classification is performed on the material passing the 75 mm sieve, and the amount of “oversize” material is noted on the drill logs or data sheets. Particles greater than 300 mm equivalent diameter are termed boulders, while materials between 75 mm and 300 mm are called cobbles. Coarse- grained soils, sands, and gravels are those having 50% or more material retained on the No. 200 sieve. These fractions have been arbitrarily but conveniently subdivided as shown in Table 3-1. Fine-grained soils are those having more than 50% passing the No. 200 sieve. The highly organic soils and peat can generally be identified visually. The symbols in Table 3-1 are combined to form soil group symbols which correspond to the names of typical soils as shown in Table 3-2. The coarse-grained soils are subdivided into gravels and gravelly soils (G) and sands and sandy soils (S). The gravels are those having the greater Percentage of the coarse fraction (particles larger than 4.75 mm diameter) retained on the No. 4 sieve, and the sands are those having the greater Portion passing the No. 4 sieve. Both the gravel (G) and the sand (S)TABLE 3-1 USCS Definitions of Particle Size, Size Ranges, and Symbols ‘Soll Fraction ‘or Component Symbol Size Range Boulders None Greater than 300 mm Cobbles None 75 mm to 300 mm (0) Coarse-grained soils: Gravel c 75 mm to No. 4 sieve (4.75 mm) Coarte 75 mmo 19 mm Fine 19mm toNo. 4 sieve (4.75 mm) Sand s No. 4 (4.75 mm) to No. 200 (0.075 mm) Coarse No.4 (4.75 mm) to No. 102.0 mm) Medium No. 102.0 mm) to No. 40 (0.425 mm) Fine No. 40 (0.425 mm) to ‘No. 200 (0.075 mm) Q) Fine-grained soils: Fines Less than No. 200 sieve (0.075 mm) sit M (No specitic grain size— use Aterberg limits) Chay c (No specific grain size— use Atterberg limits) @) Organic Soils: ° (No specific grain size) Peat Pt (No specitic grain size) Gradation Symbols ‘Liquid Limit Symbols Well-graded, W High LL, Poorly-graded, P Low LLL groups are divided into four secondary groups, GW and SW, GP and SP, GM and SM, GC and SC, depending on the grain size distribution and nature of fines in the soils. Well-graded (W) soils have a good representa- tion of all particle sizes whereas the poorly graded (P) soils are either uniform or skip- or gap-graded (Fig. 2.4). Whether a gravel or sandy soil is well graded can be determined by plotting the grain size distribution curve and computing the coefficient of uniformity C, and the coefficient of curvature C,. These coefficients are defined in Chapter 2 as a3 9) and the coefficient of curvature is Di "Dx Da (2:20)3.2. The Unified Soll Classification System (USCS) st where Deg = grain diameter at 60% passing, Dy = grain diameter at 30% passing, and Dyo = grain diameter at 10% passing by weight (or mass). Gradation criteria for gravelly and sandy soils are shown in Table 3-2 (column 6). The GW and SW groups are well-graded gravelly and sandy soils with less than 5% passing the No. 200 sieve. The GP and SP groups are poorly graded gravels and sands with little or no nonplastic fines. The fine-grained soils, those having more than 50% passing the No. 200 sieve, are subdivided into silts [M for the Swedish terms mo (= very fine sand) and mila (= silt)] and clays (C) based on their liquid limit and plasticity index. Organic soils (O) and peat (Pt) are also included in this fraction although, as shown in Table 3-1, no grain size range is specified. Fine-grained soils are silts (M) if their liquid limits and plasticity indices plot below the A-line on Casagrande’s (1948) plasticity chart (Fig. 3.2). The fines are clays (C) if the LL and PI plot above the A-line. The A-line generally separates the more claylike materials from those that are silty and also the organics from the inorganics. The exception is organic clays (OL and OH) which plot below the A-line. However, these soils do behave similarly to soils of lower plasticity, The silt, clay, and organic fractions are further subdivided on the basis of relatively low (L) or high (H) liquid limits. The dividing line between the low and high liquid limits has been arbitrarily set at 50. Representative soil types for fine-grained soils are also shown in Fig. 3.2. This figure, columns 4 and 5 of Table 3-2, and Table 3-3, will be helpful in the visual identification and classification of fine- grained soils. You can see from Fig. 3.2 that several different soil types tend to plot in approximately the same area on the LL-PI chart, which means that these soils tend to have about the same engineering behavior. This is why the Casagrande chart is so useful in the engineering classifica- tion of soils. For example, Casagrande (1948) observed the behavior of soils at the same liquid limit with plasticity index as compared with their behavior at the same plasticity index but with an increasing liquid limit, and he obtained the following results: Soils at Equal LL Soils at Equal PI Characteristic with Increasing PI with Increasing LL Dry strength Increases Decreases ‘Toughness near PL Increases Decreases Permeability Decreases Increases Compressibility ‘About the same Increases Rate of volume change Decreases es Toughness near the PL and dry strength are very useful visual classification properties, and they are defined in Table 3-3. The other characteristics are engineering properties, and they are discussed in great82 Soll Ciassitication TABLE 3-2 Unified Soil Classification System* r Te EE Tg [em [ eueomennimnsnmnin | SIRMUSMRR nt on, g Phare” | entor | tconmency g 7 ce | “pnncays gavel cure oer Iesum torn) Monee % | weston vm | Sotosaarin ary sow [oownne| Sue 5 8 Seat neon oom ii | een pen |e on | Oamesensenetn joorsosin| Secerer | Satie ‘Aiter US. Army Engineer Waterways Experiment Station (1960) and Howard (1977).3.2 The Unified Soll Classification System (USCS) TABLE 3-2 Continued : i ee? 5 x y 3 7 = b waxy &| | | | | i | (us ese wos) pedojenep) e8dh) os exneyuesoides (Bienes Bumous “YeUD Ai ua pinoy puesesey Ze MOH PLE ‘By6L '@PUBIE Bia {—— pue —a1aeb ou! of ov 09 xepuy AvDaselg3.2 The Unified Soll Classification System (USCS) ss TABLE 3-3 Field Identification Procedures for Fine-Grained Soils or Fractions These procedures are to be performed on the minus No. 40 sieve size particles, approximately 0.4 mm, For field classification pur- poses, screening is not intended; simply remove by hand the coarse particles that interfere with the tests. Dilataney (eaction to shaking) Dry Stength (crushing characteristics): After removing particles larger than No. 40 sieve size, prepare a pat of moist soil with a volume of about $ cm. Add enough water if necessary to make the soil soft but ‘ot sticky. Place the pat in the open palm of ‘one hand and shake vigorously against the cother hand several times. A positive reaction ‘consists of the appearance of water on the surface of the pat, which changes to a livery consistency and becomes glossy. When the sample is squeezed between the fingers, the ‘water and gloss disappear from the surface, the pat stiffens, and finally it cracks or crum- bles. The rapidity of appearance of water during shaking and of its disappearance dur- ing squeezing assist in identifying the char- acter of the fines in a soil Very fine clean sands give the quick est and most distinct reaction, whereas a plastic clay has n0 reaction. Inorganic sits such as a typical rock flour show a mod- erately quick reaction After removing particles larger than No. 40 sieve size, mold a pat of soil to the consistency of pulty, adding water if neces- sary. Allow the pat to dry completely by ‘oven, sun, or ar, and then test its strength by breaking and crumbling between the fingers. This strength is a measure of the character and quantity of the colloidal fraction con- tained in the soil. The dry strength increases with increasing plasticity. High dry strength is characteristic for clays of the CH group. A typical inorganic silt possesses only very slight dry strength. Silty fine sands and silts have about the same slight dry strength but can be distinguished by the feel when powdering the dried speci- men. Fine sand feels gritty, whereas a typical silt has the smooth feel of flour. “Toughness (Consistency near plastic Limit: ‘Alter removing particles larger than the No. 40 sieve size, «specimen of soil about 4 in. cube in size is molded to the consistency of puty. If too dry, water must be added and, it sticky, the specimen should be spread out in a thin layer and allowed to lose some moisture by evaporation. Then the specimen is rolled out by hand on a smooth surface or between the palms into a thread about 3 mm in diameter. The thread is thea folded and refolded repeatedly. During this manipulation the moisture content is gradually reduced and the specimen stiffens, finally loses its plasticity, and crumbles when the plastic limit is reached. ‘Atier the thread crumbles, the pices should be lumped together and a slight kneading action continued until the lump crumbles. ‘The tougher the thread near the plastic limit and the stiffer the lump when it finally crumbles, the more potent the colloidal clay fraction in the soil. Weakness of th thread atthe plastic limit and quick loss of coherence of the lump below’ the plastic limit indicate either inorganic clay of low plasticity or materials such as kaolin-type clays and organic clays which occur below the A-line Highly organic clays have a very weak and spongy feel atthe plastic limit. “After U'S. Army Engineer Waterways Experiment Station (1960) and Howard (1977).56 ‘Soll Classitication detail later in this book. For now, just rely on your general knowledge and ingenuity to figure out what those words mean. ‘The upper limit line (U-line) shown in Fig. 3.2 indicates the upper range of plasticity index and liquid limit coordinates found thus far for soils (A. Casagrande, personal communication). Where the limits of any soil plot to the left of the U-line, they should be rechecked. Some highly active clays such as bentonite may plot high above the A-line and close to the Ucline. It is shown in Chapter 4 that Casagrande’s plasticity chart can ‘even be used to identify qualitatively the predominant clay minerals in a soil. Coarse-grained soils with more than 12% fines are classified as GM and SM if the fines are silty (limits plot below the A-line on the plasticity chart) and GC and SC if the fines are clayey (limits plot above the A-line). Both well-graded and poorly graded materials are included in these two groups. Soils having between 5% and 12% passing the No. 200 sieve are classed as “borderline” and have a dual symbol. The first part of the dual symbol indicates whether the coarse fraction is well graded or poorly graded. The second part describes the nature of the fines. For example, a soil classified as a SP-SM means that it is a poorly graded sand with between 5% and 12% silty fines. Similarly a GW-GC is a well-graded gravel with some clayey fines that plot above the A-line. Fine-grained soils can also have dual symbols. Obviously, if the limits plot within the shaded zone on Fig. 3.2 (PI between 4 and 7 and LL between about 12 and 25), then the soil classifies as a CL-ML. Howard (1977) makes the practical suggestion that if the LL and PI values fall near the A-line or near the LL = 50 line, then dual symbols should be used. Possible dual symbols then are: ML-MH cLCH oL-oH cLML con cHMH CHOH Borderline symbols can also be used for soils with about 50% fines and coarse-grained fractions. In this case possible dual symbols are: OMML GM-MH sec. Gc-cH SM-ML SM-MH sCCL SC-CH3.2. The Unified Soll Classification System (USCS) 87 Figure 3.3 is a practical guide for borderline cases of soil classifica- tion UNIFIEO SOIL CLASSIFICATION SYSTEM (Borderline Classifications) coms Grtned Sots +—-——+ Fin rained Sl i me ow |g Gravel { lg sik 4Mi ca wi owe | et sind fe | 8 Clay 4H | ou os 2 a5 505 10 percent passing the 0200 Seve Note: Only two group symbols may be used to describe a soil Borderline classifications ean exist within each of the above groups. Fig. 93 Guide for borderline cases of soil classification (after Howard, 1977) A step-by-step procedure for USCS classification of soils, conveni- ently presented in Fig. 3.4, shows a process of elimination of all the possibilities until the only one left indicates the specific classification. The following steps, adapted from the Corps of Engineers, may help in this process (U.S. Army Engineer Waterways Experiment Station, 1960). Clas- sification should be done in conjunction with Table 3-2 and Fig. 3.4: 1. Determine if the soil is coarse grained, fine grained, or highly organic. This is done by visual inspection and/or by determining the amount of soil passing the No. 200 sieve. If coarse grained: a. Perform a sieve analysis and plot the grain size distribution curve. Determine the percentage passing the No. 4.sieve and classify the soil as gravel (greater percentage retained dn No. 4) or sand (greater percentage passing No. 4). b. Determine the amount of material passing the No. 200 sieve. If less than 5% passes the No. 200 sieve, examine the shape of the grain size curve; if well graded, classify as GW or SW; if poorly graded, classify as GP or SP. c. If between 5% and 12% of the material passes the No. 200 sieve, it is a borderline case, and the classification should have dualss Soll Classification ica roe) | pas i r Fig. 2.4 Auxiliary laboratory identification procedure (after USAEWES, 1960).3.2 The Unified Soll Classincation System (USCS) socy Soll Classification symbols appropriate to grading and plasticity characteristics (GW-GM, SW-SM, etc). 4. If more than 12% passes the No. 200 sieve, perform the Atter- berg limits on the minus No. 40 sieve fraction. Use the plastic- ity chart to determine the correct classification (GM, SM, GC, SC, GM-GC, or SM-SC). 3. If fine grained: a. Perform Atterberg limits tests on minus No. 40 sieve material If the liquid limit is less than 50, classify as L, and if the liquid limit is greater than 50, classify as H. b. For L: If the limits plot below the A-line and the hatched zone on the plasticity chart, determine by color, odor, or the change in liquid limit and plastic limit caused by oven-drying the soil, whether it is organic (OL) or inorganic (ML). If the limits plot in the hatched zone, classify as CL-ML. If the limits plot above the A-line and the hatched zone on the plasticity chart (Fig. 3.2), classify as CL. c. For H: If the limits plot below the A-line on the plasticity chatt, determine whether organic (OH) or inorganic (MH). If the limits plot above the A-line, classify as CH. 4. For limits which plot in the hatched zone on the plasticity chart, close to the A-line or around LL = 50, use dual (border- ine) symbols as shown in Fig. 3.3. ‘Although the letter symbols in the USCS are convenient, they do not completely describe a soil or soil deposit. For this reason, descriptive terms should also be used along with the letter symbols for a complete soil classification. Table 3-4 from U.S. Army Engineer Waterways Experiment Station (1960) provides some useful information for describing soils. In the case of all soils, such characteristics as color, odor, and homogeneity of the deposit should be observed and included in the sample description. For coarse-grained soils such items as grain shape, mineralogical content, degree of weathering, in situ density and degree of compaction, and presence or absence of fines should be noted and included. Adjectives such as rounded, angular, and subangular are commonly used to describe grain shape (see Fig. 2.5). The in situ density and degree of compaction is normally obtained indirectly by observing how difficult the material is to excavate or to penetrate with devices called penetrometers. Terms such as very loose, loose, medium, dense, and very dense are used to describe in situ density. A granular deposit which can, for example, be excavated readily3.2. The Unified Soll Classification System (USCS) Cy TABLE 34 Information Required for Describing Soils* ‘Coarse-grained soils: ‘or undisturbed soils add informa- tion on stratification, degree of compact- ness, cementation, moisture conditions and drainage characteristics. Give typical name. Indicate ap- proximate percentages of sand and gravel, maximum size, angularity, surface condi- tion, and hardness of the coarse grains, local or geologic name and other pertinent descriptive information, and symbol in parentheses. Example: ‘Silty sand, gravelly. About 20% hard, angular gravel’ particles 12 mm_ maxi- mum size, rounded and subangular sand Fine-grained soils: ‘Give typical name. Indicate degree and character of plasticity, amount and maximum size of coarse grains, color in wet condition, odor if any, local or geo- logic name, and other pertinent descriptive information, and symbol in parentheses. For undisturbed soils add informa- tion on structure, stratification, con- sistency in undisturbed and remolded states, moisture and drainage conditions. Example: Clayey silt, brown, slightly plastic, small percentage of fine sand, numerous vertical root holes, firm and dry in place, loess. (ML). grains coarse to fine, about 15% nonplastic fines with low dry strength, well com- pacted and moist in place, alluvial sand, (SM). Note: Be prepared for wide variations in soil description among organizations and testing laboratories. They all have their own ways of doing things. ‘After USS. Arty Engineer Waterways Experiment Station (1960), by hand would be considered very loose, whereas a deposit of the same material which requires power tools for excavation would be described as very dense or perhaps cemented. For the fine-grained fraction, natural water content, consistency, and remolded consistency should be noted in the sample description. Con- sistency in the natural state corresponds in some respects to degree of compaction in coarse-grained soils and is usually evaluated by noting the ease by which the deposit can be excavated or penetrated. Such terms as very soft, soft, medium, stiff, very stiff, and hard are employed to describe consistency. (Sometimes the word firm is used synonymously with the term stiff.) Fine-grained soils may be additionally described by using the tests explained in Table 3-3 for dilatancy, toughness, and dry strength. Other techniques for visual classification of soils should be learned and practiced in the laboratory. Excellent descriptions of visual classification and identi- fication procedures are found in the U.S.B.R. (1974) Earth Manual, Ap- pendix E-3, and ASTM (1980) Designation D 2488.EXAMPLE 3.1 Given: * Sieve analysis and plasticity data for the following three soils. Soil 1, Soil2, Soil, Sieve Sie % Finer. % Finer’, % Finer No.4 8 7 100 No. 10 92 90 100 No. 40 %6 0 100 No. 100 8 8 ” No. 200 o 5 7 LL 20 = 14 PL 15 = a Pr 3 ‘Pt n *Nonplastic. Required: Classify the three soils according to the Unified Soil Classification System. Solution: Use Table 3-2 and Fig. 3.4. 1, Plot the grain size distribution curves for the three soils (shown in Fig. Ex. 3.1). 2. For soil 1, we see from the curve that more than 50% passes the No. 200 sieve (60%); thus the soil is a fine-grained soil and the Atterberg limits are required to further classify the soil. With LL = 20 and PI = 5, the soil plots in’ the hatched zone on the plasticity chart. Therefore the soil is a CL-ML. 3. Soil 2 is immediately seen to be a coarse-grained soil since only 5% asses the No. 200 sieve. Since 97% passes the No. 4 sieve, the soil is a sand rather than a gravel. Next, note the amount of material passing the No. 200 sieve (5%). From Table 3-2 and Fig. 3.4, the soil is “borderline” and therefore has a dual symbol such as ‘SP-SM or SW-SM depending on the values of C, and C.. From the grain size distribution curve, Fig. Ex. 3.1, we find that Dj = 0.71 mm, Djy = 0.34 mm, and Dy = 0.18 mm. The coefficient ofvexa bia 8 = sens re a atte ma Sos a Seer as a ae a T T 5 a lee F ox i ve 2 ol ei i le = a i fa - es + oy —» 2 2 105, 1 i 2 ol € i i vm lou & H ; le ol |“ Solt Classification, uniformity C, is and the coefficient of curvature C, is (DY _ (0.34) Dy X Deo 018X071 =091~1 For a soil to be considered well graded, it must meet the criteria shown in column 6 of Table 3-2; it does not, so the soil is considered poorly graded and its classification is SP-SM. The soil is SM because the fines are silty (nonplastic). A quick glance at the characteristics for soil 3 indicates the soil is fine grained (97% passes the No. 200 sieve). Since the LL is greater than 100 we cannot directly use the plasticity chart (Fig. 3.2). Use instead the equation for the A-line on Fig. 3.2 to determine if the soil is a CH or MH. PI = 0.73(LL ~ 20) = 0.73(124 — 20) = 75.9 Since the PI is 78 for soil 3, it lies above the A-line and thus the soil is classified as a CH. ES 3.3 THE AASHTO SOIL CLASSIFICATION SYSTEM In the late 1920's the U.S. Bureau of Public Roads (now the Federal Highway Administration) conducted extensive research on the use of soils especially in local or secondary road construction, the so-called “farm-to- market” roads. From that research the Public Roads Classification System was developed by Hogentogler and Terzaghi (1929). The original system was based on the stability characteristics of soils when used as a road surface or with a thin asphalt pavement. There were several revisions since 1929, and the latest in 1945 is essentially the present AASHTO (1978) system. The applicability of the system has been extended considerably; AASHTO states that the system should be useful for determining the relative quality of soils for use in embankments, subgrades, subbases, and bases. But you might keep in mind its original purpose when using the system in your engineering practice. (See Casagrande, 1948, for some comments on this point.)3.3. The AASHTO Soll Classification System 6 TABLE 3-5 AASHTO Definitions of Gravel, Sand, and Silt-Ciay Soil Fraction ‘Size Range Boulders “Above 75 mm. Gravel 75 mm to No, 10 sieve (2.0 mm) Coarse sand ‘No. 10 2.0 mm) to No. 40 (0.425 mm) Fine sand ‘No. 40 (0.425 mm) to No, 200 (0.075 mm) Silt-clay (combined Material passing the 0.075 mm (No. 200) silt and clay) sieve Soil fractions recognized in the AASHTO system are listed in Table 3-5. Boulders should be excluded from the sample to be classified, but as with the USCS the amount of boulders present should be noted. Fines are silty if they have a PI less than 10 and clayey if the PI is greater than 10. The AASHTO system classifies soils into eight groups, A-I through ‘A-8, and it includes several subgroups. Soils within each group are evaluated according to the group index, which is calculated by an empirical formula. The only tests required are the sieve analysis and the Atterberg limits. Table 3-6 illustrates the current AASHTO (1978) soil classification. Granular materials fall into classes A-1 to A-3. A-I materials are well graded, whereas A-3 soils are clean, poorly graded sands. A-2 materials are also granular (less than 35% passing the No. 200 sieve), but they contain a significant amount of silts and clays. A-4 to A-7 are fine-grained soils, the silt-clay materials. They are differentiated on the basis of their Atterberg. limits. Figure 3.5 can be used to obtain the ranges of LL and PI for groups A-4 to A-7 and for the subgroups in A-2. Highly organic soils including peats and mucks may be placed in group A-8. As with the USCS, classification of A-8 soils is made visually. The group index is used to further evaluate soils within a group. It is based on the service performance of many soils, especially when used as pavement subgrades. It may be determined from the empirical formula given at the top of Fig. 3.6, or you may use the nomograph directly. Using the AASHTO system to classify soils is not difficult. Once you have the required test data, proceed from left to right in the chart of Table 3-6, and find the correct group by the process of elimination. The first group from the left to fit the test data is the correct AASHTO classifica- tion. A complete classification includes the group index to the nearest whole number, in parentheses, after the AASHTO symbol. Examples are A-2-6(3), A-4(5), A-6(12), A-7-5(17), ete. 7 Figure 3.7 will be helpful in classifying’ soils according to the AASHTO system.66g 209 of mana 71 wa ee drosaae 9 1-V J Xp AVDHINE OF SHURE TT URS J0 07 tebe Ht doodaRES-L-y JO 2p KDHE, fom 9¢ “um gg “orm 9g “om 9 oma’ fq peel ‘cbt HINO WoneLodsueN| pur Kong aS Jo UONNDOETY HIT OS spnager se Hones eeD, poe pur one uy “esa 0015 sree wanusocs yoy Jo md tr, sa st rst porary TY ww ew ow ro (=o one rune ADS S2innW e1866/B8y-105 PUR BIOS J0 LONROUIEEN OE SHEVA3.3 The AASHTO Soll Classification System or 70 60 50 40}— 30}-— Plasticity index 20 10 “| a4 las. 010 2030 40 50 60 70 80 90 100 Liquid timit Fig. 3.5 _Atterberg limits ranges for subgrade subgroups A-4, A-5, A6, land 4-7. Note that Casagrande's A-line and U-line have been superimm posed on the chart (after Liu, 1970, and Al-Hussaini, 1977). EXAMPLE 3.2 Given; The following data for soils 4 and 5. See Fig. Ex. 3.1 for the grain size distribution curves. Soil, Soil, SieveSie % Finer Finer 3 a 96 18 » 9 No. 100 p 5 No. 200 70 LL ° _ PL 4 = Pr 25 NP Required: Classify the soils according to the AASHTO Soil Classification System.Group index (GI) = (F ~ 35)(0.2 + 0,005(LL - 40)} + 0.01(F ~ 15)(PI~ 10), where F = % passing 0.075 mm sieve, LL = liquid limit, and PI = plasticity in When working with A:26 and A-2-7 subgroups the Partial Group Index (PGI! is determined from the PI only. When the combined partial group indices are negative, the group index shoul reported as zero. 50: 40 30 Partial group index Example: {82% passing 0.075 mm sieve 8.9 for LL LL=38 7.4 for PL Pr = 21 16 Fig. 36 Group index chart (after AASHTO, 1978). ©American Associa- tion of State Highway and Transportation Officials, 1978. Used by permis- dex ) id be Ba es RVG ond R27 50 70 Percent passing No. 200 sieve 80 100| (0161 ‘017 s0ye) wershg UoNROUIEERID NOS OLHSWY OUI 40) se:npenosd UONBoYISseIO A:oVEsoqe) KrBHIXNE 10) YEU LE “Sl hee ae & ia)‘TABLE 3-7 Comparable Soil Groups in the AASHTO and USCS Systems* Soil Group ‘Comparable Soil Groups in in AASHTO System Unified Most Possible but System Probable Possible Improbable cow Aba = A24, ADS, A26, ADT or Abe Abb om A26 4 AS, ADS,ADT 6, ATS, A746, Aba cc A26,A27 = ADS AG Ad, A-T-6, sw Ade SP Abe 24, ADS, A246, ADT SM A26,A-4, AG, ATS, AS AT, Alea sc A246, ATS AA, ATG ML AAAS AG ATS = co Ab, AT6 AS - OL AAAS AGATS, = A168 MH ATS, AS - ATS cH Ans = oH = A Pt = _ - ‘After Liu (1970).TABLE 3-7 Continued Soil Group ‘Comparable Soil Groups in in Unified System AASHTO Most Posibie But ‘System Probable Possible Improbable Aa GW, GP SW, SP GM, SM Abb sw, SP, or - GM, SM AB SP = Sw, GP Add GM,SM GC, SC ow, GP sw, Ars OM,SM = GW, GP, ‘SW, SP A26 Gc, sc GM, SM ow, GP SW, SP Aad oM,cc, — SM, SC ‘SW, SP Aa ML, OL CL, SM, om, Gc sc AS OH,MH, ‘SM, GM ML, OL As cL ML, OL, GC,GM, sc sM Ans OH,MH = ML,OL, GM, SM, cH Gc, SC AIG cH, cL ML, OL, OH, MH, sc GC, GM, SM Solution: 1, Because more than 35% of soil 4 passes the No. 200 sieve, from Table 3-6 we see that the soil is an A-4 or higher. Since the LL is 49, the soil is either A-5 or A-7. with a PI of 25, the soil is an A-7. A check of Fig. 3.5 shows the soil is classified as an A-7-6. 2, Because soil 5 has less than 35% passing the No. 200 sieve, it is granular. (A glance at Fig. Ex. 3.1 provides the same information!) Proceeding from left to right in Table 3-6, we see that the first group from the left that meets the criterion is A-l-a. n3.4 COMPARISON OF THE USCS AND AASHTO CLASSIFICATION SYSTEMS There are several significant differences between the USCS and AASHTO soil classification systems, which is not surprising considering the differences in their history and purpose. You can see the differences in the treatment of coarse-grained soils by comparing Table 3-1 with 3-5. The major differences in the fine-grained soils are shown in Fig. 3.5, where we have superimposed both the A-line and the U-line on the LL-PI chart. AASHTO (1978) actually plots LL versus PI, but we have turned the chart 90° for easy comparison with the Casagrande plasticity chart (Fig. 3.2). The differences are significant. Also, use of PI = 10 as the dividing line between silty and clayey soils seems rather arbitrary and probably does not realistically relate to the engineering properties of fine-grained soils. ‘Al-Hussaini (1977) discusses several other significant differences between the two systems. Table 3-7 shows a comparison of the two systems in terms of the probable corresponding soil groups. PROBLEMS 3-1. Classify soils 4 and 5 in Example 3.2 according to the Unified Soil Classification System. Explain your steps. Compare with Table 3-7 and Example 3.1. 3-2. Classify soils 1, 2, and 3 in Example 3.1 according to the AASHTO Soil Classification System. Compare with Table 3-7 and Example 3.2. 3-3. Given the grain size distribution curves of Problem 2-33 and the Atterberg limits data of Problem 2-35, classify soils A through F using the USCS and AASHTO soil classification systems. 3-4. For the data below, classify the soils according to the USCS. (@) 100% material passed No. 4 (b) 65% material retained on sieve, 25% retained on No. No. 4 sieve, 32% material 200 sieve. retained on No. 200 sieve. Fines exhibited: Gr3G=L Medium to low plasticity. Dilatancy—none to very slow. Dry strength—medium to high.Problems (c) 100% passed No. 4 sieve, 90% passed No. 200 sieve. Dry strength—low to medium. Dilatancy—moderately quick. LL = 23, PL = 17. (©) 100% material passed No. 4 sieve, 20% retained on No. 200 sieve. Fines exhibit high plasticity. Dilatancy—none. Dry strength—high. (g) 5% material retained on No. 4 sieve, 70% passed No. 4 sieve and retained on No. 200 sieve. Fines exhibited medium dry strength, medium toughness. Dilatancy—none. LL = 25, PL = 15. 3 (@) 5% retained on No. 4 sieve, 10% passed No. 4 and re- tained on No. 200 sieve. Fines exhibited low plasticity and high dilatancy. () 90% material passed No. 4 sieve, and retained on No. 200 sieve. Ga3Gel. 10% passed No. 200 sieve. Fines exhibited high dilatancy. (h) 70% material retained on No. 4 sieve, 27% retained on No. 200 sieve. C= 5, C= 15. 3-5. For the soils of Problem 3-4, estimate the compressibility, permeabil- ity, and toughness. 3-6. Grain size distributions and Atterberg limits are given for 16 soils in the six graphs comprising Fig. P3.6. Classify the soils according to (a) USCS and (b) AASHTO systems. (All data from USAEWES, 1960.)™ gages = Percent finer by weight (or mass) Bsas US, Standard sive size Bin. 219. No.4 No.10_No.40 No. 200 10 Sno = o L >= 10 04 001 ‘0.001 Gaain size (mm) Se ee Garve 1” Uniform fine snd; nonplastic. Curve 2: Poorly graded gravelly sand mixture; nonplastic. Approximately 7% fines makes this a borderline soil, symbol SP-SM. Curve 3: Coarse to medium sand: nonplastic. Approaching uniform gradation a USS. Standard sieve size ain bo, 4 No.10_No-40__No. 200 100 - _ 9 3 Eo 5 70 3 0 : 2 0 i «0 2» E 0 ° Goin sie (mn) ala alee Gone 1 Orgenie cy (al ats. Li=95, I-39 Game 2 Alva slay wit gn matter, L186, 71-27 Gare 3 Oroare sty LUTE P-SS ature eter content 33, P19 (oven dried). Fig. P3-6a US. Stindard sieve size io. 4 No. 10 No.40 No. 200 3 wf! & wol+ 2 aor = a |+—+— b xf 10 fs ° an : 100 6 ie Toor Grin sie toh < Eee eee oe Girve § Glayey snd: L1=29, PIO. Poorly gaded mir of sand clay ad in at sand Girve 2 Lioock ond snd mixture; LU=28, PL. Poorly graded US. Standard seve size ain J Nod No.10_No.40__ No. 200 100 _ fF 3 2 3» . % colt + 2 sol! ' = 40 ' a0 20 |! ! wr ' 0 ii See 166 6 v8 at ot Too Grin siz (mm) mv] meee | Curve 1: Silty gravely sand; nonplastic. Mieaceous sit stabilized with sandy chert gravel Curve 2" Mixture of gravel sand and fine silty sand; nonplastic. Poorly graded mixture; note absence of coarse and medium sand. Curve: Silty fine and, LL=22, PIeS, Fig. P3-6 Continued 5Percent finer by weight (or mas! US, Standard sieve size Bin FIM No.4 No.10 No.4 _ No. 200 Grain ize (mm) Tow [te [eon] weren [ Curve 1: Pit run gravel; nonplastic; well graded; small percentage of fines, Curve 2: Sandy gravel; nonplastic; no fines. o US. Standard sieve size 3 in, 20. No.4 No. 10_No.40__No. 1002 No.4 No. 10_ No. 40 200 Percent finer by weight (or mass) asses 70 or Grain size (mm) T001 Sal curve 1 Gurve 2: Uniform sandy s Clayey silt; LL= 46, PI =16. LL 30, Pre. Gurve 3: Sandy silt; LU= 34, PI=3 General. Note curves 2 and 3 have about the same plasticity but vary in grain size distribution. 76 Fig. P3-6 Continued,four Clay Minerals and Soil Structure 4.1 INTRODUCTION At this stage it is useful to again define the term clay. Clay can refer to specific minerals such as kaolinite or illite, as are discussed in detail in this chapter. However, in civil engineering, clay often means a clay soil—a soil which contains some clay minerals as well as other mineral con- stituents, has plasticity, and is “cohesive.” Clay soils are fine grained, as indicated in Chapter 2 (Table 2-2), but not all fine-grained soils are cohesive or clays. Silts are both granular and fine grained. The individual silt grains, like clays, are invisible to the naked eye, but silts are noncohe- sive and nonplastic, Rock flour is another example of a very fine-grained cohesionless soil ‘Also remember that certain characteristics of granular soils such as the grain size distribution and the grain shape affect the engineering behavior of these soils. On the other hand, the presence of water, with a few important exceptions, is relatively unimportant in their behavior. In contrast, for clay soils the grain size distribution has relatively little influence on the engineering behavior, but water markedly affects their behavior. Silts are an “in between” material. Water affects their behavior —they are dilarant—yet they have little or no plasticity (PI ~ 0), and their strengths, like sands, are essentially independent of water content. ‘As we indicate in this chapter, clay minerals are very small particles which are very active electrochemically. The presence of even a small amount of clay minerals in a soil mass can markedly affect the engineering, “properties of that mass. As the amount of clay increases, the behavior of the soil is increasingly governed by the properties of the clay. When the clay content is about 50%, the sand and silt grains are essentially floating in a clay matrix and have little effect on the engineering behavior.nm Clay Minerals and Soll Structure In this chapter, we briefly describe the important clay minerals, how they are identified, and how they interact with water and with each other. We also, describe some of the latest thinking about soil fabric and struc- ture, concepts which are fundamentally important for a good understand- ing of cohesive soil behavior. Finally, cohesionless soil structures and the concept of relative density are discussed. Oily one new symbol is introduced in this chapter. ‘Symbol Dimension ‘Unit Definition Dior ly = @ Relative density oF density index 4.2 CLAY MINERALS Clay minerals are very tiny crystalline substances evolved primarily from chemical weathering of certain rock-forming minerals. Chemically, they are hydrous aluminosilicates plus other metallic ions. All clay minerals are very small, colloidal-sized crystals (diameter less than | zm), and they can only be seen with an electron microscope. The individual crystals look like tiny plates or flakes, and from X-ray diffraction studies scientists have determined that these flakes consist of many crystal sheets which have a repeating atomic structure. In fact, there are only two fundamental crystal sheets, the fetrahedral or silica, and the octahedral or alumina, sheets. The particular way in which these sheets are stacked, together with different bonding and different metallic ions in the crystal lattice, constitute the different clay minerals. The tetrahedral sheet is basically a combination of silica tetrahedral units which consist of four oxygen atoms at the comers, surrounding a single silicon atom. Figure 4.1a shows a single silica tetrahedron; Fig. 4.1b shows how the oxygen atoms at the base of each tetrahedron are combined to form a sheet structure. The oxygens at the bases of each tetrahedron are in one plane, and the unjoined oxygen corners all point in the same direction. A common schematic representation of the tetrahedral sheet which is used later is shown in Fig. 4.le. A top view of the silica sheet showing how the oxygen atoms at the base of each tetrahedron belong to two tetrahedrons and how adjacent silicon atoms are bonded is shown in Fig. 4.1d. Note the hexagonal “holes” in the sheet. The octahedral sheet is basically a combination of octahedral units consisting of six oxygen or hydroxyls enclosing an aluminum, magnesium, iron, or other atom. A single octahedron is shown in Fig, 4.2a, while Fig. 4.2b shows how the octahedrons combine to form a sheet structure. The ows of oxygens or hydroxyls in the sheet are in two planes. Figure 4.2c isZN Cc) © Oxygens in plane above silicons + Silicons @ —_Oxygens linked to form network — Outline of bases of silica tetrahedra — Outline of hexagonal silica network (two dimensional); also indicates bonds from silicons to oxygens in lower plane (fourth bbond from each silicon is perpendicular to plane of paper) (@ Fig, 4.1 (@) Single silica tetrahedron (after Grim, 1959). (b) tsomet- fic view of the tetrahedral or silica sheet (after Grim, 1959). (c) ‘Schematic representation of the silica sheet (after Lambe, 1953). (4) ‘Top view of the silica sheet (after Warshaw and Roy, 1961) a schematic representation of the octahedral sheet which we use later. For a top view of the octahedral sheet showing how the different atoms are shared and bonded see Fig. 4.2d. Substitution of different cations in the octahedral sheet is rather common and leads to different clay minerals. Since the ions substituted are approximately the same physical size, such substitution is called isomor- phous. Sometimes not all the) octahedrons contain a cation, which results in ‘a somewhat different crystalline structure with slightly different physical properties and a different clay mineral. If all the anions of the octahedral sheet are hydroxyls and two-thirds of the cation positions are filled with aluminum, then the mineral is called gibbsite. If magnesium is substituted co20 Clay Minerale and Soll Structure = Hydroxyls or @ Aluminunts, magnesiums, ete ‘oxygens al al © © Hydronyisin upper plane © Atuminums Vacant octahedral positions (would be filled in brucite ayer) © Hyaroxyls in tower plane Outline of those faces of alumina octahedra paralle! to lower plane of hydroxyls Outline of those faces of vacant octahedra paraliel to lower plane of hydroxyls Bonds from aluminums to ‘a Irony (6 from each Fig. 42 (@) Single aluminum (or magnesium) octahedron (after Grim, 1959), (b) Isometric view of the octahedral sheet (after Grim, ‘f959). (e) Schematic representation of the octahedral or alumina (or magnesia) sheet (after Lambe, 1953). (6) Top view of the octahedral Sheet (after Warshaw and Roy, 1961) for the aluminum in the sheet and it fills all the cation positions, then the mineral is called brucite. The variations in the basic sheet structures make up the dozens of clay minerals which have been identified. All clay minerals consist of the two basic sheets which are stacked together in certain unique ways and with certain cations present in the octahedral and tetrahedral sheets. For engineering purposes it is usually sufficient to describe only a few of the more common clay ‘minerals which are found in clay soils.42 Clay Minerate a i 5 Al + 4 [vt AL ~ Fig. 43 Schematic diagram of the structure of Kaolinite (after Lambe, 1953), Si SSS Kaolinite consists basically of repeating layers of one tetrahedral (ilica) sheet and one octahedral (alumina or gibbsite) sheet. Because of the stacking of one layer of each of the two basic sheets, kaolinite is called a 1:1 clay mineral (Fig. 4.3). The two sheets are held together in such a way that the tips of the silica sheet and one of the layers of the octahedral sheet form a single layer, as shown in Fig. 4.4. This layer is about 0.72 nm thick © omveens @ Hytroxyis © Aurinums © OSilicons Fig. 44 Atomic structure of kaolinite (after Grim, 1950).a2 Clay Minerals and Soll Structure Fig. 4.5 Scanning electron micrograph of a well-crystallized kaolinite from Georgia. The Jength of the light bar is 5 zm (photograph by R. D. Holtz) and extends indefinitely in the other two directions. A kaolinite crystal, then, consists of a stack of several layers of the basic 0.72 nm layer. Successive layers of the basic layer are held together by hydrogen bonds between the hydroxyls of the octahedral sheet and the oxygens of the tetrahedral sheet. Since the hydrogen bond is very strong, it prevents hydration and allows the layers to stack up to make a rather large crystal A typical kaolin crystal can be 70 to 100 layers thick. Figure 4.5 is a scanning electron micrograph (SEM) of kaolinite. Kaolinite is the primary constituent in china clay; it is also used in the paper, paint, and pharmaceutical industries. For example, as a phar- maceutical it is used in Kaopectate and Rolaids. Another 1:1 mineral related to kaolinite is Aalloysite. It differs from kaolinite in that when it was formed it somehow became hydrated between the layers, causing a distortion or random stacking in the crystal lattice so that it is tubular in shape (Fig. 4.6). The water can easily be driven out from between the layers by heating or even air drying, and the process is irreversible. That is, the halloysite will not rehydrate when water is added. Halloysite, although not very common, occasionally plays an important Fig. 4.6 Scanning electron micro- graph of halloysite trom Colorado. The length of the light bar is 5 xm (photograph by R. D. Holtz)42 Clay Minerale J engineering role. Classification and compaction tests made on air-dried samples can give markedly different results than tests on samples at their natural water content. If the soil will not be air dried in the field, it can be extremely important that laboratory tests be carried out at the field water contents so that the results will have some validity. Montmorillonite, also sometimes called smectite, is an important mineral composed of two silica sheets and one alumina (gibbsite) sheet (Fig. 4.7). Thus montmorillonite is called a 2:1 mineral. The octahedral sheet is between the two silica sheets with the tips of the tetrahedrons combining with the hydroxyls of the octahedral sheet to form a single layer, as shown in Fig. 4.8. The thickness of each 2:1 layer is about 0.96 nm, and like kaolinite the layers extend indefinitely in the other two directions. Because the bonding by van der Waals’ forces between the tops of the silica sheets is weak and there is a net negative charge deficiency in the octahedral sheet, water and exchangeable ions can enter and separate ‘Si al HO layersand Le —_ exchangeable cations ‘ si \ al - | 0.96 am Fig. 47 Schematic diagram of the structure of montmorillonite (after Lambe. 1953).ry Clay Minerals and Soll Structure © oxyaens @ Hysroxvs — @ Aluminum, ion, magnesium O and @ Silicon, occasionally aluminum Fig. 4.8 Atomic structure of montmorillonite (after Grim, 1959) the layers. Thus montmorillonite crystals can be very small (Fig. 4.9), but at the same time they have a very strong attraction for water. Soils containing montmorillonite are very susceptible to swelling as they change (increase) water content, and the swelling pressures developed can easily damage light structures and highway pavements. Montmorillonite is also the primary constituent in drilling mud and kitty litter, and it has many4.2 Clay Minerals 35 Fig. 49. Scanning electron micrograph of Na-montmorilionite from Wyoming. The length of the light bar is 5 um (photograph by R. 0. Holtz) other important industrial and pharmaceutical applications. It is even used in chocolate bars! Illite, discovered by Prof. R. E. Grim of the University of Illinois, is another important constituent of clay soils. It also has a 2:1 structure similar to montmorillonite, but the interlayers are bonded together with a potassium atom. Remember the hexagonal hole in the silica sheet (Fig. 4.1d)? It has almost exactly the right diameter so that a potassium atom just fills that hexagonal hole and rather strongly bonds the layers together (Fig. 4.10). In addition, there is some isomorphous substitution of aluminium for silicon in the silica sheet. lites have a crystal structure similar to the mica minerals but with less potassium and less isomorphous substitution; thus they are chemically much more active than the other micas. Figure 4.11 is a SEM of illite. Chlorite, relatively common in clay soils, is made of repeating layers of a silica sheet, an alumina sheet, another silica, and then either a gibbsite (Al) or brucite (Mg) sheet (Fig. 4.12). It could be called a 2:1:1 mineral, Chlorite can also have considerable isomorphous substitution and be missing an occasional brucite or gibbsite layer; thus it may be susceptible to swelling because water can enter between the sheets. Generally, how- ever, it is significantly less active than montmorillonite. ‘As mentioned previously, there are literally dozens of clay minerals, with virtually every conceivable combination of substituted ions, interlayer water, and exchangeable cations. Some of the more important from an engineering viewpoint include vermiculite, which is similar to montmoril- lonite, a 2:1 mineral, but it has only two interlayers of water. After it is dried at high temperature, which removes the interlayer water, “expanded” vermiculite makes an excellent insulation material. Another clay mineral, attapulgite, Fig. 4.13, does not have a sheet structure but is a chain silicate; it consequently has a needle or rodlike appearance. Mixed-layer minerals0.96 nm 4 Fig. 4.10 Schematic diagram of ilite (atter Lambe, 1953) Fig. 4.11 Scanning electron ‘micrograph of ilite from Fithian, Minois. The length of the Nant bar is 5 ym (photograph by RD. Holtz)‘Sh Al si Brucite Gibbsite si a Fig. 4.12 Schematic diagram of chlorite (after Mitchell, 1976), Fig. 4.13 Scanning electron micro- {graph of attapulgite trom Florida. The length of the space between the bars of light is 0.5 pm (photograph by R. D. Holtz) a28 Clay Minerals and Soll Structure are relatively common; they would include, for example, montmorillonite mixed with chlorite or illite. Because allophane is an alumino-silicate, it is often classified as a clay mineral. However, it is amorphous, which means it thas no regular crystalline structure. Under specialized conditions of weathering, it may be a locally important constituent of clay soils. 4.3 IDENTIFICATION OF CLAY MINERALS Since the clay minerals are so very small, their identification by the usual optical mineralogical techniques used in geology is not possible so other means must be employed to identify them. From your engineering materials courses, you may remember that materials with regular or repeating patterns of crystal structure will diffract X-rays. Different miner- als with different crystalline structures will have different X-ray diffraction patterns, and in fact these different patterns were how the minerals were identified in the first place. The patterns for the common minerals are published, and it is relatively simple to compare the diffraction pattern of your unknown with the patterns of known minerals. There is a problem, however, with soils which are mixtures of clay minerals, soils which contain organics and other non-clay mineral constituents, and soils with mixed-layer minerals. Usually a detailed quantitative analysis is impossible —about all that one can tell is which minerals are present and roughly how much of each. ‘Another technique that is sometimes used to identify clay minerals is differential thermal analysis (DTA). A specimen of the unknown soil along with an inert control substance is continuously heated to several hundred degrees in an electric furnace, and certain changes in temperature occur because of the particular structure of the clay minerals. The changes occur at specific temperatures for specific minerals, and the record of these changes may be compared with those of known minerals. Electron microscopy, both transmission and scanning, can be used to identify clay minerals in a soil sample, but the process is not easy and/or quantitative. ‘A simple approach suggested by Prof. Casagrande is to use the Atterberg limits. It was mentioned (Sec. 2.8) that the activity has been related to specific active or inactive clays. Montmorillonites will be highly active since they are very small and have large plasticity indices. Use of Casagrande’s plasticity chart (Fig. 3.2) can also tell you just about as much, at least from an engineering point of view, as the more sophisticated diffraction and DTA analyses. The procedure is shown in Fig. 4.14. You simply locate your sample on the LL-PI chart, and compare its location44 Specitic Surtace 80 Plasticity index Chlorites Liquid timit Fig. 4.14 Location of common clay minerals on Casagrande’s plasticity chart (developed from Casagrande, 1948, and data in Mitchell, 1976) with those of known minerals. If your sample has Atterberg limits that plot high above the A-line near the U-line, then chances are that it contains a lot of active clay minerals such as montmorillonite. Even if the soil is classified as a CL, for example a sandy clay (CL), and still plots near the U-line, the clay portion of the soil is predominantly montmorillonite. The glacial lake clays from around the Great Lakes region in the United States and Canada are predominantly illitic and they plot right above the A-line. Scandinavian marine clays which are illitic also plot in this region. Kaolinites, which are relatively inactive minerals, plot right below the Aline. Even though they are technically clays, they behave like ML-MH materials, 4.4 SPECIFIC SURFACE Specific surface is the ratio of the surface area of a material to either its mass or volume. In terms of volume specific surface = surface area/volume (41)a Clay Minerals and Soll Structure The physical significance of specific surface can be demonstrated using a 1X 1x 1 em cube. 6(1 em?) 3 = 6/em = 0.6/mm Tem’ specific surface = If the cube is 1 mm on a side, the specific surface would be 6(1 mm?) Ima? If the cube is I um on a side, the specific surface would be 6(1 wm?) tam? This illustrates that large particles, whether cubes or soil particles, have smaller surface areas per unit of volume and thus smaller specific surfaces than small particles. To obtain the specific surface in terms of mass, you just divide the value in terms of volume by the mass density p,; units ‘would then be m?/g or m?/kg. Now, if sufficient water was present to just dampen the surface area of the cubes in the above example, it would take ten times as much water to wet the surface of all the grains when the cubes were 1 mm on a side than when the same volume occupied a single cube of 1 em’, Note also that if one were trying to remove water from the surface wet soil, there would be ten times as much water to remove from the smaller grains. Specific surface is inversely proportional to the grain size of a soil. We generally do not compute the specific surface for practical cases since the soil grains are too irregular in shape to do so. But it should be clear that a soil mass made up of many small particles will have on the average a larger specific surface than the same mass made up of large particles. From the concept of specific surface, we would expect larger mois- ture contents for fine-grained soils than for coarse-grained soils, all other things such as void ratio and soil structure being equal. You may recall from your materials courses that specific surface is a primary factor in concrete and asphalt mix design. In both cases it is necessary to provide sufficient cement paste or asphalt to coat the aggre- gate surfaces. =6/mm = 6/um = 6000/mm 4.5 INTERACTION BETWEEN WATER AND CLAY MINERALS ‘As mentioned previously, water usually doesn’t have much effect on the behavior of granular soils. For example, the shear strength of a sand is approximately the same whether it is dry or saturated. An important45 Interaction Between Water and Clay Minerals 1 exception to this fact is the case of water present in loose deposits of sand subjected to dynamic loadings such as earthquakes or blasts. ‘On the other hand, fine-grained soils, especially clay soils, are strongly influenced by the presence of water. The variation of water content gives rise to plasticity, and the Atterberg limits are an indication of this in- fluence. Grain size distribution only rarely is a governing factor in the behavior of fine-grained soils. Why is water important in fine-grained soils? Recall the discussion of specific surface, in the previous section, where the smalier the particle, the larger the specific surface. Clay minerals, being relatively small particles, have large specific surfaces, and everything else being equal, you might expect that they would have very active surfaces. ‘The relative sizes of four common clay minerals and their specific surfaces are shown in Fig. 4.15. Kaolinite, the largest clay mineral, has a Typical Typical Specific Thickness Diameter Surface Edge View {nen} (om) (em? kal Montmorillonite 3 100-1000 os ite 20 10.000 0.08 Chlorite 30 10.000 0.08 Kaolinite 50-2000 300-4000 0015 Fig, 4.15 Average values of relative sizes, thicknesses, and specitic surfaces of the common clay minerals (after Yong and Warkentin, 1975),na Clay Minerals and Soll Structure thickness or edge dimension of about 1 um, while montmorillonite, the smallest clay mineral, has a thickness of only a few nanometres. Since the crystals have roughly the same average “diameter,” at least within an order of magnitude, it is not surprising that the specific surfaces are so different. Of course, there are rather wide variations in the sizes of the crystals depending on weathering and other factors, but the values given are average values. Since surface activity is related to the particle size, you can see why montmorillonite, for example, is more “active” than kaolinite. Similarly, the surface activity of a sand or silt grain is practically zero. In Sec. 2.8, we defined the activity of a clay as PL 4 = Gay fraction (2-24) where the clay fraction is usually taken as the percentage of the sample less than 2 1m (Skempton, 1953). We mentioned that there was a pretty good correlation between activity and the type of clay mineral. This correlation is shown in Table 4-1. Now, it seems that clay particles are almost always liydrated in nature; that is, there are layers of water surrounding each crystal of clay. ‘This water is called adsorbed water. As discussed in the next section, the structure of clay soils and thus their engineering properties ultimately depend on the nature of this adsorbed water layer. How is water adsorbed on the surface of a clay particle? First, you may recall from chemistry or materials courses that water is a dipolar molecule (Fig. 4.16). Even though water is electrically neutral, it has two separate centers of charge, one positive and one negative, Thus the water molecule is electrostatically attracted to the surface of the clay crystal. Secondly, water is held to the clay crystal by hydrogen bonding (hydrogen ‘Na-montmorillonite Ca-montmorillonite Mite Kaolinite Hialloysite (dehydrated) Halloysite (hydrated) ‘Attapulgite ‘Allophane Mica (muscovite) Calcite Quarz ‘After Skempton (1953) and Mitchell (1976).4.5. Interaction Between Water and Clay Minerals 93 Fig. 4.16 Schematic diagram of a water molecule (atter Lambe, 4953) of the water is attracted to the oxygens or hydroxyls on the surface of the clay). The third factor is that the negatively charged clay surface also attracts cations present in the water. Since all cations are hydrated to some extent, depending on the ion, cations also contribute to the attraction of water to the clay surface. Of these three factors, hydrogen bonding is probably the most important factor. The attraction of water to the clay surface is very strong near the surface and diminishes with distance from that surface. It seems that the water molecules right at the surface are very tightly held and strongly oriented. Measurements show that some thermodynamic and electrical properties of the water next to the clay surface are different than that of “free water” (Mitchell, 1976). The source of the negative charge at the surface of the clay crystal results from both isomorphous substitution, mentioned earlier, and imper- fections in the crystal lattice, especially at the surface. “Broken” edges contribute greatly to unsatisfied valence charges at the edges of the crystal. Since the crystal wants to be electrically neutral, cations in the water may be strongly attracted to the clay, depending on the amount of negative charge present. Different clays have different charge deficiencies and thus have different tendencies to attract the exchangeable cations. They are called exchangeable since one cation can easily be exchanged with one of the same valence or by two of one-half the valence of the original cation As might be expected from their relative sizes and specific surfaces, montmorillonite has a much greater charge deficiency and thus a much greater attraction for exchangeable cations than kaolinite. Ilite and chlo- rite are intermediate in this respect. Calcium and magnesium are the predominant exchangeable cations in soils, whereas potassium and sodium are less common. Aluminium and hydrogen are common in acidic soils. The depositional environment as well as subsequent weathering and leaching will govern what ions are present in a particular soil deposit. As might be expected, marine clays are predominately sodium and magnesium since these are the most common cations in sea water. Cation exchange or replacement is further com- plicated by the presence of organic matter.Co Clay Minerale and Soll Structure ‘The ease of replacement or exchange of cations depends on several factors, primarily the valence of the cation. Higher valence cations easily replace cations of lower valence. For ions of the same valence, the size of the hydrated ion becomes important; the larger the ion, the greater the replacement power. A further complication is the fact that potassium, even though it is monovalent, fits into the hexagonal holes in the silica sheet. Thus it will be very strongly held on the clay surface, and it will have a greater replacement power than sodium, for example, which is also mono- valent. The cations can be listed in approximate order of their replacement ability. The specific order depends on the type of clay, which ion is being replaced, and the concentration of the various ions in the water. In order of increasing replacement power the ions are Lit ae aProblems 107 tests on samples of the sand at the same relative density as in the field. Sampling of loose grai.ulat materials, especially at depths greater than a few metres, is ve.y difficult. Since the materials are very sensitive to even the slightes? vibration, one is never sure the sample has the same density as the satural soil deposit. Therefore different kinds of penetrometers are used in engineering practice, and the penetration resistance values are roughly correlated with relative density. For deposits at shallow depths where direct access is possible, other techniques have been developed to measure the in-place density of compacted soils. These techniques are discussed in detail in Chapter 5. Finally, it should be noted in this discussion of the structure of granular soils that relative density alone is not sufficient to characterize their engineering properties. It is possible for two sands, for example, to have identical void ratios and relative densities but significantly different fabrics and thus significantly different engineering behaviors. Figure 4.26 is a two-dimensional example of such a fabric. Both “sands” are identical —they have the same grain size distributions and the same void ratios. But their fabrics are obviously very different. Stress history is another factor that must be considered when dealing with sands and gravels in engineer- ing practice. Deposits of granular materials which have been preloaded by nature or man will have very different stress-strain properties and therefore very different settlement responses (Lambrechts and Leonards, 1978). PROBLEMS: 4-1. Calculate the specific surface of a cube (a) 10 mm, (b) 1 mm, (©) | um, and (é) 1 nm on a side. Calculate the specific surface in terms of both areas and m?/kg. Assume for the latter case that p, 2.65 Mg/m’, 4-2. Calculate the specific surface of (a) tennis balls, (b) ping pong balls, (©) ball bearings | mm in diameter, and (d) fly ash with approximately spherical particles of 50 am in diameter. 4-3. The values of eng and ema, for a pure silica sand (p,= 2.65 Mg/m?) were found to be 0.46 and 0.66, respectively. (a) What is the corre- sponding range in dry density? (b) If the in situ void ratio is 0.63, what is the density index?108 46. 47. 48, Clay Minerals and Soll Structure Describe briefly the crystalline or atomic structure of the following ten minerals. Also list any important distinguishing characteristics. (a) Smectite (b) Brucite (©) Gibbsite (d) Attapulgite (©) Bentonite _(f) Allophane (g) Halloysite (h) Illite (i) Mica @) Chlorite . Describe the following types of bonding agents found with clay minerals. (a) Hydrogen bond (b) Covalent bond (c) van der Waals’ forces (d) James bond ‘The wet density of a sand in an embankment was found to be 1.9 Mg/m? and the field water content was 10%. In the laboratory, the density of the solids was found to be 2.66 Mg/m’, and the maximum and minimum void ratios were 0.62 and 0.44, respectively. Calculate the relative density of the sand in the field. ‘Which sheet, silica or alumina, would you wear to a toga party? Why? Given the particles in Fig. 4.26, is it realistic to show that all the particles are in contact with each other for this given plane? Any given plane? Why?five Compaction 5.1 INTRODUCTION In geotechnical engineering practice the soils at a given site are often less than ideal for the intended purpose. They may be weak, highly compressible, or have a higher permeability than desirable from an en- gineering or economic point of view. It would seem reasonable in such instances to simply relocate the structure or facility. However, considera- tions other than geotechnical often govern the location of a structure, and the engineer is forced to design for the site at hand. One possibility is to ‘adapt the foundation to the geotechnical conditions at the site. Another possibility is to try to stabilize or improve the engineering properties of the soils at the site. Depending on the circumstances, this second approach may be the most economical solution to the problem. Stabilization is usually mechanical or chemical, but even thermal and electrical stabiliza- tion have occasionally been used or considered. In this chapter we are primarily concerned with mechanical stabiliza- tion or densification, also called compaction. Chemical stabilization in- cludes the mixing or injecting of chemical substances into the soil. Port- land cement, lime, asphalt, calcium chloride, sodium chloride, and paper mill wastes are common chemical stabilization agents. Other methods for stabilizing unsuitable foundation soils include dewatering, which is the removal or reduction of unwanted excess ground water pressures, and preloading, in which the foundation soils are surcharged with a temporary overload so as to increase the strength and decrease anticipated settlement. The details of these and other, methods are de- scribed in textbooks on foundation and highway engineering. A good state 109110 ‘Compaction of the art discussion and reference to methods for improving the engineer- ing characteristics of soils is provided in the ASCE (1978) committee report “Soil Improvement—History, Capabilities, and Outlook.” ‘Compaction and stabilization are very important when soil is used as an engineering material; that is, the structure itself is made of soil. Earth dams and highway embankments are examples of earth siructures. If soils are dumped or otherwise placed at random in a fill, the result will be an embankment with low stability and high settlement. In fact, prior to the 1930's, highway and railroad fills were usually constructed by end-dumping soils from wagons or trucks. There was very little attempt to compact or densify the soils, and failures of even moderately high embankments were common. Of course earthworks such as levees are almost as old as man, but these structures, for example in ancient China or India, were con- structed by people carrying small baskets of earth and dumping them in the embankment. People walking over the dumped materials compacted and thus strengthened the soils. Even elephants have been used in some countries to compact soils, but research has shown that they are not very good at it (Meehan, 1967). The following symbols are introduced in this chapter. Symbol Dime Unit Defi L mm _ Depth of influence (Eq. 52) L/T? m/s? Acceleration of gravity, 9.80665m/s? = (% Relative compaction (Ea. 53) M 1 Mass of falling weight (Ea. $2) = (%) Optimum water content; sometimes called the ‘optimum moisture content (OMC) ‘M/L>— Mg/m? Maximum dry density ‘M/L?_Mg/m? Field dry density 5.2 COMPACTION Compaction is the densification of soils by the application of mecha- nical energy. It may also involve a modification of the water content as well as the gradation of the soil. Cohesionless soils are efficiently com- pacted by vibration. In the field, hand-operated vibrating plates and motorized vibratory rollers of various sizes are very efficient for compact- ing sand and gravel soils. Rubber-tired equipment can also be used efficiently to compact sands. Even large free-falling weights have been used to dynamically compact loose granular fills. Some of these techniques are discussed later in this chapter.5.3 Theory of Compaction a Fine-grained and cobssive soils may be compacted in the laboratory by falling weights and hammers, by special “kneading” compactors, and even statically in a common loading machine or press. In the field, common compaction equipment includes hand-operated tampers, shzepsfoot rollers, rubber-tired rollers, and other types of heavy compac- tion equipment (Sec. 5.5). Considerable compaction can also be obtained by proper routing of the hauling equipment over the embankment during construction. The objective of compaction is the improvement of the engineering properties of the soil mass. There are several advantages which occur through compaction Detrimental settlements can be reduced or prevented. Soil strength increases and slope stability can be improved. Bearing capacity of pavement subgrades can be improved. Undesirable volume changes, for example, caused by frost action, swelling, and shrinkage may be controlled. 5.3 THEORY OF COMPACTION The fundamentals of compaction of cohesive soils are relatively new. R. R. Proctor in the early 1930's was building dams for the old Bureau of Waterworks and Supply in Los Angeles, and he developed the principles of compaction in a series of articles in Engineering News-Record (Proctor, 1933). In his honor, the standard laboratory compaction test which he developed is commonly called the Proctor test. Proctor established that compaction is a function of four variables (1) dry density p,, (2) water content w, (3) compactive effort, and (4) soil type (gradation, presence of clay minerals, etc.). Compactive effort is a measure of the mechanical energy applied to a soil mass. In countries where British Engineering units are used, compactive effort is usually reported in ft-Ibf/ft?, whereas the SI units are J/m’ (J = joules). Since 1 J=1 Nem, and using the conversion factors in Appendix A, we can determine that | ft-Ibf/ft = 47.88 J/m’, In the field, compactive effort is the number of passes or “coverages” of the roller of a certain type and weight on a given volume of soil. In the laboratory, impact or dynamic, Kneading, and static compaction are usually employed. During impact compaction, which is the most common type, a hammer is dropped several times on a soil sample in a mold. The mass of the hammer, height of drop, number of drops, number of layers of soil, and the volume of the mold are specified. For example, in the standard Proctor test [also standard AASHTO2 ‘Compaction (1978), Designation T 99, and ASTM (1980), Designation D 698}, the mass of the hammer is 2.495 kg (5.5 Ib) and the height of fall is 304.88 mm (1 fu). The soil is placed in three layers in an approximately 1 litre (0.944 x 10-? m? or 1/30 ft?) mold, and each layer is tamped 25 times. Compactive effort can then be calculated as shown in Example 5.1. EXAMPLE 5.1 Given: Standard Proctor test hammer and mold. Required: Calculate the compactive effort in both SI and British Engineering units. Solution: a. SI units: compactive _ 2.495 ka(9.81 m/s?)(0.3048 m)(3 layers)(25 blows/layer) effort 0.944 x 10-3 m? 92.7 kJ /m If exact values of g and the volume are used, the standard Proctor compactive effort is 592.576 kJ /m?. b. British Engineering units: 5.5 TAC f1)(3)(25) ge fe-lbf 1 compactive effort = = 12,375 This calculation is, strictly speaking, incorrect since the 5.5 1b hammer is really a mass not a weight. However, the differences are negligible. For other types of compaction, the calculation of compactive effort is. not so simple. In kneading compaction, for example, the tamper kneads the soil by applying a given pressure for a fraction of a second. The5.3 Theory of Compaction 13 kneading action is supposed to simulate the compaction produced by a sheepsfoot roller and other types of field compaction equipment. In static compaction, the soil is simply pressed into a mold under a constant static stress in a laboratory testing machine. The process of compaction for cohesive soils can best be illustrated by considering the common laboratory compaction or Proctor test. Several samples of the same soil, but at different water contents, are compacted according to the standard Proctor compaction test specifications given previously. Typically, the total or wet density and the actual water con- tent of each compacted sample are measured. Then the dry density for each sample can be calculated from phase relationships we developed in Chapter 2. mM on 26) a | Pa ew (2-14) When the dry densities of each sample are determined and plotted versus the water contents for each sample, then a curve called a compaction curce for standard Proctor compaction is obtained (Fig. 5.1, curve A). 20 ———————— Deas of 60% "BOR 1005 Tor p,= 270 Main sturation Zero air 120 19 a wide" Lineof = z cotimus g : 5s 3 2 st = 2 (2) Modifies 0 § 8 Proctor = é g w ros (a) Standard Proctor 16 100 10 15 20 25 Water content w (%) Fig. 5.1 Standard and modified Proctor compaction curves for Crosby 8 til— : 4 ‘Compaction Each data point on the curve represents a single compaction test, and usually four or five individual compaction tests are required to completely determine the compaction curve. This curve is unique for a given soil type, method of compaction and (constant) compactive effort. The peak point of the curve is an important point. Corresponding to the maximum dry density é max i8 A Water content known as the optimum water content Wag (also known as the optimum moisture content, OMC). Note that the maximum dry density is only a maximum for a specific compactive \effort and method of compaction. This does not necessarily reflect the maximum dry density that can be obtained in the field. Typical values of maximum dry density are around 1.6 to 2.0 Mg/m?> (100 to 125 Ibf/ft?) with the maximum range from about 1.3 to 2.4 Mg/m? (80 to 150 Ibf/ft?). (Densities are also given in British Engineering unit because you are likely to encounter them in practice.) Typical optim. water contents are between 10% and 20%, with an outside maximum range of about 5% to 40%. Also shown on Fig. 5.1 are curves representing different degrees of saturation of the soil. From Eqs. 2-12 and 2-15, we can derive the equation for these theoretical curves. (S-1) ‘The exact position of the degree of saturation curves depends only on the value of the density of the soil solids p,. Note that at optimum water content for this particular soil, S is about 75%, Note too that the compac- tion curve, even at high water contents, never actually reaches the curve for “100% saturation” (traditionally called the zero air voids curve). And this is true even for higher compactive efforts, for example, curve B of Fig. 5,1. Curve B is the compaction curve obtained by the modified Proctor compaction test (modified AASHTO (1978), Designation T 180, and ASTM (1980), Designation D 1557]. This test utilizes a heavier hammer (4.536 kg, or 10 Ib), a greater height of fall (457 mm or 1.5 ft), and 5 layers tamped 25 times into a standard Proctor mold. You should verify that the compactive effort is 2693 kI/m? or 56,250 ft-Ibf/ft?. The modified test was developed during World War II by the U.S. Army Corps of Engineers to better represent the compaction required for airfields to support heavy aircraft. ‘The point is that increasing the compactive effort tends to increase the maximum dry density, as expected, but also decreases the optimum water content. A line drawn through the peak points of several compaction curves at different compactive efforts for the same soil will be almost parallel to a 100% S curve. It is called the line of optimums.5.3 Theory of Compaction 15 Typical compaction curves for different types of soils are illustrated in Fig. 5.2. Notice how sands that are well graded (SW soils, top curve) have a higher dr Jensity than more uniform soils (SP soils, bottom curve). For clay s°.., the maximum dry density tends to decrease as plasticity incre ses Why do we get compaction curves such as those shown in Fig. 5.1 and 5,2? Starting at low water contents, as the water content increases, the particles develop larger and larger water films around them, which tend to “lubricate” the particles and make them easier to be moved about and reoriented into a denser configuration. However we eventually reach a water content where the density does not increase any further. At this point, water starts to replace soil particles in the mold, and since p, 2 “| Unconsoidated- undrained tess Bg | Confining pressure = 10 kg/m? 10268 ae T ‘Strength (stress required to cause \ sarin vs. mater content ‘0 — =a ~e fo i _ Layers Tamps per Foot layer pressure Stress required to cause 5% strain (kg/em? 7 15276 psi 7 18 136 psi 7 248 65 psi 2 + ol 10 116 104 Ory density (tot? 8 100 \Ofmeetta ees 14 eer6 Molding 1820 water content (%) (o) 2 26 26 Fig. 5.7(0) Strength as a function of compactive effort and molding water content (atter Seed and Chan, 1958).“asset aque] yy, 2antswas 29 0) ade asoun opis Ki sayead8 gona op Aig 98H 9p 2M Javad Sq 30 ores ay nogy apis Kacy Ponrad Buyjons 108 9q ov ops 1 portonaid ‘Foyoas 9 soyfy yeqnaas0s apts Kigq fig eAamos ope hag 2291 gona api Kg Atpides aso sayeproruoo apis Kacy fw amssoud W8ry Um oprs Kap ‘o8ues amssosd ‘no| w ajqissesdaioo asou 2p%s 12m wonvomsd £q ‘10ur ganar poonpas Auiqvatuied opis Aig ajqeemisad atom apis aq, 280g 0} santsuas amionns opis Ki ameroad aod 2mo} 4 ‘stot seas sem 23000 s9qiqur sq :iuaEy9pasou api ict ‘mopar: i001 2p Ky 1 aunssasd s01eM 210g “2 penreig (4) epmimseyy “y wosedwe Auadorg suonoedwoo wnundo 40 1844 pue wnwndo J0 Aig uBemieg soMLedoid 05 JO UOSUEdWOD 1-§ BTEVL 123124 ‘Compaction due to swelling, especially dry of optimum. The strength curves for a silty clay compacted by kneading compaction for three different compactive efforts are shown in Fig, 5.7b. They show the stress required to cause 25% strain (upper) and 5% strain (middle) for the three compactive efforts. The strengths are about the same wet of optimum and increase significantly on the dry side of optimum. ‘Note too that at a given water content wet of optimum, the stress at 5% strain is actually less for the higher compaction energies. This fact is also shown in Fig. 5.8, where strength is measured by the CBR (California bearing ratio) test. In this test, the resistance to penetration of a 3 in? piston developed in a compacted specimen is compared to that developed by a standard sample of densely compacted crushed rock. The CBR is a common pavement design test. In Fig. 5.8 a greater compactive effort produces a greater CBR dry of optimum, as you would expect. But notice how the CBR is actually less wet of optimum for the higher compaction ‘energies. This fact is important in the proper design and management of a ‘compacted earth fill; we shall discuss its implication later in this chapter. Table 5-1 from Lambe (1958b) is a summary of the effects of wet versus dry of optimum compaction on several engineering properties. 5.5 FIELD COMPACTION EQUIPMENT AND PROCEDURES Soil to be used in a compacted fill is excavated from a borrow area. Power shovels, draglines and self-propelled scrapers ot “pans” are used to excavate the borrow material. A self-loading scraper is shown in Fig. 5.9a and an elevating scraper in Fig. 5.9b. Sometimes “dozers” are necessary to help load the scraper. Scrapers may cut through layers of different materi- als, allowing several grain sizes to be mixed, for example, The power shovel mixes the soil by digging along a vertical surface, whereas the scraper mixes the soil by cutting across a sloping surface where different layers may be exposed. ‘The borrow area may be on site or several kilometres away. Scrapers, off the road vehicles, are often used to transport and spread the soil in lifts on the fill area. Trucks may be used as well, on or off the highway, and they may end dump, side dump, or bottom dump the fill material (Fig. 5.10a). For economic reasons, the hauling contractor usually tries to spread the fill material when dumping in order to reduce spreading time. However, unless the borrow materials are already within the desired water content range, the soil may need to be wetted, dried, or otherwise reworked. Where possible, the contractor directs his earth-moving equipment over previously100 120 115 100 90 10 is 20 25 Water content (6) Legend P05 blows per layer 026 blows per layer 22 12 blows per layer a——o 6 blows per layer Note: 10 1b hammer, 18" drop (modified Proctor) Fig. 5.8 Strength as measured by the CBR and the dry density versus ‘water content for laboratory impact compaction (after Turnbull and Fos- ‘er, 1956), 125Fig. 5.9 Two types of scrapers: (a) Conventional or solt-loading scraper. ‘Sometimes a “dozer” or two helps push the “pan” to load up. Here the two scrapers are assisting each other load up in push-pull. (b) Elevating scraper where the elevating machine loads by itself and eliminates the need for a pusher. (Photographs courtesy of Caterpillar Tractor Co.)cy Fig. $.10 Examples of equipment used for hauling and spreading tit ‘materials: (a) fill material being hauled by end dump truck; (0) motor grader spreading and preparing fill subgrade (photographs courtesy of Caterpillar Tractor Co.) uncompacted soil, thereby reducing the amount of compactive effort required later. Once borrow material has been transported to the fill area, bull- dozers,” front loaders, and motor graders. called blades (Fig. 5.10b), spread the material to the desired layer or lift thickness. Lift thickness may range from 150 to 500 mm (6 to 18 in.) or so, depending on the size and type of compaction equipment and on the maximum grain size of the fill. The kind of compacting equipment or rollers used on a job will depend on the type of soil to be compacted. Equipment is available to apply pressure, impact, vibration, and kneading. Figure 5.11 shows two types of rollers. ‘A smooth wheel, or drum, roller (Fig. 5.11a) supplies 100% coverage under the wheel, with ground contact pressures up to 380 kPa (55 psi) and may be used on all soil types except rocky soils. The most common use for “Genus bovinas masculinus sonambulorum, az© Fig. 5.11 Types of rollers: (a) smooth-whee! roller: (b) rubber-tired roller (eelt-propelied) (photographs courtesy of Hyster Company, Construction Equipment Division). 12855 Field Compaction Equipment and Procedures, 120 large smooth wheel rollers is for proofrolling subgrades and compacting asphalt pavements. The pneumatic, or rubber-tired, roller (Fig. 5.11b) has about 80% coverage (80% of the total area is covered by tires) and tire pressures may be up to about 700 kPa (100 psi). A heavily loaded wagon with several rows of four to six closely spaced tires is self-propelled or towed over the soil to be compacted. Like the smooth wheel roller, the rubber-tired roller may be used for both granular and cohesive highway fills, as well as for earth dam construction. Probably the first roller developed and perhaps the most common type of compactor used today is the sheepsfoot roller. This roller, as its name implies, has many round or rectangular shaped protrusions or “feet” (Fig. 5.12a) attached to a steel drum. The area of these protrusions ranges from 30 to 80 cm? (5 to 12 in.*), Because of the 8% to 12% coverage, very high contact pressures are possible, ranging from 1400 to 7000 kPa (200 to 1000 psi) depending on the drum size and whether the drum is filled with water. The drums come in several diameters. Surprisingly enough, a “4 by 4” (which means 4 ft long and 4 ft in diameter) roller provides a higher strength compacted fill in clay soils than a heavier, higher pressure “5 by 5” roller because there is less kneading or shearing action with the “4 by 4” than the “S by 5” roller, which produces a different soil structure (see Fig. 5.8). Sheepsfoot rollers are usually towed in tandem by crawler tractors or are self-propelled, as shown in Fig. 5.12b. The sheepsfoot roller starts compacting the soil below the bottom of the foot (projecting about 150 to 250 mm from the drum) and works its way up the lift as the number of passes increases. Eventually the roller “walks out” of the fill as the upper part of the lift is compacted. The sheepsfoot roller is best suited for cohesive soils. Other rollers with protrusions have also been developed to obtain high contact pressures for better crushing, kneading, and compacting of a rather wide variety of soils. These rollers can either be towed or self- propelled. Tamping foot rollers (Fig. 5.13) have approximately 40% cover- age and generate high contact pressures from 1400 to 8400 kPa (200 to 1200 psi), depending on the size of the roller and whether the drum is filled for added weight. The special hinged feet (Fig. 5.13a) of the tamping foot roller apply a kneading action to the soil. These rollers compact similarly to the sheepsfoot in that the roller eventually “walks out” of a well- compacted lift. Tamping foot rollers are best for compacting fine-grained soils. Still another kind of roller is the mesh, or grid pattern, roller with about 50% coverage and pressures from 1400 to 6200 kPa (200 to 900 psi) (Fig. 5.14). The mesh roller is ideally suited for compacting rocky soils, gravels, and sands. With high towing speed, the material is vibrated, crushed, and impacted.Bureau of Reclamation)" "8 Bureau of Reclamation 2S ‘and Corps of Engineers coro ot Engineers? TL | A. Bureau of, —— 92" Reclamation ~ Corps of Upper portion Engineers of shank. Lower portion of shank @ Oy Fig. 5.12 Shoepstoct rollers: (a) detail of a rectangular sheepstoot (draw: ing provided by Hyster Company, Construction Equipment Division); (©) sell-propelied sheepstoot roller in foreground (“pan in background) (photograph courtesy of Caterpillar Tractor Co.)Hardness Re 55 (25.5 mm) @ Fg. 5.19 Tamping tot rollr(e deta of a tamping foot () set propaliod tamping foot compactor. Note how the blade is used to spread the material before compaction by the rollers (drawing and photograph courtesy of Caterpillar Tractor Co.).132 ‘Compaction Fig. 5.14 Grid roller (photograph courtesy of Hyster Company, Construc- tion Equipment Division). Several compaction equipment manufacturers have attached vertical vibrators to the smooth wheel and tamping foot rollers so as to more efficiently densify granular soils. Figure 5.15 shows a vibrating drum on a smooth wheel roller compacting a gravelly material. Also available are vibrating plates and rammers that range in size from 230 to 122 mm® (9 to 45 in.2) and weigh from 50 to 3000 kg (100 to 6000 Ib). Compaction depth for even the larger plates is less than 1 metre. These devices are used in Fig. 5.15 Vibrating drum on smooth-whee! roller (photograph courtesy of the Hyster Company, Construction Equipment Division).5.5 Fleld Compaction Equipment and Procedures 133 areas where the larger rollers cannot operate, Broms and Forssblad (1969) have listed the different types of vibratory soil compactors (see Table 5-2) and also indicate the common frequency of operation. In Table 5-3, they illustrate the practical application of these machines. Probably the best explanation of why roller vibration causes densifi- cation of granular soils is that particle rearrangement occurs due to cyclic deformation of the soil produced by the oscillations of the roller. In addition, vibratory compaction can work in materials with some cohesion (Selig and Yoo, 1977). When oscillation is added to a static component, compaction is significantly increased, as shown in Fig. 5.16. For soils compacted on the dry side of optimum, adding the dynamic component results in increased density ‘There are many variables which control the vibratory compaction or densification of soils. Some are compactor dependent and some depend on the soil being compacted. The list of variables would include: Characteristics of the compactor: Mass, size Operating frequency and frequency range Characteristics of the soil: Initial density Grain size and shape Water content Construction procedures: ‘Number of passes of the roller Lift thickness Frequency of operation of vibrator Towing speed ‘The compactor characteristics influence the stress level and depth of influence of the dynamic force, and the initial density strongly influences the final density. For example, the upper 30 cm of medium dense sand may never become compacted higher than the initial density, whereas dense sands will be vibrated loose in the top 30 cm. Once the compactor is chosen, the actual construction procedures essentially govern the results. The influence of operating frequency for various soil types is shown in Fig. 5.17. Note how a peak in the density-frequency curve develops for most of the soils, even clays. The frequency at which a maximum density is achieved is called the optimum frequency. It is a function of the compactor- soil system, and it changes as the density increases during the process of compaction. Clearly, it is desirable for a compactor to have the capability to vary its operating frequency and have the range required to obtain3ye05 powEN WO P2sA), “(6961 palassiod pur stoug Ve THOs-02 sU0nSI-S1 ‘warp sore 7H 09-0 sam soqqnu ‘papadoud-§ 7H 08-06 ‘dév-wwopar ‘payadoid-y§ (strap of 30 200) 7H 08-0» popint-puvy paqpdosd es s4aqjou Bunsosgi, aH si-o1 — suoroz dn pevunow- 20055) (a1, 009-008) ‘19 “s201983 7H OL-O€ 84 00-007, wo payunour ‘adky-apdnin yy Ca st-6) (@1.0009-001) THOEMogY wD sE-E —__poromows-seesy| zHOg-zt BL OOOE-OF _papINa-pRY “Ponedosd-js ib wonofouta oroodon sod BunDugt (ar 9-2) PanMow-20)981 (@100¢~001) ‘aH 0ee-001 MHSI-F so parendo AtenTEHY | 7H OI IRoaY #4 951-05 Poapini-pasH Hvoo1g7 38.000 (soures saduoy Bu01q14 fowonbaig seem eumeyyjoady | Kovenbosa Sse SaRNDEN 0 20K HOTegIA PET ‘svoresqia 298g {s#0}9edwOD Hos ArovesaIA Jo SodKL WEBI ZS STOVETABLE §-3 Applications af Vibratory Soil Compaction* “Type of Machin ‘Applications Vibrating tampers (cammers): Street repair. Fills behind bridge abutments, retaining, and basement walls, ete. Trench fill. Vibrating plate compactors: Self-propelled, hand-guided Base and subbase compaction for streets, sidewalks, etc. ‘Street repair. Fill behind bridge abutments, retaining and basement walls, etc. Fills below floors. Trench fills. Multiple-type Base and subbase compaction for highways, Vibrating rollers Self-propelied, hand-guided Base, subbase, and asphalt compaction for streets, side- walks, parking areas, garage driveways, etc. Fills behind bridge abutments and retaining walls. Fills below floors. Trench fills. Self-propelied, tandem type Base, subbase, and asphalt compaction for highways, streets, sidewalks, parking areas, garage driveways, etc. Fills below floors. Self-propelled, rubber tires Base, subbase, and embankment compaction for highways, streets, parking areas, airfields, etc. Rock-fill dams. Fills (sol of rock) used as foundations for residential and industrial buildings. ‘Tractordrawn Base, subbase, and embankment compaction on highways, streets, parking areas, airfields, etc. Earth- and rock- fill dams. Fills (soil or rock) used as foundations for residential and industrial buildings. Deep compaction of natural deposits of sand. “After Broms and Forssblad (1968). 128-4 > 1 2.00 Increased | generated 2010 air voids es ] 120} force 192 g 3 = 2 ein shou 184 2 vibration 3 8 3 E10 16 § Static component 105 168 10 tae Water content (%) 3 Fig. 5.16 Compaction results on 30 cm (12 in) layers of silty sand, with land without vibration, using a 7700 kg (17,000 Ib) towed vibratory roller (atter Parsons, ot al., 1962, as cited by Selig and Yoo, 1977), 135136 ‘Compaction Symbol Roller wt Symbo!_ Roller wt f—27,5001b we 8.60015 o—o 7.500 o> 5,800 0 17,000 x——x 2600 150 | Gravel-sand clay 1 Well raded Cohesive sandy grave! 140 — 422 oa > Slag o we | — 130+ — ae -s Giver sandetay 42.08 q 1 5 £ Gravely sone | = 2 | | | = z sandy clay — 4192 = é 1 3 ° 4 = ite © 100 > 41.60 Unitormsand F \ | Heavy clay \ Heavy clay i ol pd iaa 1000 2000 3000 Vibration frequency (com) Fig. 5.17 Variation with frequency of compaction by smooth-drum vibra tory rollers (alter several sources as cited by Selig and Yoo, 1977), maximum density. However, the peaks are gentle and, percentagewise, a wide frequency range is not all that important. The influence of the number of passes of a roller and the towing speed are shown in Fig. 5.18 for a 7700 kg roller compacting a “heavy” (high LL) clay and a well-graded sand, Notice how the density increases as the number of passes or coverages increases, up to a point. Not so obvious is that, for a given number of passes, a higher density is obtained if the vibrator is towed more slowly! The effect of lift thickness may be illustrated by the work of D'Appolonia, et al. (1969), shown in Fig, 5.19. Here a 5670 kg roller operating at a frequency of 27.5 Hz is used to compact a 240 cm thick layer of northern Indiana dune sand. The initial relati5.5 Fleld Compaction Equipment and Procedures 137 0 176 isa 100 1.52 90 (a) Heavy clay uso e {rotstre = 21%) a = = B80 h2e 2 z z i l200 § 130 a g dom : Ji.80 120 1.20 ry (0) Well raded san (moisture = 75%) 1.70 100 1.60 o «6 6 20 Roller coverages Fig. 6.18 Etfect of roller travel speed on amount of compaction with 7700 kg (17,000 Ib) towed vibratory roller (after Parsons, et al. 1962, as cited by Selig and Yoo, 1977). 50% to 60%. Field density tests were made in test pits before and after compaction. Note how the density varies with depth. In the upper 15 cm (6 in.), the soil is vibrated loose, whereas the soil reaches its maximum density for a given number of coverages at about 45 cm; thereafter the increase in density tapers off. When compacting past five or so coverages, there is not ‘a great increase in density. EXAMPLE 5.2 Given: It is decided, based on economics, that five coverages of a certain roller and operating frequency shall be used.138 ‘Compaction Required: What would be the maximum lift thickness to obtain a minimum relative density? Use the data shown in Fig. 5.19. Solution: Trace the relative density versus depth curve for five passes on onionskin paper. Superimpose that drawing over the original one, and slide it up and down until the desired relative density is obtained (shown in Fig. Ex. 5.2). ‘About 45 cm (18 in.) is indicated as the maximum thickness. Actually, however, the lift could be thicker as compaction of the top layer densifies the lower layer the second time around. Dv density bt) Dry densy 41°) ee eee st 1080 ° eee eee o ao ' | ' > ; - 23) | 3} % a, | at , Zak 2s von {0 Bole oom 5} - A. — ste : 7 7 i oA ' 4 «6% 8 80 10 oo % 7 #9 10 reiave ersy Flav deity Dry censty ao) ory desi th) 1 15 m0 858 tts : : gs: | te) 15 rote oases ; ak 7 1 (ah 8 oer pas j ' ie Fig. 5.19 Density-depth relationship for a 5670 kg roller operating at 27.5 Hz for a 240 cm lift height (after D'Appolonia, et al., 1969)5.5 Fleld Compaction Equipment and Procedures, 130 Relative density (X) Relative density () 50 60 70 8 9% (50 60 70 80 © ia 9 —— 1 =? B3 g a x g relative \ density = 75% 4 Density-depth 5 rolationship for tare it height using 5 roller passes Fig. Ex. 5.2 Approximate method for determining litt height required to achieve a minimum compacted relative density of 75% with five roller asses, using data for a large lift height (after D'Appolonia, et al, 1968). Figure 5.20 summarizes the applicability of various types of compac- tion equipment as a function of soil type, expressed as a percentage of sand to clay. These “zones” are not absolute, and it is possible for a given piece of equipment to compact satisfactorily outside the given zone. When structures are to be founded on relatively deep deposits of loose granular materials, densification by even heavy surface vibratory rollers is usually insufficient, and other techniques must be employed. Excavation and replacement of the soil in compacted layers may be eco- nomical under certain conditions. Blasting has also been used at some sites (Mitchell, 1970). Vibro-flotation (Mitchell, 1970) is often used to increase the density of building foundations on loose sand. Another technique which is gaining in popularity is dynamic compaction. Basically, the method consists of repeatedly dropping a very heavy weight (10 to 40 tons mass) some height (10 to 40 m) over the site. The impact produces shock waves that cause densification of unsaturated granular soils. In saturated granu- lar soils, the shock waves can produce partial liquefaction of the sand, a condition similar to quicksand (discussed in Chapter 7), followed by consolidation (discussed in Chapter 8) and rapid densification. The varia bles include energy (drop height and weight of pounder), the number of drops at a single point (3 to 10), and the pattern of the drops at the surface (5 to 15 m center-to-center). Figure 5.21 shows a pounder just impacting the surface of a loose sand layer. Eventually this site will look like a set of140 Compaction cOMPACTIVE COMPACTOR ZONES OF APPLICATION EFFORT, 100% 100% Clay Sand Rock sl Static wt, kneading Static wt, kneading Static wt, vibration Towed tamping foot | Static wt, kneading Fig. 520 Applicability of various types of compaction equipment for a given soil type (modified atter Caterpillar Tractor Co., 1977) organized moon craters. The craters can be filled with sand and addition- ally tamped or the area between them smoothed out by the pounder itself. Dynamic compaction was apparently first used in Germany in the mid-1930's during construction of the Autobahns (Loos, 1936). It has also been used in the USSR to compact loessial soils up to 5 m deep (Abelev, 1957). The technique was further refined and promoted in France and elsewhere by Louis Ménard (Ménard and Broise, 1975), who pioneered in the development of very heavy pounders (up to 200 metric tons mass) and massive cranes and tripods for lifting them to drop heights up to 40 m. Improvement is claimed to depths down to 40 m. In the United States, dynamic compaction has been used on a more modest scale by contractors using ordinary equipment (Leonards, Cutter, and Holtz, 1980; Lukas, 1980). The depth of influence D, in metres, of the soil undergoing compac- tion is conservatively given by Leonards, et al. (1980) as D=i(wn)'? (5-2) where W =mass of falling weight in metric tons, and ‘h =drop height in metres. The heavier the weight and/or the higher the drop height, the greater the depth of compaction. Leonards, et al. (1980) also found that the amount of improvement due to compaction in the zone of maximum improvement correlates best with the product of the energy per drop times the total energy applied per unit of surface area.5.6 Fleld Compaction Control and Specifications: 1“ Fig. 5.21 Dynamic compaction at a site in Bangladesh, The 100 ton crane is dropping a 16 metric ton weight 30 m (courtesy of S. Varaksin, Techniques Louis Ménard, Longiumeau, France). 5.6 FIELD COMPACTION CONTROL AND SPECIFICATIONS Since the objective of compaction is to stabilize soils and improve their engineering behavior, it is important to keep in mind the desired engineering properties of the fill, not just its dry density and water content. This point is often lost in earthwork construction control. Major emphasis1a Compaction is usually placed on achieving the specified dry density, and little consider- ation is given to the engineering properties desired of the compacted fill. Dry density and water content correlate well with the engineering proper- ties (Sec. 5.4), and thus they are convenient construction control param ters The usual design-construct procedure is as follows. Laboratory tests are conducted on samples of the proposed borrow materials to define the properties required for design. After the earth structure is designed, the compaction specifications are written. Field compaction control tests are specified, and the results of these become the standard for controlling the project. Construction control inspectors then conduct these tests to see that the specifications are met by the contractor. ‘There are basically two categories of earthwork specifications: (1) end-product specifications and (2) method specifications. With the first type, a certain relative compaction, or percent compaction, is specified. Relative compaction is defined as the ratio of the field dry density py nag to the laboratory maximum dry density pym,x, according to some specified stan- dard test, for example, the standard Proctor or the modified Proctor test; or relative compaction (R.C.) = 2%! x 100(%) (5-3) You should note the difference between relative compaction and relative density D, or density index Ip, defined in Chapter 4. Relative density, of course, applies only to granular soils. If some fines are present, it is difficult to decide which type of test is applicable as a standard test. ASTM (1980), Designation D 2049, suggests that the relative density is applicable if the soil contains less than 12% fines (passing the No. 200 sieve); otherwise the compaction test should be used. A relationship between relative density and relative compaction is shown in Fig. 5.22. A statistical study of published data on 47 different granular soils indicated Pg" 0 Dry density Pa min Pa Pamex ere ‘oid ratio e © enn Density index Ip or relative density D, (%) © " te Relative compaction R.C. (%) __F-C- * 80 190 ee ey Fig. 5.22 Relative densityiand relative compaction concepts (after Lee and Singh, 1971)5.8 Field Compaction Control and Specifications “3 that the relative compaction corresponding to zero relative density is about 80%. With end-product specifications, which are used for most highways and building foundations, as long as the contractor is able to obtain the specified relative compaction, how he obtains it doesn’t matter, nor does the equipment he uses. The economics of the project supposedly ensure that the contractor will utilize the most efficient compaction procedures. The most economical compaction conditions are illustrated in Fig. 5.23, showing three hypothetical field compaction curves of the same soil but at different compactive efforts. Assume that curve I represents a compactive effort that can easily be obtained by existing compaction equipment. Then Line ot 100% saturation ‘optimus. _ Dry density > pbb Water content (%) Fig. 5.23 Dry density versus water content, illustrating the most etlicient conditions for field compaction (after Seed, 1964),