A NUMERICAL AND EXPERIMENTAL STUDY OF SAMPLING DISTURBANCE BY ABU SIDDIQUE B.Sc., M.Sc. A Thesis submitted to the University of Surrey for the degree of Doctor of Philosophy in the Department of Civil Engineering MAY. 1990 SUMMARY This thesis contains numerical and experimental investigations of sampling disturbance. An approximate numerical technique has been developed to predict the strain paths of soil elements due to undrained penetration of a sampler, in order to evaluate the effects of cutting shoe designs on soil disturbance. Investigations of the cutting shoe designs of the NOI 54 mm dia. and SOl 50 mm dia. piston samplers, and two typical UlDO open-drive samplers indicate that soil disturbance depends not only on the thickness of the sampler but also to a great extent on the precise geometry of the cutting shoe. A parametric study of the effects of various design parameters for samplers shows that (a) an increase in area ratio and outside edge taper angle increased soil disturbance and (b) increasing the inside cutting edge taper angle and a decreasing the inside clearance ratio reduced soil disturbance. Results from the investigation of BIt ratio of the flat-ended samplers indicate that soil disturbance depends on thezyxwvutsrqponmlkjihgfedcbaZYXWV samplers and that the characteristics of the strain paths of these samplers are markedly different from those of other samplers (e.g., NOI, SOl and UlOO) with identical BIt ratio and thickness. A sample of soft London Clay (LLzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIH = 69, PI = 45) was prepared in the laboratory by Ko-consolidation from a slurry to a vertical effective stress of 100 kPa. The experimental programme consisted of carrying out (a) incremental loading oedometer tests and (b) stress and strain path tests. Stress and strain path tests, investigating the effects of tube penetration disturbances on undrained stress-strain characteristics of the clay, were carried out using a computer controlled stress path system, incorporating devices for monitoring local deformations and pore pressures. The most pronounced effects due to simulated tube penetration disturbances in unaged London Clay were found to be significant reductions in mean effective stress, initial stiffness parameters and pore pressure changes. ACKNOWLEDGEMENTS I am extremely grateful to my supervisor Dr. C.R.I. Clayton for his encouragement and moral support throughout the course of this research. I would like to express my profound gratitude and indebtedness to Dr. C.R.!. Clayton for his close supervision, suggestions, guidance, discussions and for reviewing the text. I sincerely thank Professor N.E. Simons, Head of the Department of Civil Engineering for allowing me to undertake this research. Sincere thanks are extended to Mr. M.J. Gunn for his help and suggestions in connection with the numerical analyses, using the LUSAS finite element package. The author is indebted to the members of staff in the Geotechnical Engineering Section in the Civil Engineering Department, namely, Mr. M.A. Huxley, Mr. M.C. Matthews, Mr. R.1. Woods, and to his research colleagues, S.A. Khatrush, J. Sadrekarimi, M. Vaziri, N. Saffari-Shocshtari, A. Bica, S. Instone, R. Hopper, R.P. Hillier, A. Ponnampalam, C.S. Russell, Mrs. M. Wicks, P.D. Williams and A.M. Thome. Special thanks are also due to Mr. P.F. Cheesman and Mr. C.G. Sivewright of the Soil Mechanics Laboratory for their assistance in various stages of laboratory investigations. The author also wishes to thank the members of staff within the Civil Engineering Department workshop and in particular Mr. K. LeHuep for his fabrication of the Hall effect axial strain devices and to Mr. D. Cleaver and Mr. B. Inch for their help in fabricating other components. Special thanks to Mr. M. Jackman of the Heavy Structures I. Rankin of the Laboratory for taking the photographs presented in the thesis and to Mr.zyxwvutsrqponmlkjih Structural Polymers Laboratory for his help in preparing the rubber grommet used for the installation of miniature pore pressure transducer. Many thanks to Mrs. J. Finch for typing part of the manuscript. The author also gratefully acknowledges the financial support of the Commonwealth Scholarship Commission of the United Kingdom. Finally, I would like to thank my wife, Shaheen, for her tireless patience, encouragement, support and understanding during the period of research. To My Parents and Wife CONTENTS SUMMARY ACKNOWLEDGEMENTS CHAPTER 1 1 INTRODUCTION 1.1 General 1 1.2 Objectives of the Present Work 4 1.3 The Research Scheme 5 1.4 Thesis LayoutzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA 5 CHAPTER 2 SAMPLING DISTURBANCE 7 2.1 Introduction 7 2.2 Causes of Sampling Disturbance 7 2.3 Investigations 8 of Sampling Disturbance 2.3.1 Disturbance During Drilling 8 2.3.2 Disturbance During Sampling 9 2.3.3 Disturbance Caused by Transportation 2.3.4 Disturbance Caused by Sample Preparation and Storage 12 15 2.4 Assessment of Sampling Disturbance 17 2.5 Sampler Design and its Effect on Sample Disturbance 21 2.5.1 Effect of Area Ratio and Cutting Edge Taper Angle 21zyxwvutsrqponml 2.5.2 Effect of Inside and Outside Clearances 22 2.6 2.7 Effect of Sampler Dimensions, Sampler Types and Sampling Methods on Sample disturbance 24 Sampling Effects 29 2.7.1 "Perfect" Sampling 2.7.2 . Block Sampling 2.8 30 34 2.7.3 Tube Sampling 34 2.7.4 Ideal Sampling 37 2.7.5 Methods of Correcting Sampling Disturbance Effects 39 Strain Path Method for Predicting Soil Disturbance 43 2.8.1 Comparison of the Strain and Stress Path Methods 43 2.8.2 Applications of Strain Path Method 44 TableszyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA 47 Figures 50 CHAPTER 3 FINITE ELEMENT ANALYSIS 67 67 68 3.1 The Main Objectives 3.2 Development of an Analytical Technique for Strain Path Computation 3.2.1 Introduction 68 3.2.2 Problem Description 69 3.2.3 Theory for Analysis 69 3.2.4 Finite Element Model Computation of Strains and Deformations 73 73 74 Errors in the Analyses 78 Minimisation of Errors 79 3.2.5 3.2.6 3.2.7 3.2.8 Material Properties and Boundary Conditions 3.2.9 Concluding Remarks 3.3 Details of Computational Programme 3.4 Analysis of NOI, sal and Ul00 Samplers 3.4.1 Cutting Shoe Design of the Samplers 3.4.2 Finite Element Models and Boundary Conditions 3.4.3 Strain Paths of Soil Elements 3.5 Parametric Study of Cutting Shoe Designs 3.5.1 Analyses With Different Area Ratios 81 82 83 83 84 85 85 3.5.1.1 Dimensions and Characteristics of the Samplers 3.5.1.2 Finite Element Model and Boundary Conditions 86 86 86 3.5.1.3 Strain Paths of Soil Elements 87 3.5.2 Analyses With Different Inside Clearance Ratios 87 3.5.2.1 Dimensions and Cutting Shoe Designs 87 3.5.2.2 Finite Element Model and Boundary Conditions 88 3.5.2.3 Strain Paths of Soil Elements 88 3.5.3 Analyses With Different Inside and Outside Cutting Edge Taper Angles 89 3.5.3.1 Dimensions and Characteristics of the Samplers 89 3.5.3.2 Finite Element Model and Boundary Conditions 89 3.5.3.3 Strain Paths of Soil Elements 90 3.6 Study of Flat-Ended Samplers 91 Tables 92 H~s W CHAPTER 4 LABORATORY INVESTIGATIONS, EQUIPMENTS AND INSTRUMENTATION 138 4.1 The Main Objectives 138 4.2 Preparation of Reconstituted soil 138 4.2.1 Introduction 138 4.2.2 Soil Used 139 4.2.3 Preparation of Slurry 140 4.2.4 Consolidation of Slurry 141 4.3 Sampling of Clay 143 4.4 Automated Stress/Strain Path Test Equipment 143 4.4.1 Introduction 143 4.4.2 Basic Features of the Stress/Strain Path Test Equipment 145 4.4.3 The Measuring Devices 147 4.5 Local Deformation Measurement 149 4.5.1 Introduction 149 4.5.2 Development of a Local Axial Strain Measuring Device 150 4.5.2.1 StageszyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA in the Development 151 4.5.2.2 Gauge Characteristics and Calibrations 4.5.3 The Local Radial Strain Measuring Caliper 154 155 4.6 Volume Change Measurement 156 4.7 Measurement of Porewater Pressure 157 4.7.1 Introduction 157 4.7.2 The Miniature Pore Pressure Transducer 158 4.7.3 Validity of Measured Pore Pressure at Mid-Height 159 4.8 Software for Stress and Strain Path Testing 160 4.8.1 Introduction 160 4.8.2 Stress and Strain Path Control Programs 161 4.8.2.1 Stress Path Control Sub-Program 161 4.8.2.2 Strain Path Control Sub-Program 164 4.9 Oedometer Tests 165 Table 166 Figures 167 CHAPTER 5 STRESS AND STRAIN PATH TESTS 191 5.1 Details of Testing Programme 191 5.2 193 Theoretical Investigations of the Test Rates for Ko-Consolidation 5.3 Test Procedure 5.3.1 Preparation and Set-up of Specimen 5.3.2 5.3.3 194 194 Installation Procedure for the Miniature Pore Pressure Transducer 195 Mounting Procedure of Local Axial Strain Devices and Caliper 196 5.3.4 Test Set-up and Execution 197 5.4 Processing and plotting of Test Data 198 Table 200 R~reszyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA WI CHAPTER 6 RESULTS AND DISCUSSIONS 208 6.1 Introduction 208 6.2 Predicted Strain Paths Due to Undrained Penetration of Samplers 208 6.3 Comparison of NOI, SOl and UIOO Samplers 211 6.4 Comparison of Cutting Shoe Designs from the Parametric Study 212 6.4.1 Samplers With Different Area Ratios 213 6.4.2 Samplers With Different Inside Clearance Ratios 214 6.4.3 Samplers With Different Inside Cutting Edge Taper Angles 215 6.4.4 Samplers With Different Outside Cutting Edge Taper Angles 217 6.4.5 Summary 218 BIt Ratios 218 6.5 Comparison of Flat-Ended Samplers and Samplers of IdenticalzyxwvutsrqponmlkjihgfedcbaZYXWV 6.6 One-Dimensional Consolidation and Permeability Properties of Normally Consolidated London Clay 6.7 Stress and Strain Path Test Results 220 222 6.7.1 Ko-Consolidation 222 6.7.2 Investigations of Various Approaches to Correct Test Results 223 6.7.3 Observed Behaviour in Compression and Extension for "Undisturbed" specimens 227 6.7.3.1 Stress Paths 227 6.7.3.2 Stress-Strain Behaviour 228 6.7.3.3 Pore Pressure Response During Shearing 230 6.7.4 Strain Path Tests Modelling Tube Penetration Disturbances 232 6.7.4.1 Stress Paths 232 6.7.4.2 Stress-Strain Behaviour During the Application of Strain Paths 235 6.7.4.3 Stress-Strain and Pore Pressure Characteristics After Tube Penetration Disturbances 237 6.8 Concluding Remarks 241 Tables 244 Figures 252 CHAPTER 7 CONCLUSIONS AND RECOMMENDATIONS 306 7.1 Conclusions 306 7.2 Recommendations for Further Study 314 REFERENCES 316 APPENDIX • A Listing of Computer Programs for Strain Path Computation 333 APPENDIX - B Basic Algorithm for Scanning Signals from Various Measuring Devices 348 APPENDIX. 350 C Summary of Stress and Strain Path Test Results CHAPTER 1 INTRODUCTION 1.1 GENERAL The engineering properties of soils needed for geotechnical analyses and designs are estimated either from results of laboratory or in-situ testing. Both procedures involve penetration of samplers or other rigid devices in the ground that inevitably cause disturbance to the soil. Soil disturbance due to sampling operations is of major concern to geotechnical engineers attempting to determine in-situ properties of soil by means of laboratory tests. Soil disturbance is often regarded as a significant problem because it is thought to prevent acquisition of realistic soil parameters. Deep penetration of soil samplers causes considerable shearing and distortion of the surrounding soil. In deep penetration problems, experimental observations (Robinsky and Morrison, 1964) indicate that soil deformations caused by penetration of a rigid indenter are similar in different soils even though the penetration resistance can be drastically different (i.e., soil stresses are different). This implies that deep steady penetration of a soil sampler is basically strain-controlled and that the associated deformations are not very sensitive to soil behaviour. Major improvements have been achieved to predict the behaviour of shallow foundations due to better understanding and identification of the important mechanisms governing foundation behaviour. The essential elements in improving predictive capabilities have been newly developed analytical procedures, better methods to characterise in-situ soil conditions and more reliable observations of field prototype behaviour. However, the same improvements could not be directly utilised in "deep" geotechnical problems. "Deep" geotechnical problems are referred to situations where soil of interest is relatively deep below ground surface compared to its lateral extent, for example, long piles, cone penetrometers, in-situ tests and soil sampling. The importance of soil disturbances caused by deep penetration of soil samplers has long been recognised. Significant research has been carried out to investigate the effects of sampling disturbance on the behaviour of clayey soils. Many investigators have attempted to establish the extent and nature of the disturbances associated with 1 sampling and laboratory testing. However, in view of the absence of analytical techniques to predict the effects of sampling on soil deformations and strains, most of this research has been limited to comparative experimental investigations. By treating soil as an incompressible, inviscid fluid an approximate analytical method was developed by Baligh (1975) for the prediction of deformations and strains caused by deep penetration of cones in saturated isotropic clays. The analytical method was later called Strain Path Method (Baligh, 1985). The Strain Path Method has also been used to predict deformations and strains for other "deep" geotechnical problems, e.g., closed and open-ended long piles, cone penetrometers and samplers (Chin and Baligh, 1983; Levadoux and Baligh, 1980; Baligh, 1985; Chin, 1986). This method is based on concepts similar to the Stress Path Method (Lambe, 1967) and consists of four basic steps: (a) initial stresses are estimated; (b) an approximate strain field satisfying conservation of volume, compatibility and boundary velocity requirements is estimated; (c) the deviatoric stresses at a selected number of elements are evaluated by performing laboratory tests on specimens subjected to the same strain paths or, alternatively, by using an appropriate computer based soil behavioral model; and; (d) the octahedral (isotropic) stresses are estimated by integrating the equilibrium equations. Solutions in accordance with the Strain Path Method consider the external diameter (B) to thickness ratio (B/t), also called aspect ratio, of the sampler, as the prime variable controlling the overall distortion pattern of the soil around the sampler. Although the influence of the exact geometry of thin-walled samplers has not been analysed, analyses conducted on simple samplers with round-ended walls and flatended walls have suggested no significant effect of sampler geometries on the strain history of soil elements on the centreline of the sampler. This conclusion certainly needs to be qualified as the comparison made by Baligh (1985) only implies that the two geometries studied are equivalent in terms of distortion. As recognised by many other researchers, the geometry of the cutting edge is a fundamental characteristic of a good sampler. There is strong evidence that the precise geometry of the cutting shoe of a sampler is of significant importance for quality sampling (Hvorslev, 1949; Kallstenius, 1958;zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA La Rochelle, 1973; La Rochelle et al., 1981). The geometry of the cutting shoe of a sampler has a marked influence on the strains and deformations of the soil around the sampler. This strongly suggests that extensive research should 2 be carried out to predict deformations and strains accurately due to penetration of samplers having different cutting shoe designs. During the sampling process soil is disturbed in two major ways:(a) Mechanical disturbance caused when the sample tube is pushed into the soil. This disturbance is termed as tube penetration disturbance. (b) Disturbance caused by the release of the total in-situ stress after the soil has been sampled. Such a disturbance is called "perfect" sampling disturbance or stress release disturbance. The effects of "perfect" sampling disturbance on the undrained stress-strain and strength properties of clays have been reported extensively by numerous investigators (Skempton and Sowa, 1963; Ladd and Lambe, 1963; Ladd and Varallyay, 1965; Seed et al, 1964; Noorany and Seed, 1965; Davis and Poulos, 1967; Adams and Radakrishna, 1971; Ladd and Foott, 1974; Kubba, 1981; Gens, 1982; Kirkpatrick and Khan, 1984; Jardine, 1985; Hight et al, 1985; Kirkpatrick et al, 1986; Graham and Lau, 1987; Baligh et al, 1987; Graham et aI, 1988; Hight and Burland, 1990). In contrast to "perfect" sampling, the stress-strain paths involved during penetration of tube samplers and subsequent extrusion are complex. The levels of distortion which occur as soil enters a sampler during penetration have been determined analytically by the application of the Strain Path Method. (Baligh, 1985). The strains involve triaxial compression followed by triaxial extension and triaxial recompression. Little experimental work has been done to understand the effect of tube penetration disturbances only on the undrained behaviour of clays. Limited results have been reported by Baligh et al (1987) and, Lacasse and Berre (1988) from tests performed on samples of reconstituted Boston Blue Clay and Drammen Clay respectively. The tests evaluated the effect of tube penetration disturbances on the undrained stressstrain and strength characteristics of normally consolidated and overconsolidated clays. The effects of varying degrees of tube penetration disturbances on subsequent stressstrain and strength behaviour of soils have not yet been studied. If the degree of tube penetration disturbances depends on the design of the cutting shoe of a sampler, it is essential that attempts should be made to evaluate experimentally the effects of imposing different degrees of tube penetration disturbances on subsequent stress- strain behaviour of clays. Such a study would also be useful in understanding thezyxwvutsrq 3 importance of cutting shoe designs in controlling the degree of tube penetration disturbances. 1.2 OBJECTIVES OF THE PRESENT WORK The main objectives of the present research work are as follows: (i) To develop an approximate numerical method of predicting strain paths of soil elements due to undrained penetration of samplers into soil. The numerical technique should enable the computation of the magnitudes of both radial and axial strains of soil elements. (ii) To study the strain histories of soil elements at different locations within the sampling tube when soil samplers of different cutting shoe geometries are penetrated into the soil. This should provide an integrated and systematic framework for assessing sampling disturbance effects due to penetration of samplers of various cutting shoe designs. Cutting shoe geometry of three types of samplers will be investigated. These are: a) the Norwegian Geotechnical Institute (NGI) 54 mm diameter piston sampler; b) the Swedish Geotechnical Institute (SGI) 50 mm diameter piston sampler; and; c) two typical British Standard General Purpose 100 mm diameter samplers. (iii) A parametric study of the effects of area ratio, inside clearance ratio, inside cutting edge taper angle and outside cutting edge taper angle of samplers on strain paths of soil elements will be carried out. Strain paths due to undrained penetration Bit ratio will also be investigated. of flat-ended samplers of different thickness andzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFE be made to perform stress and strain path tests on (iv) Finally, attempts willzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA reconstituted soft London Clay specimens in a computer controlled triaxial apparatus in the laboratory. In stress path tests, Ko-normally consolidated specimens will be subjected to monotonic undrained shearing in compression and extension to determine the "undisturbed" normally consolidated behaviour of the soil before disturbance. In strain path tests, different degrees of tube penetration disturbances will be imposed on Ko-normally consolidated specimens followed by undrained compression. shearing in These tests will evaluate the effect of applied disturbances on 4 subsequent undrained stress-strain, stiffness and strength properties of normally consolidated soft London Clay. Deformations, both radial and axial, will be monitored directly on the specimen. Pore pressure will also be measured locally at the mid-height of the specimen. 1.3 THE RESEARCH SCHEME In order to reach the goal, the whole research programme has been divided into the following phases: Phase 1: The current research proposes to perform strain path tests on soft clay specimens, using local strain measurement. A soft clay sample is likely to undergo large axial deformations during consolidation and shearing. Therefore, an axial strain measuring device which can measure relatively large strains was required. As a result, the first phase of the research considers the development of a gauge that can measure axial strains up to 9 to 10%. Phase 2: The development of an approximate numerical procedure to compute the value of radial and axial strains of soil elements due to axisymmetric undrained penetration of a sampler and then performing the necessary analyses as outlined in the research objectives earlier. Phase 3: The development of a computer program to run stress and strain path tests on Ko-nonnally consolidated reconstituted soft London Clay specimens automatically, in a triaxial apparatus in the laboratory. 1.4 THESIS LAYOUT A brief review of the previous work on sampling disturbance of cohesive soils is given in Chapter 2. The review of sampling disturbance is directed mainly towards previous work on soft clays although some significant work concerning other cohesive materials has also been considered. Chapter 3 presents the work done towards the development of an approximate analytical method to predict strain paths of soil elements due to axisymmetric 5 undrained penetration of a sampler.zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJ In this Chapter the strain paths of all the samplers are shown and their nature is discussed briefly, wherever necessary. In Chapter 4, preparation of Ko-normally consolidated soft London Clay in the laboratory is described. A brief introduction of local strain measurement and pore pressure measuring techniques for triaxial specimens is presented. The design of a new local axial strain measuring device capable of monitoring larger strains, and its development, are discussed in this Chapter. Calibration and programming of the automated stress path testing equipment are also described. In Chapter 5, theoretical investigations of test rates for Ko-consolidation have been presented. The test procedure followed to conduct stress and strain path tests has also been described in this chapter. Results from finite element analyses given in Chapter 3 and from laboratory experiments (reported in Chapter 4 and 5) are presented and discussed in Chapter 6. Firstly, the strain paths of samplers obtained from different series of analyses are compared separately. Secondly, the one-dimensional consolidation and permeability properties of normally consolidated London Clay are presented and discussed. Finally, stress and strain path tests results are shown. Undrained stress-strain behaviour observed in compression and extension are discussed and a comparison is made between the two responses. The effects of imposing varying degrees of tube penetration disturbance on the resulting undrained stress-strain, stiffness and strength properties are discussed, together with a comparison with previous investigations. Chapter 7 presents the conclusions of the research and includes general recommendations for future work. 6 CHAPTER 2 SAMPLING DISTURBANCE 2.1 INTRODUCTION The availability of good engineering parameters for geotechnical design depends on careful testing. Testing may be performed in the field or in the laboratory, but in both the cases the most significant factor controlling the quality of the results is likely to be the avoidance of soil or sample disturbance. The mechanisms of sample disturbance have been well understood since 1940s (Hvorslev, 1940 and 1949; Rutledge, 1944; Kallstenius, 1963; Broms, 1980; Clayton et al, 1982; Hight and Burland, 1990). Disturbances to soil in its widest sense occur during drilling, during the process of sampling itself and after sampling. A number of different procedures are adopted for measuring, analysing and correcting the effects of sampling disturbance and, in order to highlight the importance of the present research, it is necessary to review previous investigations on the problem of sample disturbance. There has been a wide range of reported observations on the effects of sampling procedures on different types of soils. Some direct investigations considered the effects of major causes of disturbances on the stress-strain and strength properties of soils while other indirect observations were concerned more with the design, use and maintenance of samplers and the development of sampling techniques. In this chapter the previous investigations done on the subject of sampling disturbance are reviewed. The effects of sampling disturbance on the engineering properties of soils, particularly soft clays, and methods of correcting sampling disturbance effects are discussed. Analytical methods for predicting soil disturbance are also presented. 2.2 CAUSES OF SAMPLING DISTURBANCE The physical process of obtaining samples has been recognised as a prime cause of sample disturbance. Causes of sampling disturbance have well been identified in the past (Hvorslev, 1949; Rutledge, 1944; Kallstenius, 1958; La Rochelle et aI, 1981). The main causes of sampling disturbance can be stated as follows: (i) Disturbance of the soil to be sampled before the beginning of sampling as a resultzyxwvuts 1 of poor drilling operation. (ii) Mechanical distortion during the penetration of the sampling tube into the soil. (iii) Mechanical distortion and suction effects during the retrieval of the sampling tube. (iv) Release of the total in-situ stresses. (v) Disturbance of the soil during transportation, storage and sample preparation. The first cause can be reduced by sampling with properly cleaned boreholes advanced by using bentonite slurry. sampler design unavoidable The second and third causes are directly associated with and can be controlled to certain extent. The fourth cause is even though its effects may be different depending on the depth of sampling and soil properties. The fifth cause can be reduced by storing samples for minimum atmosphere time in controlled transportation and careful handling of samples during and preparation. Mechanisms and causes of sampling disturbance have been summarised by Clayton (1986). Methods of amelioration of soil disturbance have also been outlined. 2.3 INVESTIGATIONS OF SAMPLING DISTURBANCE 2.3.1 DISTURBANCE DURING DRILLING A soil sample can be disturbed during drilling operation. Compaction, remoulding and displacement of soil beneath or around casing or sampler tubes driven ahead of an open borehole can occur as a deliberate method of advancing a borehole. boring rigs operate on the percussion drilling principle. Many This type of displacement drilling leads to significant remoulding and compaction of the soil around and ahead of the bit. Similar effects can be caused during the most common types of site investigation drilling, principally when using augers or light percussion drilling in soft soil. Most rigs using continuous flight augers are capable of providing considerable downward thrust. In very soft clays the soil may block flights and fail to travel up to the ground surface. Soil displacement then becomes inevitable. Light percussion boring can induce similar problems if casing is advanced below the bottom of the open hole. A plug of soil will form inside the base of the casing and lead to compaction, compression and bearing capacity failure immediately below the bottom 8 of the casing. Casing should never be allowed to go below the bottom of the borehole at any time during drilling; in this case samples taken through the bottom of the casing will probably be highly remoulded if clays, or compacted if sand or gravels (Clayton et al, 1982). Reduction in total vertical and total lateral stresses due to removal of soil from the borehole is another principal cause of sample disturbance during drilling. Swelling at the base of borehole occurs as a consequence of stress relief. The process is fast and unavoidable in granular soil; in cohesive soils, however, swelling can be reduced by sampling as quickly as possible following boring. The amount of swelling that occurs is proportional to the change of total stress occurring at the base of a borehole. Thus if the borehole is substantially empty of water there is likely to be more swelling than if the borehole is kept full of mud or water. Other severe effects of stress relief during drilling on soil are base heave, piping and caving (Clayton et al, 1982). Base heave can be thought of as foundation failure under decreased vertical stress. When the total stress relief at the base of a borehole is very great compared with its undrained shear strength, plastic flow of soil may take place upwards into the borehole. Failure in a borehole by base heave can occur in very soft soils if the water level is kept too low (Begemann, 1977). When a borehole is inducing total stress relief, and water balance is insufficient to prevent high seepage pressure gradients in the soil at the base of the hole, large volumes of fine granular soil may move up into the casing. Soil below the bottom of the casing will be brought to a very loose state. This phenomenon is called piping. Both base heave and piping can be reduced by keeping the hole full of water. Caving typically occurs when boreholes are advanced into soft, loose or fissured soils. Material from the sides of the borehole collapses into the bottom of the hole and must be cleaned out before sampling can take place. 2.3.2 DISTURBANCE DURING SAMPLING The change of volume resulting from the intrusion of a sampling tube into a soil mass produces appreciable distortions. Hvorslev (1949) described the forces acting on an element of soil while it is being tube sampled. There are two main forces associated with sampling. The first is that occurring as the soil is displaced by the advancing cutting edge. This can cause quite considerable shear strains, and possiblyzyxwvutsrqp 9 large forces. This disturbing effect is reduced by decreasing the cross-sectional area of the cutting edge. The second disturbing force in the soil during tube sampling is that caused by friction or adhesion between the soil and walls of the sampler. Hvorslev (1949) considered that friction on the internal wall would be more significant than that on the outside wall, causing the structure of the sample to be altered. Bjerrum (1973) also reported that due to friction between the clay and the sampling tube, the outer zone of the sample becomes remoulded. The volume of these zones of badly disturbed clay and the degree to which the original structure of the clay in these zones is destroyed are.however, not the same in all types of clay. The greatest amount of disturbance is, for instance, experienced in clays of low plasticity. Clays with pronounced cohesive properties will undergo less disturbance. The same is the case with highly sensitive or quick clays, the remoulded strength being so low that the friction between clay and sampling tube is practically eliminated. In a soft clay, remoulding at the periphery produces large positive pore pressures. During the period following sampling the pore pressures tend to equalise with those in the core of the sample causing an overall increase in porewater pressure (Kallstenius, 1971; Bjerrum, 1973; Schjetne, 1971). Kallstenius (1971) found that the outer zone of soft clay samples (wzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGF = 113.3-117.6%, LL = 135-145, PL = 38-41) was on average 1.5% dryer than the core of the sample, and this was due to water redistribution associated with the equalisation of pore water pressure. Bjerrum (1973) has shown that due to remoulding and moisture migration, the outer 5 mm of extruded plastic Drammen Clay specimens (w = 52%, LL = 61, PL = 32) typically have a moisture content,about 3 to 4% lower than at the centre. Bjerrum concluded that the swelling of the core of the sample associated with water content redistribution was one of the major factors causing disturbance. Schjetne (1971) measured the porewater pressures in soft clay samples of different sensitivities with a NOI piston sampler during and immediately after sampling. He found that immediately after sampling the porewater pressures were negative and of significant magnitude. However, they began to increase and eventually were only slightly negative or zero. Schjetne (1971) considered that this occurred because of the equalisation of pore pressures, following the shearing of a thin peripheral zone of the sample. Apted (1977) investigated the fundamental causes of sample disturbance for 10 overconsolidated stiff London Clay in relation to undrained strength. It has been reported that during tube sampling the outside of the sample is intensively sheared. This causes a decrease in porewater pressure in this zone, which on equalisation with the rest of the sample causes an increase in effective stress in the sample. This process leads to the increased water contents measured at the periphery of tube samples. This finding, however, contrasts with those reported by Kallstenius (1971) and Bjerrum (1973) who found decreased water contents in the outer zone of plastic soft clay samples. Alonso et al (1981) carried out theoretical investigations upon soil stressing and straining around the sampler during the sampling operation in saturated cohesive soil. The research was carried out with the aid of a viscoplastic model of the saturated clay which was implemented via the Finite Element Method. Preliminary results show that the model may be used to investigate the effects of sampler geometry, side friction, velocity of driving and constitutive behaviour of soil. Another important contributory factor to disturbance during sampling is due to release of in-situ total stresses. In response to the reduction of applied total stresses, the pore pressures in a sample will reduce and may normally be expected to become If the sample is coarse-grained, it will have a high coefficient of negative.zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA permeability and a large average pore size and water or air will rapidly penetrate it and dissipate the negative pore pressures. Thus, with total and effective stresses reduced to zero, a granular soil has little strength and is very difficult to sample or prepare for laboratory testing. In a cohesive soil. however. a small average pore size normally precludes the penetration of air. Because of low permeability a considerable period of time may be required for water to penetrate and dissipate the negative pore pressures set up in the sample. Disturbance caused due to the release of in-situ total stresses is called stress release or "perfect" sampling disturbance. A sample which has received no disturbance other than that involved with the release of in-situ total stresses is termed "perfect" sample. This is an imaginary sample used as a concept by researchers who are trying to look separately at one of the two main parts of sampling disturbance. Several workers have investigated the effects of "perfect" sampling disturbance on the undrained stress-strain, stiffness and strength properties of clays. These will be discussed in detail in section 2.7.1. 11 2.3.3 DISTURBANCE CAUSED BY TRANSPORTATION AND STORAGE The disturbance caused by vibrations and shock loads during transport of samples to the laboratory was investigated by Kallstenius (1971) by loading samples of soft sensitive clay in axial compression. The average reduction of shear strength which was measured by fall cone tests was about 30% when the axial compression of the sample was 20%. The effect of shock loads was investigated by dropping samples stored in plastic tubes from a height of 1.0 m on a floor of soft asphalt. The reduction of the undrained strength as measured by the fall cone test increased with increasing liquidity index of the soil. For a clay with a liquidity index of 1.3 the maximum reduction was approximately 30%. Both an increase (maximum 26%) and a reduction (maximum 22%) of shear strength were observed when samples of soft clay were vibrated to simulate the disturbance from transport by train. Of the secondary disturbances or changes of samples which may occur during transportion, Kallstenius (1971) found that the greatest and most easily avoided are shocks, frost and great deformation of the soil. Of great importance are further partial remoulding and redistribution of water within the samples, partial freezing and strong vibrations. Samples are usually sealed and stored for some period of time before testing and this delay may cause further alternations to the clay structure. Migration of water within the sample may still lead to significant changes of properties such as compressibility and undrained strength. Two types of effects have been noted. Firstly, water migrates from one type of soil to another (Kimball, 1936; Rowe,. 1972) and secondly, differential residual pore pressures in the samples equalise with time (Kallstenius, 1971; Schjetne, 1971; Bjerrum, 1973). These two effects have been explained by Clayton et al (1982) for a laminated soil and a very stiff fissured clay of high plasticity, such as the London Clay. In situ the laminated soil, containing alternate layers of silt grading into fine sand and clay, might have a firm consistency, but once stress relief occurs the water in the granular layers will migrate to the clay and relieve the negative excess pore pressures. Upon examination, the soil might appear to consist of very soft clay layers interbedded with relativelyzyxwvutsrqponmlkjihgfedcbaZYXWVUTS dry silty sands. In case of stiff clay, however, after sampling, the bulk of the sample will be expected to have similar effective stresses to its original state in situ. This part of the sample will also have quite large negative porewater pressures. The outside of the sample will have been remoulded and the pore pressures will be much lower. 12 During the period following sampling the pore pressures in different parts of the sample will equalise with time. Pore pressures in the outer part of the sample will decrease and the soil will consolidate. In the central part of the sample the pore pressures will increase, and the soil will swell and become weaker. Because of the relatively small volume changes required for swelling and effective stress decrease, the higher pore pressures in the outer region of the sample can have a large time dependent effect on the overall undrained shear strength of the sample. Shear strengths measured immediately after sampling will be higher than those measured after the sample has been transported to a laboratory and stored for some time. Bozozuk (1971) observed that a storage period of 15 months reduced the measured preconsolidation pressure by about 5% for a soft marine clay. Although this reduction is not great, it does indicate that consolidation tests should be performed as soon as possible after the samples are obtained. Bjerrum (1973) found that for a quick clay, the undrained shear strength decreased by 15% after 3 days in comparison with samples that were tested immediately after sampling. This effect was found to increase with decreasing plasticity of the clay. Bjerrum (1973) concluded that the gradual reduction of shear strength in soft clays with time was attributed to a gradual reduction of the initial residual porewater pressure that occurs during storage of the samples. The reduction of the porewater pressure appears to be larger for clays of low plasticity than for clays of high plasticity. A change of the water content due to internal migration of porewater will have a relatively large effect for clays of low plasticity. Also the difference in permeability of clays with different liquid limit and plasticity index will contribute to the effect. The effect of long storage times on the shear strength and consolidation characteristics were investigated by Arman and McManis (1976) for specimens obtained from thin-walled tubes and hand-cut blocks. The variations of the undrained strength and preconsolidation pressure with storage time are shown in Figs. 2.1 and 2.2 respectively. It was found that the shear strength of specimens obtained from tubes decreased as a result of long-term storage. The preconsolidation pressure of stored tube samples also followed a similar trend as is evident from Fig. 2.2. Figs. 2.1 and 2.2 also show that long-term storage does not affect the strength or 13 preconsolidation pressure for specimens obtained from blocks. Arman and McManis (1976) recommended that tube samples should be tested within 15 days after sampling to prevent erroneous results, due to deteriorating effects of long-term storage. La Rochelle et al (1976) investigated the effect storage time on strength and consolidation properties of two sensitive cemented clays from Saint-Louis and Saint Jean-Vianney in Canada. When comparing results of unconfined compression tests performed in the field immediately after sampling or in the laboratory the following weeks, it was observed that there was a decrease in the measured strength which was attributed to water migration. Triaxial tests carried out on isotropically consolidated block samples within a few weeks after sampling showed that the undrained shear = 50, strength had decreased by 10 to 15% for more plastic Saint Louis clay (LLzyxwvutsrqponm PI = 23) and by 14 to 21% for less plastic Saint-Jean-Vianney clay (LL = 29, PI = 11). The strain at failure increased appreciably in the case of Saint-Louis clay but not in the case of Saint-Jean-Vianney clay. Consolidation tests made on block samples after prolonged periods of storage indicated no change in the value of preconsolidation pressure for samples from either site. Arman and McManis (1976) also reported similar observations. Kirkpatrick and Khan (1984) reported the effect of age on the undrained stress-strain properties of "perfect" and "in-situ" samples. The study was conducted on two clays, kaolin (PI = 30) and illite (Pl = 40), prepared in the laboratory from consolidated slurries. Large losses of undrained strength were observed. These losses amount to about 34% for more plastic illite and 47% for less plastic kaolin of the "in-situ" strength after 5 or 6 hours. The losses increase with age reaching about 50% and 72% for the two clays after 50 days storage. Secant modulus values for "perfect" samples progressively dropped with sample age. Secant modulus values calculated at the maximum deviator stress were decreased by about 23% for illite and 38% for kaolin after 50 days storage although secant modulus values calculated at half the maximum deviator stress were decreased by approximately same amount (55% for illite and 56% for kaolin) for both the clays. The values of pore pressure parameter A at failure,zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA At for both "perfect" samples of illite and kaolin were also changed. For both illite and kaolin, At showed a gradual decrease in value with the increase in sample age. "Perfect" samples, however, when consolidated under Ko-conditions 14 and then tested,zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA it was found that sample age had little or no influence on behaviour up to ages of one month as long as water contents of the samples were unchanged. Graham et al (1987) investigated the effect of storage time on the undrained stressstrain characteristics of normally consolidated and overconsolidated "perfect" samples of reconstituted illite. "Perfect" samples were stored "undrained" (or rather, at constant water content), reconsolidated and sheared undrained. The storage times were 15 min., 1 day and 7 days. Results showed that different "undrained" storage had no effects on the strengths after reconsolidation. This agrees with the conclusions drawn by Kirkpatrick and Khan (1984). From the work of Graham and Lau (1988) on normally consolidated and overconsolidated "perfect" samples of illite it was found that, in general, the undrained strengths of reconsolidated samples after different "drained" storage times (15 min., 1 day and 7 days) decreased with increasing storage time. This, therefore, represents a change from the results after "undrained" storage reported by Graham et al (1987) and, Kirkpatrick and Khan (1984), where the strengths were found to be independent of storage time. 2.3.4 DISTURBANCE CAUSED BY SAMPLE PREPARATION A major contributory factor to disturbance is the extrusion of the sample from the sampling tube. The force required to extrude a soil sample was investigated by Sone et al (1971) for a clayey silt. It was much larger than the unconfined compressive strength of the soil. The undrained shear strength was reduced 10% to 20% by the extraction up to 10 cm to 20 cm from the bottom of the sample. Arman and McManis (1976) also examined the extrusion stress for tube samples of very stiff clay. Soil cores were extruded using hydraulically operated pistons. During core extrusion, the end of the sample in contact with the piston began to show measurable displacements before the opposite end. Thus internal displacements were occurring within the tube. The maximum strain at the piston end varied from 0.001 to 0.005. In all cases, the applied stress exceeded the unconfined compressive strength of the soil to a maximum of 900%. X-ray radiography was also carried out to determine the extent of disturbances in the extruded soil cores. Radiographs showed two distinct distortion effects caused by extrusion process. The first type of distortion, observed in all cores, was a gradual bending of the soil layers, with a maximum at the tube surface and decreasing toward the centre. The bending was a symmetrical 15 dome-like effect around the longitudinal axis. This effect was more pronounced in the soft marine deposits tested than it was in the stiff pleistocene clays. A second, more serious type of disturbance was a definite failure plane pattern in some of the stiff clays. These failure planes, either at a 500 angle or with a classical cone shape, occurred at intervals along the longitudinal axis of some specimens. Shackel (1971) used a nuclear technique as a useful method for assessing sample disturbance in a non-destructive way. The main advantage of the technique is that point to point measurements are possible along the sample both before and after extrusion from the sampler. Fig. 2.3 shows the variation in bulk density in a sample of stiff sandy clay taken by a thick-walled open-drive sampler before and after extrusion from the sampler tube. Fig. 2.3 shows that the extrusion process increases the densities at bottom part of the sample whereas it generally decreases the densities at the top part. The increase in density at the bottom end is believed to result from local compression at that end of the sample to which the extrusion force has been applied. Several factors affect the accuracy and usefulness of the nuclear technique in the observation of sampling disturbance. These include variations in soil composition and in the dimensions of the sample and changes in the particle size distribution. Thus the method is unsuitable for use where samples are obtained from granular materials since the dimensional variability of such samples is often large. The method is also unsuitable for samples taken from varved or laminated soils and where sampling technique may cause appreciable particle degradation. The largest reduction in shear strength is generally obtained at the end of samples from which the extrusion is performed and the disturbance increases as the length to diameter ratio increases and would be excessive when the ratio exceeds 14 (Kallstenius, 1963). Apted (1977) reported that loss of moisture during trimming, setting up, etc. of as little as 0.1% of the dry weight of the sample could produce a significant change in effective stress. The disturbance effects due to sample preparation were investigated by Kimura and Saitoh (1982). Variation of pore pressures during extrusion and trimming were monitored with an embedded small pore pressure transducer for two Kawasaki Clays 16 (PIzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA = 50 and 30). Both the negative pore pressure and mean effective decreased with the progress of sample preparation. stress Due to sample preparation the undrained strength was reduced by 15% and 30% for the clays with plasticity index of 50 and 30 respectively. 2.4 ASSESSMENT OF SAMPLE DISTURBANCE The mechanical properties of soils are modified by sampling disturbance and hence, they can be used to calculate properties of in-situ the amount soils are required of disturbance as references quantitatively. in calculating The disturbance. However, there is no way of obtaining a soil sample so as to maintain exactly the in-situ conditions. This is because its removal involves a change in the in-situ state of stress and usually some disturbance due to sampling and handling. disturbance can be estimated by investigating So, degree of the behaviour of the least disturbed sample. Because of additional disturbances other than that occurred due to total stress release, the residual effective stress of a disturbed sample,zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJ e', is usually less than the effective stress, 0'pi of a "perfect" sample. The isotropic effective stress in a "perfect" saturated sample of clay which had in-situ vertical and horizontal effective stresses of e', and Koo'y respectively, is given by the following expression (Ladd and Lambe, 1963; Ladd and Varallyay, 1965):zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHG (J ~B = (J : [ K0 + A u( Where K, is the coefficient 1 - Ko) ] .... (2.1) of earth pressure at rest' and Au is the pore pressure parameter for the undrained release of the total stresses which existed at the conditions. The parameter Au for a saturated clay (i.e., Skempton's Ko- B parameter is equal to unity) is given by .... (2.2) Where, Au is the pore pressure change; and Aa and Aab are the changes of vertical y and horizontal total stresses. unity. Equation (2.1) is valid for both Ko less and greater than Skempton and Sowa (1963), Seed et al (1964) and Noorany and Seed (1965) also presented equations similar to Equation (2.1) but their form of Au used Aa, and 17 Acr), which changes in direction when K, becomes greater than unity. A number of investigators have defined the degree of disturbancezyxwvutsrqponmlkjihgfedcba (0) in terms of cr'pe e'; These are as follows: andzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA (a) Ladd and Lambe (1963) proposed that disturbance could be defined as .... (2.3) (b) Noorany and Seed (1965) regarded the difference between cripe and cr r as a measure of disturbance, i.e., D=O"zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA .... (2.4) ps -0" r (c) Okumura (1971) and Nelson et al (1971) defined the degree of disturbance as .... (2.5) Broms (1980) pointed out the main difficulty with the above definitions of sample disturbance.is the calculation of cr'pe (Equation 2.1) which is based on assumed or measured values of K, and Au. therefore, the calculated cr pe Such measurements are difficult to carry out and, is subjected to significant errors caused by inaccurate values of K, and Au. Furthermore, it is also difficult to measure o', of a disturbed sample. Direct and indirect methods of measuring the residual effective stress of a disturbed sample wer~ proposed by Skempton (1961) and Lambe (1961). Different methods of measuring the residual or initial effective stress in clays have been summarised by Baldi et al (1988) and Hight and Burland (1990) Degree of disturbance has also been derived from the consolidation characteristics of soils. Rutledge (1944) recognised the effects of disturbance on the one-dimensional compression behaviour for a range of plastic soils. He observed that disturbance shifts the e-Iog cry curve downwards, decreases its slope and obscures the previous stress history of the soil and its preconsolidation load. On this basis. Schmertmann (1955) suggested that the e-log ay curve of a remoulded sample and a virgin sample may be used as the lower and upper limits respectively to indicate the degree of disturbance. The one-dimensional compression curve of a virgin sample may be constructed theoretically and the maximum preconsolidation pressure may be estimated. For comparison the compression curve of an undisturbed sample may be observed in the laboratory. The more the sample is disturbed, the more its position 18 is shifted towards the remoulded behaviour. Schmertmann (1955) defined this shift at the estimated preconsolidation pressure as the degree of disturbance; D =zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA LI e / LI ezyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA max .... (2.6) where Aemu is the difference in void ratio between the two limit curves at the estimated preconsolidation pressure and Ae is the difference in void ratio between the virgin behaviour and the undisturbed behaviour as shown in Fig. 2.4. Alternatively, Bromham (1971) measured disturbance in samples of soft clay from the slopes of oedometer curves in consolidation tests. Fig. 2.5 shows a typical reconstructed field (virgin) consolidation curve with laboratory and remoulded curves. The laboratory curves are normally corrected with respect to the estimated field values of preconsolidation pressure, P, and the initial void ratio eo of the soil. The corrected curves were assumed to pass through the same point at e = 0.42eo as suggested by Schmertmann (1955). Bromham (1971) defined the disturbance factor, X as follows: X = 100 (p a P I ) Pa-Pr + .... (2.7) where, PI' PI and P, are shown in Fig. 2.5. In practice, P/p. is small and hence, the disturbance factor can be expressed as follows: .... (2.8) Equations 2.6 and 2.8, defining degree of disturbance, were derived on the basis that the preconsolidation pressure is significantly affected by sampling disturbance. Recent observations, however, made by different authors tend to disagree on this particular point. It has been found that the value of preconsolidation pressure measured by oedometer test is not influenced appreciably by sampling disturbance (La Rochelle and Lefebvre, 1971; Bozozuk, 1971; La Rochelle et aI, 1981). Thus the value of preconsolidation pressure can not be considered as a reliable criterion for defining the degree of sample disturbance. Sampling disturbance is generally considered to have greater effect on the stress19 strain properties of a soil than on its in-situ shearing strength, c, (Hvorslev, 1949). Raymond et al (1971) introduced the concept of failure index as an indicator of the effect of sampling disturbance on strain. Failure index (FI) is the ratio of the deviator stress at certain compression to the deviator stress at failure and can be defined algebraically as follows: .... (2.9) In addition, the undrained modulus E, at any stress difference has been defined by the secant modulus as follows: .... (2.10) Then the value of Ejc zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA will be given by twice the failure index divided by the y corresponding strain. This definition obviously has the added advantage of being more general than using only the value at a particular failure index. property Ejc y A derived (which is proportional to the reciprocal of strain for any specified value of failure index) will thus indicate the effect of sampling on the degree of disturbance. The higher the value of Ejc y at a specified failure index, the lesser is the degree of disturbance and vice versa. For clays, the magnitude of volumetric strain when consolidating the specimen to the in-situ effective stress, £0 is a useful index to define sample disturbance. Lacasse and Berre (1988) reported the following criterion to evaluate sample quality from volumetric strain of soft sensitive onshore clay specimens, measured during anisotropic consolidation to the in-situ stresses: ~ (%) <1 Sample quality Very good to excellentzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIH 1 to 2 good 2to4 4to8 fair poor >8 very poor 20 2.5zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA SAMPLER DESIGN AND ITS EFFECT ON SAMPLE DISTURBANCE The design of a sampler is one of the most important factors that should be considered for quality sampling. The amount of disturbance varies considerably depending upon the dimensions of the sampler and the precise geometry of the cutting shoe of the sampler. Hvorslev (1949) discussed at length the importance of the design of a sampler and introduced the concepts of area ratio, inside and outside clearance ratios and cutting edge taper angle in controlling sampling disturbance. 2.5.1 EFFECT OF AREA RATIO AND CUTTING EDGE TAPER ANGLE Area ratio is considered one of the critical parameters affecting the disturbance of soil during sampling. Hvorslev (1949) defined area ratio as follows: .... {2.11) Where D. is the external diameter of the sampler tube and D, is the internal diameter of the sampler cutting edge as shown in Fig. 2.6. Increasing area ratio gives increased soil disturbance and remoulding. The penetration resistance of the sampler and the possibility of the entrance of excess soil also increase with increasing area ratio. For soft clays; area ratio is kept to a minimum by employing thin-walled tubes. For composite samplers, the area ratio, however, is considerably higher. In these cases, sample disturbance is reduced by tapering the outside of the sampler tube very gradually from a sharp cutting edge (Hvorslev, 1949, recommended a maximum 1O~, so that the full wall thickness is far removed from the point where the sample enters the tube. Jakobson (1954) investigated the effect of sampler type on the shear strength of clay samples. Samples were collected using nine different types of samplers. These types differ from one another in area ratio, edge angle, inside clearance, drive velocity and other factors. Shear strength of samples were determined by carrying out the unconfined compression tests, the cone test and the laboratory vane test. It was found that an extremely small area ratio offers no special advantages and that the cutting edge taper angle does not seemzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGF to have any great influence. However, a very large area ratio or cutting edge taper angle is not recommendable. 21 Kallstenius (1958) also studied the effect of area ratio and cutting edge taper angles on the shear strength of Swedish clays. He carried out tests similar to those reported by Jakobson (1954) on samples Kallstenius recommended obtained using six types of piston samplers. that a sampler ought to have a sharp edge and a small outside cutting edge taper angle (preferably less than 5°). He also considered that Hvorslev's concept of area ratio need not be regarded as an absolute criterion provided that the edge angle is small. The combined requirements for area ratio and cutting edge taper angle to cause low degrees of disturbance were proposed by the International Society for Soil Mechanics and Foundation Sampling Engineering's (1965). Sub-committee For samplers on Problems and Practices of about 3 inch diameter of Soil they suggested the following combinations of area ratio and cutting edge taper: Area ratio (%) Outside cutting edge taper (degree) 5 15 10 12 20zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA 9 40 5 80 4 2.5.2 EFFECT OF INSIDE AND OUTSIDE CLEARANCES Inside wall friction is one of the principal causes of disturbance (Hvorslev, 1949). One of the methods of reducing or eliminating of the sample wall friction between the soil and sampler is to provide inside clearance by making the diameter of the cutting edge, 01, slightly smaller than the inside diameter of the sampler tube, D, The inside clearance ratio is expressed as follows (Fig. 2.6): Inside Clearance Ratio =D -D1 _8 __ DI .... (2.12) Inside clearance gives the soil sample room for some swelling and lateral strain due to horizontal stress reduction. Although neither of these types of behaviour desirable, they are less undesirable than the consequences is of adhesion between the soil and the inside of the sampler tube (Clayton et al, 1982). Inside clearance should be large enough to allow partial swelling and lateral stress reduction but it should not 22 allow excessive soil swelling or loss of the sample when withdrawing from the sampling tube. Hvorslev (1949) suggests an inside clearance ratio of 0.75 to 1.5% for long samplers and 0 to 0.5% for very short samplers. Kallstenius (1958) on the basis of Swedish clays sampled by six different piston samplers, also recommends that a sampler ought to have a moderate inside clearance. The clearance reduces the wall friction and probably counteracts to a certain extent the disturbance from displacement of soil caused by the edge and sampler wall during the driving operation. If the inside clearance and the edge angle are moderate, the above positive effects outweigh the disturbance caused by deformation when the sample tends to fill the clearance. The existence of inside clearance may have detrimental effects on sample disturbance as pointed out by La Rochelle et al (1981). They reported from the work of Sarrailh (1975) that, in general, a "reshaped" 54 mm sampler without inside clearance seemed to give better results than a 54 mm sampler piston tube sampler with inside clearance. The improvement in strength was of the order of 20% or more and the tangent moduli were higher by 50-100%. Based on these observations, La Rochelle et al developed a new sampler with no inside clearance for sampling in soft sensitive soils. This sampler, called the Laval Sampler, is of large diameter (208 mm inside diameter and 218 mm outside diameter) and also without a piston. The area ratio,zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA BIt ratio and outside cutting edge taper angle of this sampler are 10%, 43.6 and 5° respectively. The sampler can recover a 600 mm length of sample and the sampler tube is over-cored to reduce disturbance it is being withdrawn. Drawing of the sampling and coring tube is shown in whenzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA Fig. 2.7. To evaluate the performance of the Laval Sampler, unconfined compression tests and consolidated undrained triaxial tests were performed on samples taken by the Laval Sampler and the results were compared with those obtained for block samples. It was concluded that stress-strain characteristics of the Laval samples were very similar to those observed on block samples. In order to reduce outside wall friction, samplers are often provided with outside clearance which is expressed as follows (Fig. 2.6): Outside Clearance Ratio = Ow-D. D. .... (2.13) An outside clearance ratio of a few per cent may decrease the penetration resistance of samplers in cohesive soils. Although outside clearance increases the area ratio, 23 a clearance of 2 to 3% can be advantageous in clay (Hvorslev, 1949). From the foregoing description, it is apparent that substantial research has been carried out to understand the effects of area ratio, outside cutting edge taper angle and inside clearance on sample disturbance. However, a systematic evaluation of how these parameters affect sample disturbance has not been carried out. It is also evident that the effect of inside cutting edge taper angle on soil disturbance has not been investigated. Of course, the best way to investigate the effects of all these parameters would be to vary each of them separately, while keeping all the other constant, in a range of uniform soil conditions. 2.6 EFFECT OF SAMPLER DIMENSIONS, SAMPLER TYPES AND SAMPLING METHODS ON SAMPLE DISTURBANCE A number of workers have reported the effect of sampler dimensions, particularly diameter of the sampler tube, upon soil disturbance. Effect of different types of samplers and the method of sampling on soil disturbance have also been reported. Hvorslev (1949) stated that the amount of disturbance would be decreased with increasing diameter of the sample. Berre et al (1969) observed that larger tube samples showed more consistent behaviour than those from small tube samples. Oedometer tests carried out on samples of soft marine clay in Norway indicated that = 14%, inside clearance ratio, IeR = 1.4%) a 95 mm piston sampler (area ratio, ARzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJ gave less disturbance than a 54 mm piston sampler (AR = 12%, IeR = 1.3%). Bozozuk (1971) performed undrained triaxial tests on 1.4 inch diameter samples of soft marine clay. Samples were obtained by the 54 mm NOI piston sampler (AR = 11%, IeR = 1%) and the 127 mm Osterberg piston sampler (AR = 6%, IeR = 0.42%). Test results showed that the undrained strengths of samples cut from 127 mm tube sample were higher than those cut from 54 mm tube samples. Samples cut from 54 mm tube samples showed lower stiffness and pore pressure responses. Stress-strain and pore pressure characteristics of two clays cut from 54 mm and 127 mm tube samples are shown in Fig. 2.8. It appears that larger diameter tube samplers provide better samples of soft clay but further comparative tests need to be carried out. 24 Conlon and Isaacs (1971) carried out unconsolidated undrained triaxial compression tests on specimens of sensitive lacustrine clay of medium to high plasticity. The clay was sampled using 73 mm outside dia. thin-walled shelby tube (ARzyxwvutsrqponmlkjihgfedcba = 9.3%) and 127 mm outside dia. fixed-rod thin-walled piston sampler (AR = 10.8%). Some 51 mm thin-walled tube samples were obtained in wash borings and auger holes. samples were also collected. Block Conlon and Isaacs (1971) observed that disturbance increased as the size of the tube sample decreased. The strengths of tube samples were found, in general, to be less than the block samples; the magnitude of scatter was, in fact, not much greater. The undrained shear strength as a function of degree of disturbance has been shown schematically in Fig. 2.9. An investigation of the difference in quality of samples taken with large diameter fixed piston samplers and the 50 mm diameter Swedish Standard piston sampler (AR = 21%, ICR = 0.4%, outside cutting edge taper angle = 5~ was carried out by Holm and Holtz (1977). The large diameter piston samplers used were the 95 mm NO! research sampler (AR = 14%, ICR = 1.4%, outside cutting edge taper angle = 10~, the 127 mm Osterberg sampler (AR angle = 7~and = 18%, ICR = 0.4%, outside cutting edge taper the 124 mm SOl research sampler (AR = 27%, ICR angle of cutting edge = 5~. = 1.2% and The investigation has shown that in general no significant differences between either the ratio (preconsolidation pressure/in-situ vertical stress) or undrained shear strength derived from laboratory tests on specimens obtained by the various devices, but there are indications that: (a) Results of oedometer tests on 50 mm samples are more scattered, supporting findings of Berre et al (1969). (b) The undrained modulus obtained from 50 mm samples are lower. Holm and Holtz (1977), however, concluded that for routine investigations in soft Swedish clays, there ~eems to be no need to perform sampling with large diameter piston samplers. Kubba (1981) investigated the effect of thickness of tube on sampling disturbance for a reconstituted spestone kaolin (LL = 51, PI = 30). Tube samples were obtained by inserting 38 mm diameter tubes of different wall thicknesses into a 102 mm diameter "perfect" sample. Three tubes of thickness to diameter ratios of 0.039, 0.072 and 0.105 were used for sampling. Kubba (1981) found that increasing the ratio of wall thickness to diameter of the tube caused a qualitative increase in the degree of 25 disturbance. sampler. Sample quality is also related to the length to diameter ratio of the One of the major factors controlling sample jamming is the length to diameter ratio of the sampler. The optimum length to diameter ratios suggested for clays of different sensitivities are as follows (the Report of the Sub-committee on Problems and Practices in Soil Sampling, 1965) Sensitivity, S, length to diameter ratio >30 20 5 to 30 12 <5 10 McManis and Annan (1979) investigated the effect of sampling on the properties of undisturbed soil specimens. The soil types studied were soft organic silty clays and stiff, fissured pleistocene clays. Sampling was performed using 76 mm and 127zyxwvutsrqpo mm thin-walled open-drive tubes and by hand cutting of block samples. The test results were found to be dependent on the sampler types and type of soil. For stiff fissured clay, the strength of the 76 mm diameter tube sample exceeded that of 127 mm diameter specimen. This was attributed to stress release and migration of moisture toward and along the fissure planes. Maguire (1975) also found that for stiff fissured overconsolidated clay the undrained strength increased with decreasing diameter of sample. However, for soft silty clay, McManis and Arman (1979) found that 127 mm tube specimens exhibited strengths greater than that of 76 tube specimens. They also observed that specimens cut from block provided higher undrained strengths than the tube specimens. The influence of sampling methods on some soil properties for two sensitive slightly overconsolidated clays was reported by Milovic (1971a). Clay samples were obtained by Shelby tubes and Norwegian piston sampler. The area ratio and inside clearance ratio for both Shelby tube and piston sampler were respectively 12 ± 1.5% and 0.8 ± 0.1%. Cubic blocks were cut by hand. The unconfined compressive strength, the secant modulus, shear strength parameters and consolidation parameters of these sensitive clays, determined on Shelby and Piston specimens, were systematically lower than those obtained for Blocks. Stress-strain curves from the consolidated undrained triaxial tests and compressibility modulus (11m.) curves from consolidation tests for St. Simon Clay (LL = 69, PI = 44, SI = 10) and Nicholet Clay (U.. = 63, 26 PI = 40, SI = 15) are shown in Figs. 2.10 and 2.11 respectively. Figs. 2.10 and 2.11 show that for both the clays the undrained strength and compressibility modulus obtained on Shelby specimens are considerably lower than those obtained on Block specimens. Figs. 2.10 and 2.11 also demonstrate the effect of index properties of the two clays on stress-strain and compressibility characteristics for different methods of sampling. La Rochelle and Lefebvre (1971) reported that for sensitive Champlain Clay, undrained shear strengths measured on samples obtained by NaI 54 mm sampler (ARzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA = 10%, ICR = 1%) were 50 to 60% of the value measured on block samples. Milovic (1971b) also studied the effect of sampling methods on some loess properties. Loess samples (LL = 41, PI = 18) were obtained by Shelby tubes (AR = 12%, ICR = 0.8%) and also from blocks, cut by hand. The unconfined compression tests and consolidation tests were carried out on both types of specimens. The unconfined compressive strength, Young's modulus, compressibility modulus and preconsolidation pressure obtained on Shelby specimens were considerably higher than those obtained on Block specimens. This was attributed to higher initial density of Shelby specimens. It is well known that the initial density affects the elastic, shear and consolidation properties of loess. Eden (1971) investigated the effect of sampling method on preconsolidation pressure and undrained shear strength for sensitive overconsolidated clay. Sampling was conducted with four types of piston samplers, and the test results were compared with those obtained from block samples. Samplers used were the Swedish Foil (Kjellman et aI, 1950), the NaI 50 mm, the sal standard 50 mm, and the 127 mm Osterberg hydraulic sampler. The in-situ strength of the clay was also measured with the field vane test. The results showed that none of the samplers nor the field vane test were successful in obtaining results that could be compared consistently with results obtained from the block samples. The main conclusion of the study is that present methods of sampling of such clays by boring from the surface do not produce satisfactory undisturbed samples in this material. Raymond et al (1971) studied the behaviour of sensitiveLeda Clay sampled by six different sampling methods to assess the significance of the different features in the design of samplers. An example is shown in Fig. 2.12, demonstrating qualitatively the differences in stress-strain relationships of block and tube samples and the qualitative similarities between different tube samplers. Of the five different tube 27 samplers used, the samplers causing least disturbance were, in order: (a) the 125 mm Osterberg hydraulic piston sampler; (b) the SGI 50 mm standard piston sampler; (c) the 50 mm thin-walled Shelby tube piston sampler with sharp outside cutting edge; (d) the 50 mm thin-walled Shelby tube piston sampler with normal cutting edge; and; (e) the 50 mm thin-walled open-drive Shelby tube. The quality of samples of soft marine clay with respect to the method of boring and sampling was reported by Adachi et al (1981). The quality of soil samples was found to depend markedly on the method of boring and sampling. The average undrained shear strengths obtained by percussion borings with the open-drive sampler were almost one-half of those obtained by rotary borings with the fixed piston, thinwalled sampler. at (1985) compared the behaviour of block samples of Norwegian marine Lacasse etzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA clays with the behaviour of 95 mm tube samplers. The block samples were taken with the University of Sherbrooke cylindrical block sampler for soft sensitive clays (Lefebvre and Poulin, 1979). Using a series of rotating blades, this sampler carves out a block of soil, 300 mm dia. by 350 mm high, at the base of a mud-filled hole. On completion of the carving, blades fan out to slice through the base of the block and these blades support the sample as it is raised from the borehole. With this sampler block samples can be obtained at much greater depths than in an open trench. During sampling with the Sherbrooke sampler, the borehole is kept full of bentonite mud to reduce drastically the stress relief. In addition to allowing block sampling from the surface, the method provides samples of equivalent or better quality than conventional block samples (Lefebvre and Poulin, 1979). The tube samples were obtained with the NGI 95 mm fixed piston sampler (AR = 14%, ICR = 1.4%, outside cutting edge taper angle = Hf). Two quick clays of low plasticity and one sensitive clay of high plasticity were sampled. The laboratory test results were compared in terms of preconsolidation pressure, oedometer curves, and stressstrain-strength behaviour from unconfined compression, triaxial and direct simple shear tests. The quality of the block samples was superior to the quality of the samples obtained by 95 mm piston sampler. However, the degree of disturbance due to tube sampling varied for different types of clays. In case of lean quick clays, block sampling resulted in 30% higher undrained strength and 4 times higher Young's modulus. In case of the plastic sensitive clay, the block and 95 mm 28 samples had similar characteristics. Only small differences were observed in the preconsolidation stress profiles derived from tests on both types of samples. The effect of sampling disturbance on the test result also varied with the type of test. The disturbance effect appeared smaller in tests which offered large confinement. The effect of sampling disturbance was indeed the least in the oedometer test, intermediate in consolidated triaxial test and the largest in unconfined compression tests. The experience in the Norwegian clays demonstrated the ability of the University of Sherbrooke cylindrical block sampler to obtain samples of excellent quality, even at large depths ( >10 m). Dietzler et al (1988) compared the effects of sampling disturbance on shear strength between samples of glacial till (LL = 24, PI = 12) obtained by three methods; = 12%, ICR = 1%), namely, samples obtained with thin-walled tube sampler (ARzyxwvutsrqponmlkjihgfedcbaZYX samples carved from blocks, and samples obtained with continuous split-barrel sampler (AR = 89%, ICR = 14%). The comparative laboratory testing program indicated that the continuous sampler may be used to provide soil samples of cohesive fill which are equivalent to those obtained using thin-walled tubes. Test results also showed that the effective shear strength parameters determined using samples obtained by any of the three methods are accurate for use in feasibility or preliminary investigations, especially where the future loading conditions and testing pressures exceed the maximum past confinement. 2.7 SAMPLING EFFECTS The effects of sampling on stress-strain characteristics can be considered by dealing separately with the following: (a) "Perfect" sampling, which is usually simulated in the laboratory by consolidating specimens anisotropically in the triaxial apparatus and then releasing the in-situ shear stress under undrained conditions. Undrained shear to failure, therefore, starts from an isotropic stress state. "Perfect" sampling is a gross simplification of the total sampling but, nevertheless, has enabled to determine the important effects of sampling. (b) Imperfect sampling, in which some arbitrary stress path is assumed to be applied before undrained shearing to failure. Imperfect sampling has been further subdivided 29 into block sampling and tube sampling. (c) Ideal sampling (Baligh et aI, 1987), which can be modelled in the laboratory by consolidating specimens anisotropically in the triaxial apparatus and then imposing predicted tube penetration disturbances, followed by undrained stress relief simulating "perfect" sampling. Undrained shear to failure, therefore, starts from an isotropic state of stresses. 2.7.1 "PERFECT" SAMPLING The influence of "perfect" sampling on undrained stress-strain and strength properties of soils has been studied by numerous investigators (Skempton and Sowa, 1963; Ladd and Lambe, 1963; Ladd and Varallyay, 1965; Seed et al, 1964; Noorany and Seed, 1965; Davis and Poulos, 1967; Adams and Radakrishna, 1971; Kubba, 1981; Gens, 1982; Kirkpatrick and Khan, 1984; Jardine, 1985; Hight et al, 1985; Kirkpatrick et aI, 1986; Graham et al, 1987; Baligh et al, 1987; Graham and Lau, 1988; Hight and Burland, 1990). Skempton and Sowa (1963) examined the effect of "perfect" sampling in remoulded Weald Clay (LL = 46, PI = 24) which has a low sensitivity (S, = 2). Pairs of specimens of Weald Clay were normally consolidated in a triaxial cell under approximately Ko-condition. One specimen ("ground") of each pair was sheared by increasing the axial stress. The second specimen ("perfect") was first unloaded by reducing the axial stress to the value of lateral stress and then loaded by increasing the axial stress with both steps being done under undrained conditions. Fig. 2.13 shows the stress paths for a pair of specimens. Skempton and Sowa (1963) found that the undrained strength of the "perfect" samples were only 1 to 3% less than that of the "ground" samples although the stress paths were entirely different. They also found that failure strain of "perfect" samples were increased. Most likely, clays with higher sensitivity will be affected more by "perfect" sampling; since Noorany and Seed (1965) observed a 6% reduction of the strength for San Fransisco Bay Mud with a sensitivity of 8 to 10. Ladd and Varallyay (1965) found a 10% decrease in undrained strength for normally consolidated Boston Blue Clay (LLzyxwvutsrqponmlkjihgfedcba = 33, PI = 15) due to "perfect" sampling. Davis and Poulos (1967) reported a 18% decrease in strength of a remoulded "perfect" kaolin (LL = 55, PI = 22) specimen tested unconfined. However, the undrained strength of the reconsolidated "perfect" specimen 30 was only 5% less than that of the "field" element. in undrained consolidated strength Kubba (1981) reported a decrease of 5 to 11zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCB % due to "perfect" sampling for normally samples of kaolin. Ladd and Lambe (1963) determined pressure parameter, Au the isotropic effective stress, ape and pore of "perfect" specimens of Kawasaki Clay and Boston Blue Clay. The resulting values of the ratio, 0'Ja zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFE zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHG v were 0.56 ± 0.05 with corresponding A,. values of 0.17 ± 0.10. a/Ja v = 0.59 Clay yielded values of Values of Au Au Similar test data on normally consolidated Boston Blue and Au = 0.11. Ladd and VaraUyay (1965) also reported and 0'Jo'v for undisturbed Kawasaki Clay and Boston Blue Clay. 0'Ja v for "perfect" sampling obtained by different and the ratio investigators are summarised in Table 2.1. Apart from leading to a decrease in strength, "perfect" sampling has a marked influence on pore pressure responses as reported by Seed et al (1964), Noorany and Seed (1965), and Ladd and Varallyay (1965). failure was found to decrease "perfect" sampling. The pore pressure parameter A at by as much as 50% for specimens subjected to Ladd and Varallyay (1965) also observed a slight reduction in stiffness and a large increase in axial strain required to mobilise the peak shearing resistance. Atkinson and Kubba (1981) also reported considerably lower stiffness for anisotropically consolidated "perfect" specimens than that for the "in-situ" specimens. Jardine (1985) explained the effect of "perfect" sampling on stiffness anisotropy. a normally consolidated compression For soil, after "perfect" sampling, a sample becomes stiffer in than in extension. For a overconsolidated soil, however, a sample becomes stiffer in extension than in compression due to "perfect" sampling. Kirkpatrick and Khan (1984) investigated the influence of stress release caused by "perfect" sampling on the undrained stress-strain behaviour of normally consolidated kaolin (PI = 30) and illite (PI = 40). The tests on both clays showed that, compared to "in-situ" soil, "perfect" samples suffered considerable loss in strength, increase in failure strain, and produced appreciably different effective stress paths to failure. The strength losses were more acute in the less plastic kaolin compared with the more plastic and less permeable illite. Similar effects of stress relief due to "perfect" sampling were observed on lightly overconsolidated (Kirkpatrick et al, 1986). 31 (OCR = 2 to 3) kaolin and illite The effects of "perfect" sampling on low plasticity clays (LLzyxwvutsrqponmlkjihgfedcbaZYX = 32, PI = 17) have been discussed by Hight et al (1985). Fig. 2.14 contrasts the undrained behaviour of a young Ko-consolidated low plasticity clay from North Sea when sheared at two OCRs (= 1 and 7.4) from either "in-situ" conditions or those resulting from "perfect" sampling. Comparing the normally consolidated (OCR = 1) "in-situ" test RI from point A (Fig. 2.14) with the "perfect" sample test PSI, it is evident that "perfect" sampling greatly reduces the initial mean effective stresses, both through shearing during unloading and subsequent creep. Peak undrained strength and undrained brittleness are reduced by "perfect" sampling. The ultimate strength is little affected but the overall stress-strain behaviour is modified considerably. The changes in the stress-strain properties were also investigated in terms of two indices (E,.)O.Ol.JP'O and L l= (E..)o.l.J(E..)o.ol ..l as proposed by Jardine (1985). E, is the secant stiffness and p'o is the initial mean effective stress prior to shear. The first index provides a measure of small strain region. The small strain zone may be considered as a region around a point in stress space, within which strains accompanying stress changes from that point are less than some small limiting value, e.g., 0.1% (Baldi et al, 1988). The second index, L is an indicator of non-linearity in the stress-strain behaviour, the higher the value of L, the greater is the degree of linearity; L = 1 indicates a linear behaviour. Due to "perfect" sampling size of the small strain zone was increased while the degree of non-linearity was reduced. For the case of over consolidated soil (OCR = 7.4), in which Ko>I, the "perfect" sampling path is identical to the initial section of the triaxial compression path from "in-situ" conditions. There is, therefore, no change in strength. However, both the size of the small strain region and degree of non-linearity were reduced due to "perfect" sampling. It is apparent from Fig. 2.14 that the effective stress changes during "perfect" sampling are completely different from the two stress histories considered. The effect of stress history on the "perfect" sampling stress path and on the changes in effective stress was reported by Hight and Burland (1990) for the case of a low plasticity clay. This is shown in Fig. 2.15. It can be seen from Fig. 2.15 that the effective stress changes reduce as the OCR increases; for an OCR of 4, there is no change in effective stress; for the heavily overconsolidated clay, there is a slight increase in average effective stress. Therefore, the effect of "perfect" sampling on undrained triaxial compression strength decreases with increasing OCR as shown in Fig. 2.16. For overconsolidation ratios greater than 4 for this clay, there is no change in strength. At all oveconsolidation ratios, there is no effect of "perfect" sampling on triaxial strength 32 in extension, since the direction of the stress path is not reversed. The effect of soil composition on "perfect" sampling stress path and undrained strength after "perfect" sampling was also reponed by Hight and Burland (1990). For heavily overconsolidated soils , the soil composition has little effect. For normally consolidated soils, the effect of soil composition on "perfect" sampling stress path is shown in Fig. 2.17. reduces. The reduction in effective stress increases as soil plasticity Accordingly, the effect of "perfect" sampling on triaxial compression strength also increases as the soil plasticity reduces. Kirkpatrick and Khan (1984) also reponed similar observations. The effect of soil composition on undrained triaxial compression strength after "perfect" sampling is shown in Fig. 2.18. The effect of "perfect" sampling disturbance on overconsolidated (OCRzyxwvutsrqponmlkjihgfedc = 2.5) plastic Drammen Clay (PI = 27) has beenreported by Lacasse and Berre (1988). "Perfect" samples were loaded to the same stresses as they were carried after the laboratory overconsolidation for the reference undisturbed specimen. They reported about 11% decrease and 2% increase in undrained shear resistance in compression and extension respectively. "Perfect" samples, however, when consolidated to maximum vertical stress of the undisturbed specimen and then unloaded to the appropriate OCR (= 2.5) provided 3% and 12% increase in shear resistance in compression and extension respectivel y. The effect of "perfect" sampling on one-dimensional compressibility characteristics of a soft saturated silty clay (LL = 88, PI = 45, S, = 10) from the San Francisco Bay area was investigated by Noorany and Poormand (1973). The effect of "perfect" sampling on compressibility for this clay is shown in Fig. 2.19 which illustrates the variations of void ratio with vertical effective stress. It is seen that mere removal of the "in-situ" stresses without physical disturbance has practically no influence on the compressibility characteristics of the clay. Also, the preconsolidation pressure and the virgin compression behaviour of the "perfect" sample closely represent the "insitu" behaviour. This shows that, in the absence of large shear strains, major changes in the effective stress system will not adversely affect the compressibility behaviour of the clay. Davis and Poulos (1967) also found slight reduction in coefficient of consolidation, Cy for "perfect" specimen of kaolin. 33 2.7.2 BLOCK SAMPLING Block sampling can be modelled in the laboratory by releasing and trimming blocks of soil from large oedometer samples. Hight et al (1985) demonstrated the behaviour of specimens of Lower Cromer Till, another low plasticity clay, due to block sampling. The results of unconsolidated undrained triaxial compression tests are presented in Fig. 2.20. The specimens were cut from the blocks having different stress histories (OCRs of 1, 2, 4, 7 and 80). It can be seen that the effect of block sampling largely obliterates the important effects of stress history on in-situ behaviour. The specimens tend towards similar initial mean effective stress levels and, as a consequence, show similar behaviour. As could have been anticipated from the results of "perfect" sampling, peak strengths and undrained brittleness are reduced in the normally and lightly overconsolidated soil. The effect of block sampling on the stress-strain behaviour was also assessed by Hight et al (1985). They reported results from two similar unconsolidated undrained tests on specimens of North Sea clay cut from reconstituted blocks (OCRzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJI = 2). Both the initial stiffness and degree of non-linearity were reduced. The quality of block samples has been compared with that of tube samples by several workers. A number of examples have been reported previously in section 2.6. It has been found that block samples are always less disturbed than tube samples, and as a consequence provide higher undrained strengths (Milovic, 1971a; La Rochelle and Lefebvre, 1971; Raymond et al, 1971; Lacasse et aI, 1985). 2.7.3 TUBE SAMPLING In contrast to "perfect" sampling, the stress or strain paths involved in tube sampling from a borehole, and subsequent extrusion in the laboratory, are complex. The levels of distortion which occur as soil enters an idealised sampler have been determined by the Strain Path Method of analysis (Baligh, 1985). The Strain Path method has been discussed in section 2.8. Fig. 2.21 illustrates the sequence of strains experienced by an element of soil on the centreline of samplers of differing geometry. The element undergoes substantial compression strains and then extension strains as it enters the sampling tube. Further straining occurs when the sample is extruded from the tube prior to testing. 34 Around the periphery, severe shear distortions may be superimposed on this strain path (Fig. 2.21). For the element on the centreline, the corresponding stress paths during tube sampling and extrusion has been predicted by Hight (1986) on the basis of behaviour pattern shown in Fig. 2.21. The predicted stress paths for normally consolidated and overconsolidated soil are shown in Figs. 2.22 and 2.23 respectively. Hight (1986) pointed out the following: (i) in the normally consolidated soil, the effective stresses are reduced, (ii) in the heavily overconsolidated soil, the effective stresses are increased, (iii) changes in pore pressure are different on the centreline and around the periphery so that a process of equalisation takes place. The level of distortion which occurs around the periphery of tube samples is often apparent when such a sample is split to expose its fabric. Although the strain paths followed in this outer zone have not been modelled in triaxial tests,zyxwvutsrqponmlkjih it can be reasonably anticipated that: (a) soil in an initially normally consolidated or lightly overconsolidated state will develop positive pore pressure increments (Fig. 2.22), (b) soil in a heavily overconsolidated state will develop negative pore pressure increments (Fig. 2.23). Extrusion involves additional distortion. Its path is indicated arbitrarily by efgh in Figs. 2.22 and 2.23. The response that could be anticipated in normally consolidated soil after tube sampling and extrusion has been shown by Hight et al (1987) for young low to medium plasticity clays. The response, shown in Fig. 2.24, is compared to the "in-situ" soil, and soil after "perfect" sampling. It can be seen that the undrained stress path and stress-strain curve of tube sample are markedly different from those of "perfect" and "in-situ" samples. Hight et al (1985) also reported the behaviour of three tube samples taken from the seabed in the North Sea. The estimated OCR's of the ftrst two samples were 1.1 and the OCR of the third sample was greater than 50. The initial mean effective stresses of the normally consolidated samples were below those estimated in situ, but the heavily overconsolidated sample showed a large overall increase in initial mean effective stress. Of the two normally consolidated samples, one was tested without 35 consolidation and the other was reconsolidated situ stresses. The overconsolidated anisotropically to the estimated in- sample was tested without consolidation. of the three intact tests provided a satisfactory model for the in-situ behaviour. the normally consolidated samples corresponding test on reconstituted None Both values of (E,.)o.o,.!P'o than the sample and only the reconsolidated normally gave higher consolidated sample reproduced strong non-linearity. Both the normally consolidated samples showed behaviour after 0.1zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA % strain that was markedly different from that expected in situ. differences For the heavily overconsolidated sample, however, despite the in initial stress path direction and the stiffnesses around 0.1 % strain, reasonable agreement with in-situ behaviour was found. Apart from stress-strain behaviour, tube sampling also affects the compressibility characteristics The effect of tube sampling disturbance of clays. on consolidation parameters of soft clay samples (bulk density =zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA 1S.7 to 16.8 kN/m3) was examined by Bromham (1971). Samples were obtained in 76 mm internal dia. thin walled sampling tubes. The major effect of sampling disturbance was to produce low values of coefficient of volume compressibility, 11\., especially near the overburden pressure. Values of I1\. calculated from the reconstructed field curve were considerably higher than those obtained directly from the laboratory consolidation consolidation, Cy test. Coefficient of for the least disturbed specimens! with disturbance factor of 15 to 20, were less than the extrapolated field values by a factor of 2 to 5. Lacasse et al (1985) carried out comparative oedometer tests on blocks taken with a Sherbrooke sampler and 95 mm dia. NO! piston samples for three sensitive clays. of the quick clays (S, > 60), the volumetric strain during reconsolidation In the case to the in- situ effective stresses (which is a measure of sample disturbance) were found to be higher for the 95 mm samples than that for the block samples. Hight et al (1987) also reported higher volumetric strains for tube samples of lightly overconsolidated Magnus Clay (OCR = 1.15) than those for reconstituted For both the tube and block samples m-values "in-situ" sample. block samples (OCR = 2). were considerably smaller than the Compression indices, Co were, however, the same for block, tube, and "in-situ" samples. 36 2.7.4 IDEAL SAMPLING Baligh et al (1987) proposed ideal sampling approach (lSA) as an extension to "perfect" sampling. Ideal sampling approach denotes an idealised method of incorporating the effects of tube penetration, sample retrieval to the surface and extrusion from the tube, but neglects all other types of disturbances, including operator dependent disturbances and water content changes in the soil. The proposed method for implementing ISA consists of the following steps: (a) Estimation of tube penetration disturbances at the centreline of sampler using the Strain Path Method. (b) Estimating the effects of sample retrieval and extrusion by assuming an idealised process of undrained stress relief from the (generally) anisotropic stress conditions in the tube to the final isotropic stress state of the sample before testing. Step (b) adopts the same simplification adopted by "perfect" sampling regarding sample retrieval and extrusion simulation. Therefore, the only difference between the proposed ISA and "perfect" sampling is the incorporation of tube penetration disturbances, i.e., step (a), and hence ISA is equivalent to "perfect" sampling when tube penetration disturbances are insignificant. This condition can be achieved by block sampling, i.e., by eliminating tube penetration effects. A limited number of tests were carried out by Baligh et al (1987) on reconstituted samples of Boston Blue Clay (LL = 42, PI = 20zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA ± 2.5) to evaluate the effects of ideal sampling disturbances and tube penetration disturbances on the undrained stress-strain, stiffness and strength behaviour of normally consolidated clays under Ko-conditions. summarised in Table 2.2. The results are In test 1 (Table 2.2), the sample was subjected to monotonic undrained shearing to determine the reference "undisturbed" normally consolidated behaviour of the soil before disturbance. In test 2, the soil was subjected to a simulated disturbance of ideal sampling in order to determine their effects on undrained behaviour. The tube penetration disturbances applied in test 2 corresponds to that obtained along the centreline of a Simpler sampler with aspect BIt = 40 (a value typical of well-designed thin-walled piston samplers) and IeR ratio,zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA ... 1%. In test 3, the soil was subjected to the same tube penetration disturbances as in test 2, but without the undrained stress relief simulating sample retrieval and extrusion in order to isolate tube penetration disturbances and determine their relative importance. Fig. 2.25 shows the stress-strain curves and stress paths after the 37 application of ideal sampling disturbances 2 and test 3 respectively. and tube penetration disturbances Fig. 2.25 also shows the "undisturbed" in test behaviour of the sample in testzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA 1. Examination of the data presented in Table 2.2 and the stressstrain behaviour of the samples shown in Fig. 2.25 indicates the following: (a) The mean effective stresses in the sample prior to shear is reduced by about 57% due to ideal sampling disturbances, compared with 59% for tube penetration disturbances. (b) Ideal sampling disturbances reduce undrained shear strength ratio (cJa'yJ by about 18% while tube penetration disturbances overpredict (c) Ideal sampling disturbances 4.42%. this value by only 2%. increase the strain at peak strength from 0.16% to Tube penetration disturbances underestimate' this value by only 1.6%. (d) Ideal sampling disturbances reduce the undrained stiffness ratio (EsJa'yJ by about 75%. Tube penetration disturbances overestimate this value by about 27%. (e) Test 2 and test 3 give virtually identical stress paths and identical stress-strain behaviour at strain levels exceeding 0.5% (Fig. 2.25) although the stress paths and stress-strains curves are entirely different from the "undisturbed" sample in test 1. The relative importance of tube penetration disturbances versus sample retrieval and extrusion disturbances, as incorporated in the ideal sampling approach can be evaluated by comparing the results of tests 2 and 3. From the points, as indicated above, it is evident that, with the exception of soil stiffness at small strain level, the effects of ideal sampling disturbances and tube penetration disturbances on subsequent stress-strain and strength properties are basically the same. Baligh et al (1987), therefore, concluded that sampling disturbances predicted by ideal sampling approach are primarily due to tube penetration effects rather than to sample retrieval and extrusion effects when the latter are simulated by an idealised process of undrained shear stress relief. The effect of tube penetration disturbance compression on undrained stress-strain behaviour in and extension was also reported by Lacasse and Berre (1988) for both normally consolidated (OCR = 1) and overconsolidated (OCR = 2.5) plastic Drammen Clay (PI = 27). Before shearing under undrained condition specimens were disturbed by imposing strain paths equivalent to that estimated at the centreline of a Simple samplerzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA (BIt = 40, ICR .. 1%). The stress-strain characteristics of the disturbed andzyxwvutsrqpon 38 undisturbed specimens are shown in Fig. 2.26. The following preliminary conclusions were reported: (i) Compression tests (OCRzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCB = 1 and 2.S) - The peak shear strength is about the same for the disturbed and undisturbed specimens while the initial moduli are much lower for the disturbed specimens. For normally consolidated disturbed specimens, however, the shear resistance at high strains is much higher. (ii) Extension tests (OCR = 1 and 2.S) - The ultimate shear strengths and the initial moduli are higher for the disturbed than for the undisturbed specimens. Baligh et al (1987) and Lacasse and Berre (1988) reported tests carried out to model tube penetration disturbances estimated only at the centreline of a Simple samplerzyxwvutsrqponm (BIt = 40, ICR = 1%). More tests should be carried out for other clays to validate and confirm their findings. Results could then be compared with those reported by Baligh et al (1987) and, Lacasse and Berre (1988). Moreover, the effect of varying degrees of tube penetration disturbances on the stress-strain-strength characteristics has not yet been investigated. Such an investigation is essential because it might reveal the relative importance of cutting shoe designs in controlling the degree of disturbance. 2.7.5 METHODS FOR CORRECTING SAMPLING DISTURBANCE EFFECTS Because of sampling disturbances, it is necessary to correct the undrained strength in order that it is representative of the in-situ material. A number of methods have been proposed for correcting the strength and these are presented in the following sections. Some of the methods involve the use of the void ratio of the soil in its natural state. Hvorslev (1949) reported a relationship between undrained strength and void ratio. Correct strength can be obtained by extrapolating the logarithm of strength versus void ratio line to the in-situ void ratio. Schmertmann (1956) and Calhoon (1956) also used a correction utilising the in-situ void ratio. Ladd and Lambe (1963) considered the difference between measured residual effective stress, cif and the residual effective stress expected with "perfect" sampling, 39 a'zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA po as being similar to an overconsolidation phenomenon which influences the measured strength. For each particular soil they established a relationship between the overconsolidation ratio, OCR and shear strength. Then by considering the OCR as being equal to cr'Ja'r, they corrected the strength measured at an effective stress of o', to the value that would have existed if the sample had been tested at a stress of a' po' An alternative correction method, proposed by Adams and Radakrishna (1971), is based on the loss of suction (the difference between tIle measured suction,zyxwvutsrqpo -e', and theoretical suction, -a' po)' The loss of suction is proportional to the change in moisture content. The change in moisture content associated with the computed loss of suction is determined from a swelling curve. The swelling curve is obtained by allowing a sample to return to its initial suction with drainage permitted after being unloaded (undrained) from the K, or "in-situ" stress condition by removing the deviator stress. The correction for strength loss is then determined from a unique moisture content-shear strength relationship using the measured failure moisture content and the computed moisture content change due to loss of suction.zyxwvutsrqp -' Okumura (1971) proposed a method similar to Ladd and Lambe (1963) to correct for a disturbed strength. In order to obtain the base for correction, a triaxial compression test, loaded repeatedly up to failure, is performed on a representative specimen consolidated under Ko-conditions and with its deviator stress released in an undrained condition ("perfect" sample). Test results are plotted as disturbed strength ratio (S..JSu~ against disturbance ratio (a' /a'r), where Suris the undrained strength after each cycle, Supis the undrained strength of the "perfect" sample, a', is the residual effective stress after each cycle and, cr'pis the residual effective stress of the "perfect" sample. Such a plotting is shown in Fig. 2.27. Fig. 2.27 also presents the results from repeated loading simple shear tests plotted as disturbed strength ratio and disturbance ratio. Fig. 2.27 shows that all the test results, as a whole, lie on a unique curve with relatively little scatter for various kinds of test. In order to find out the undisturbed strength of a sample, the residual effective stress of the actual sample is first measured to find its disturbance ratio. The, sample is then sheared to find its disturbed strength. The correction curve (Fig. 2.27) obtained by the above process gives the perfectly undisturbed strength of each sample. A comprehensive way of correcting the measured value of the undrained shear strength for sample disturbance has been reported by Nakase et a1 (1985). 40 An expression has been proposed to evaluate the disturbance ratio (a ratio of the undrained strength of the "perfect" sample to the undrained strength of the actual sample) of a soil sample from the measured values of plasticity index, secant modulus, Eso and the in-situ effective overburden pressure. The disturbance ratio could then be used to correct the measured undrained strength value. The proposed method of correction is applicable to soils of wide range of plasticity.zyxwvutsrqponmlkjihgf It is possible to reduce the effects of sampling disturbance on the undrained behaviour of clays by reconsolidating the sample to a more appropriate stress level prior to shearing. Raymond et al (1971) reported that the optimum consolidation pressure to be used depends on sample disturbance. For general use where an extensive set of tests is not undertaken, an isotropic consolidation to 50-75% of the preconsolidation has been recommended. Kirkpatrick and Khan (1984) adopted two methods of isotropic reconsolidation to examine whether the "in-situ" undrained behaviour could be reproduced. Hydrostatic reconsolidations to pressures equal to 0' pi and "in-situ" vertical effective pressure, o'YC were applied to samples of kaolin and illite. It was found that, reconsolidation to 0' pi resulted in underestimation of "in-situ" strength by as much as 14%. However, hydrostatic reconsolidation to a'yC had the effect of producing fairly large overestimations of "in-situ" strength of 16% or more. Failure strains and porewater pressures were heavily overestimated by both the methods of reconsolidation. Graham et al (1987) found that in both normally consolidated and overconsolidated samples of kaolin, isotropic reconsolidation to a'yC overestimated the strength of "insitu" specimens while isotropic reconsolidation to 0.60'yo underestimated it. In both cases the strains to failure and pore pressure parameter at failure were higher than the "in-situ" specimens. These findings agree with those reported by Kirkpatrick and Khan (1984). Similar results have also been reported by Graham and Lau (1988) for normally consolidated kaolin. Anisotropic reconsolidation has been proposed by several investigators as an effective method of reducing sampling disturbance effects. Ko-consolidation to the in-situ stresses has been suggested by Davis and Poulos (1967) and Bjerrum (1973). 41 Ladd and Foott (1974). proposed that samples should be reconsolidated anisotropically to a pressure at least equal to 1.5 to 2 times the in-situ vertical effective stress,zyxwvutsrqponmlkjihgf 0',,,. This method of reconsolidation is called the SHANSEP (Stress History and Normalised Soil Engineering Properties) method. The effect of anisotropic reconsolidation in recovering the in-situ behaviour has been studied by many reconsolidation strength research workers. La Rochelle et al (1976) reported that of the samples to the in-situ stresses restored at least part of the and stiffness lost by sampling negligible in case of good quality samples. disturbance. However, this effect was Kirkpatrick and Khan (1984) found that compared with the "in-situ" soil, anisotropic reconsolidation to "in-situ" stresses gave a good simulation of strength and stress-strain behaviour and closely similar stress paths. Graham and Lau (1988) also obtained significantly better results for samples reconsolidated to "in-situ" isotropically. Anisotropic stresses than for samples reconsolidation those were consolidated to "in-situ" stresses produced overall estimate of strength, pore water pressure parameters and stiffness. the best Atkinson and Kubba (1981), however, found considerably lower stiffness for anisotropically reconsolidated normalised specimens effective than that for the "in-situ" specimens, although the stress paths were the same for both "perfect" and "in-situ" specimens. Hight et al (1985) found consolidated from a series of experiments that in the normally soil the features of the in-situ behaviour were not fully recovered by anisotropic reconsolidation of samples to the in-situ stresses. in stiffness and post peak behaviour. There were differences The effects of a sampling cycle were only fully removed when reconsolidation was continued to vertical effective stress levels greater than 1.75 times the previous maximum vertical stress. This finding is consistent with that on which the SHANSEP approach (Ladd and Foott, 1974) is based. Gens (1982) from his investigation on low plasticity clays reported that anisotropic reconsolidation to approximately 1.8 times the previous maximum vertical stress was required before the effects of sampling and preparation were eliminated. Baligh et al (1987) also reported that most effects of ideal sampling disturbance on the undrained behaviour of Ko-normally consolidated Boston Blue Clay could be reduced by reconsolidating the soil and could, in effect, be virtually eliminated by the SHANSEP method. 42 2.8 STRAIN PATH METHOD FOR PREDICTING SOIL DISTURBANCE The Strain Path Method (Baligh, 1975 and 1985) is an approximate technique to predict soil disturbances objects in the ground. caused by the installation analytical of various rigid The method provides a framework that enables such problems to be approached in a realistic, systematic and rational manner. The Strain Path Method is based on concepts similar to the Stress Path Method (Lambe, 1967). 2.8.1 COMPARISON OF THE STRAIN AND STRESS PATH METHODS Table 2.3 describes the essential features of the Stress and Strain Path Methods and emphasises their strong similarities in approaching geotechnical problems. The Stress Path Method as described by Lambe (1967), is an approximate analytic technique for predicting the stability and deformation of shallow foundations, e.g., footings, mats, excavations, natural slopes, earth dams and situations where the depth of the soil of interest below the ground surface is relatively small compared with its lateral extent. The Stress Path method is approximate because even under ideal conditions, using an infinite number of samples, the compatibility of strains is not satisfied. compatible strain field would be obtained if, and only if, the estimated increments were identical to those actually experienced A stress in the field. The latter depends on the soil behaviour and cannot, therefore, be known a priori. The Strain Path Method is also approximate because the estimated stresses will not in general, satisfy the equilibrium requirements, the actual one. performance unless the estimated strain field is identical to The Stress Path Method has proved successful in predicting of surface structures, slopes, earth dams etc. e.g., excavations, shallow foundations, the natural (Lambe and Marr, 1979). The Stress Path Method has also been proved reliable in analysing settlement problems (Simons and Sam, 1970). In concept, the Strain Path Method is virtually identical to the Stress Path Method except for one fundamental shallow problems versus versus aspect that really represents deep problems, namely the strain-controlled the difference the stress-controlled nature nature of deep problems of shallow (which represents the most rigorous definition of shallow versus deep problems). between in fact, Clearly the basic simplification introduced by the Strain Path Method consists of hypothesising 43 that estimates properties. of strain (instead of stress) increments For deep penetration problems is based on simple soil it is argued that this will introduce reasonably small errors that may be tolerated in view of other major uncertainties in soil behaviour. 2.8.2zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA APPLICATIONS OF STRAIN PATH METHOD The Strain Path Method has been applied to predict soil deformations, strains, stresses and pore pressures during deep penetration of cones, open and closed ended piles and samplers (Levadoux and Baligh, 1980; Chin and Baligh, 1983; Baligh, 1985; Chin, 1986). Deep steady penetrations state problems. of cones, piles and samplers are all axisymmetric steady Steady state means that, to an observer moving with the indenter (cone, pile or sampler), the deformations and strains in the soil do not vary with time. and viscosity of the soil, the process of By neglecting the compressibility penetration has been reduced to a flow problem where soil particles move along streamlines around a fixed body. A streamline is defined as a continuous line drawn in the direction of velocity vector at each point in the flow. consists of obtaining the deformations, elements along different superimposing streamlines. strains, and pore pressures at various soil Distortions the stream functions corresponding sinks on to that of a uniform flow. A solution, therefore, and strains to a combination are obtained by of sources and Stream functions serve as means of establishing streamlines of the flow and remain constant along a streamline. Using the Strain Path Method, the distortions caused by a blunt 6fP cone and a sharp 18° cone were obtained by Levadoux and Baligh (1980). The predicted deformation patterns around the cones are shown in Fig. 2.28. The deformed grids illustrate the magnitude and distribution of the shear strains. The deformed grids show that the sharp (18~ cone cuts its way through the soil and causes smaller shear strains than the blunt (6fP) cone which causes severe straining in the vicinity of the tip and near the shaft. The analysis assumes a frictionless believed to underestimate Deformations, soil-probe interface and, hence, is actual soil distortions. strains and strain rates in the soil due to closed and open-ended pile 44 penetration have been investigated by Chin and Baligh (1983). (BIt = 20 to 40) it has been found that large deformations pileszyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA For typical offshore and strains in the vicinity of closed-ended pile walls are controlled by the radiuszyxwvutsrqponmlkjihgfedcbaZYXWVUT (B/2) of the pile. For open-ended piles, however, deformations and strains near the pile wall are basically controlled by the wall thickness (t), This means that for a given wall thickness, the ratio BIt (or the diameter) has minor effects on deformations and strains in the soil near the pile wall. The general trend observed from the analyses is that the magnitudes of deformations and strains tend to decrease as BIt increases. The soil which moves outside the pile is more heavily strained and more affected by aspect ratio (BIt) than that moving inside the pile. Predictions in the inner soil of open- ended pile penetration can also be used to estimate sampling disturbance effects due to sampler intrusion. Due to lack of reliable experimental results on pile penetration effects, predictions can not be evaluated accurately. However, qualitative comparisons with experimental results obtained by Randolph et al (1979) for closed-ended penetration and the study of sampling distortions reported by Hvorslev open-drive samplers indicate that the predicted deformations pile 0949) for are reasonable. Distortions resulting from deep penetration of samplers have also been obtained by Baligh (1985) using the Strain Path Method analysis. treated as an incompressible In the analysis soil has been and inviscid fluid flowing past the sampler. A Simple sampler with round-end wall and a sampler with flat-ended wall were investigated. The samplers were modelled using the method of sources and sinks of Potential Theory. The predicted deformation pattern around the sampler with flat-ended wall and for a Simple sampler having the same external diameter to wall thickness ratio, BItzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA = 20, are shown in Figs. 2.29(a) and 2.29(b) respectively. The Simple sampler solution corresponds to the superposition of one single ring source and a uniform velocity field. The flat-ended sampler is simulated by superimposing 42 ring sources of which 10 sources are located directly distributed along each side of the sampler. behind the flat end and 16 sources Fig. 2.29 shows the following: (a) the effect of cutting shoe geometry on soil distortions vicinity of the sampler walls (with a region of width (b) deformations := is only visible in the 3t); of the soil penetrating into the sampler (inner soil) are different from the soil outside; (c) visual inspection of Fig. 2.29 reveals 45 no soil distortion near the sampler centreline. The strain history of an element at the centreline of Simple samplers having B/t ratio (i.e., aspect ratio) equal to 10, 20 and 40 has already been presented in Fig. 2.21. Fig. 2.21 shows that the sample straining (disturbance) depends on the aspect ratio, B/t, of the sampler. From the strain contours during undrained Simple sampler (B/t = 40, IeR "'" 1%) zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA penetration in saturated clays, Baligh et al (1987) also conclude that reasonable estimates of soil disturbances within the inner half of the tube can be obtained from the results at the sampler centreline. Chin (1986) shows that, for thin-walled Simple samplers (B/1»I), both the maximum axial strain in compression and extension at the centreline of the sampler is approximately given by the following expression: .... (2.14) Analyses conducted on samplers with flat-ended walls (Fig. 2.29(a» also indicated no significant effect of the sampler geometry on the strain history at the centreline of the sampler. According to Baligh (1985), the aspect ratio, B/t, of the sampler is the prime variable that controls the strain history of soil elements at the centreline of the sampler. But the exact geometry of thin-walled samplers has not been analysed by Baligh (1985). There is, however, strong evidence that the precise geometry of the cutting edge of a sampler is an important factor that controls the quality of sampling (Hvorslev, 1949; Kallstenius, 1958; La Rochelle, 1973; La Rochelle et al, 1981). The geometry of the cutting shoe of a sampler has a significant influence on the degree of disturbance of the soil around the sampler during sampling and subsequent timedependent equilibration of pore pressures after sampling lead to changes in effective stress on the centreline of the sample. This strongly suggests the need for research to be carried out to predict deformation and strains accurately for samplers having different cutting shoe designs. Attempts should also be made to predict strain histories at various locations within the sampling tube in order to assess the dependence of disturbance across the diameter of the sampler tube. 46 Au-values and stress ratioszyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONM (0'p/o'y) for "perfect" sampling Table 2.1 of normally consolidated clays Clay Index type properties Undisturbed LL = 48-106% Kawasaki PI == 16-46% .07-.28 0.47 0'p/o'y Reference .50-.61 Ladd and Lambe Clay (1963) Ladd and Varallyay (1965) Undisturbed LL = 33% Boston Blue PI = 0.54zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA .11 Ladd and .59 14% Lambe Clay (1963) Remoulded Boston LL = 33% Blue Clay PI = .12-.24 0.54 .57-.67 Varallyay 15% SI = 7 ± Ladd and (1965) 2 Skempton Remoulded Weald LL = 46% Clay PI = 24% and Sowa SI = 2 (1963) 0.59 -.02 to-.l .16-.24 .57-.61 .58-.62 Seed et Undisturbed LL = 88% San Francisco PI = 45% al (1964) Bay Mud SI = 10 Noorany 0.50 and Seed (1965) Note: 0'pi = isotropic effect stress in the "perfect" samplezyxwvutsrqponmlkjihgfedcbaZYX o', = vertical effective "in-situ" stress SI = sensitivity 47 Table 2.2 Effects of ideal sampling disturbance and tube penetration disturbances on undrained behaviour of Ko-normally consolidated resedimented Boston Blue Clay. (after Baligh et al, 1987) UNDRAINED DISTUR- BANCE SIMULATION Test UNDRAINED Descr- Tube Sample Effective Effective iption/ pene- retr- stresses stresses ieval after prior to and distur- shear extru- bance effect tration SHEAR BEHAVIOUR sionzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA a 'Jaye' ab'Jay.' a ;; ayeI'zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQP ab'j ayeI' cja:" Ep(%) EsJa:" 1 Undisturbed 2 Ideal No No .654 .481 .391· .16· 350- Yes Yes .278 1.0 .278 1.0 .263 4.42 88 Yes No .267 .426 .267 .426 .253 4.35 17 sampling 3 Tube penetration Note: • Average of six tests a: zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA = initial vertical effective "in-situ" stress at the end of Ko-consolidation a: = mean effective stress after disturbance a: = mean effective stress prior to shear a: = horizontal effective stress after disturbance v: and a:' = horizontal and vertical effective stresses prior to shear c c Cu = undrained compression strength e, = strain at peak strength Eso = undrained secant modulus 48 Table 2.3 Comparison of Stress Path and Strain Path Methods (after Baligh, 1985) STRAIN PATH METHOD STRESS PATII METIIOD APPLICA TIONS Shallow Problems: Depth of soil of Deep Problems: Soil of interest is interest is relatively small relatively deep below ground compared to its lateral extent. surface compared to its lateral extent. STEPS 1. Estimate initial stresses. 1. Estimate initial stresses.zyxwvutsrqponmlkjihgfedcbaZ 2. Estimate incremental stresses. 2. Estimate incremental strains. 3. Perform stress path tests on 3. Perform strain path tests on samples (or use adequate soil samples (or use adequate soil model) to obtain strains at model) to obtain deviatoric selected locations. stresses at selected locations. 4. Estimate deformations by 4. Estimate octahedral (isotropic) stresses by integrating strains integrating equilibrium equations. APPROXIMATIONS In step 2, stresses are In step 2, strains are approximate thus leading to approximate thus leading to strains not satisfying stresses not satisfying all compatibility requirements, equilibrium conditions, i.e., i.e., deformations octahedral stresses in step 4 in step 4 depend on strain integration depend on equilibrium path. integration path. 49 .. tzyxwvutsrqponmlkjihgfedcbaZYXW / BloCk zyxwvutsrqponmlkjihgfedcbaZYXWV IIlm ... d to 2.5 in. _zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPO t. 4. A zyxwvutsrqponmlkjihgfedcbaZYXWV a .----~~.~.----.----.~.----.-.. ..--.----.~~~ .. ....zyxwvutsrqponmlkjihgfedcbaZYXW x l- o I!) Z .. IIJ II: ----.-.. • • '--e--_ :;; ZO ..J "X:5 trimmed -..... to 2.5 in. , /~ II: l- ... 15 Q 2.8 In. _ carl' trimm.d to 2.5 in. II Z :( II: Q Z :> 10 \ 5LI------~---L--~J-~~-LL-----~--~~-L~--'~~~------~---L--~~500 TIME (DAYSI Fig. 2.1 Effect of storage time on undrained triaxial compression strength (after Arman and McManis, 1976) .. .. i:. .---.~.---.---.~.----.~.---.. .. /BIOCk" trimmed 10 2.5 A ...a: :> Cl) Cl) ... II: Q. Z o ~ o :::i o Cl)3 Z o ...a: u Q. z~-L~L5 ~~~-L-L~~~------~~~~ 10 50 100 150 LOG TIME (DAYS) Fig. 2.2 Effect of storage time on preconsolidation (after Arman and McManis, 1976) 50 pressure 0 0'4 O'S 1·2zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA u,zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA ""'c Q. 0 1·6 0 2·0 0 l- :z :z S! U1 :::l e:: lw x e:: w u 2·4 0 I- 2'8 ~ 0 .J UJ CD J: l- 3'2 3·6 4'0 BEFORE EXTRUSION Q. UJ 0 4'4 0 AFTER 4'S EXTRUSION 5·2 5·6 126 126 BULK 134 132 136 3 DENSITY, '( , lb./ ft. Fig. 2.3 Variation of bulk density in field sample before and after extrusion (after Shackel, 1971) " laboratory " _/ -. -. undisturbed Degree of disturbance. -~'(~.) .."" P,essure in '7. (108 scale) Fig. 2.4 Method of determining degree of disturbance (after Schmertmann, 1955) 51 R. ~ ~I\ lOOzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA ~ t-.... 90 I- ~ ~ ~ '" r-.... f'-. ~ f\ \,~ \\. ~~ ~ "J ~ ~ (~~ 80 t"o -?~): ~ zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPO zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA C'".... ('\c.... zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA ~ ...... ....)' 0" 1 2 < '~ "~. ' 70 "" 0: o (5 1\ r..... > 60 ~ ~ ,\l\ ~ 0 5zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA -, 42\ 40 0.06 0.1 0.2 0.4 Q6 PRESSURE. 1.0 2 ~ "0 4 6 10 20 (ton/ft~) Fig. 2.5 Definition of disturbance factor (after Bromham, 1971) Fig. 2.6 Dimensions of a tube sampler 52 internal guide internal guide izyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA COUPE-AA (mm)zyxwvutsrqponmlkjihgfedc Fig. 2.7 Drawing of the sampling and coring tubes of the Laval sampler (after La Rochelle et al, 1981) 53 (164-2S-3C) from 5" 1124 mm) die sampler ( ........ o-.Q....! zyxwvutsrqponmlkjihgfedcbaZYXWVUTS o/ 0-0'""0-0 !,_,~.--._.----.. o-°-2_g_o zyxwvutsrqponmlkjihgfedcbaZYXWVUT ~(" o \,( ~ • 164-2-5) from -a-o_o- 2" (54mm) die sampler zyxwvutsrqponmlkjihgfedcbaZYXWVU 0.6 I~l£ ~ Cl ~- ~ <I e! .. Cl- Cl .. 0.4 :; zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA Cl- o~--~----~----~--~----J---~ o 4 5 6 Axial Strain, Cl' % Fig. 2.8 Effect of sampling method on the measured strength, strain and pore pressure parameter A in soft marine clay (after Bozozuk, 1971) COMPLETELY UNOIS TlJR8EO SOIl. S' ANO 3" TUBES Z' TUBES DEGREE OF mSTURBANCE INCREASING Fig. 2.9 Influence of sample disturbance on undrained shear strength (after Conlon and Isaacs, 1971) 54 0.7 0.6zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPO --_ 0.5 N E ....u .... Cl "W'" zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA I w: I/) I/) L>.I a: l- I/) ST-SIMON a: 0 I- < 5 L>.I Cl BLOCK 2 0.1 CLAY NICOLET CLAY PISTON bc 2 0.3kg/cm2 AXIAL 3.0 2.0 1.0 E •. STRAIN. percent Fig. 2.10 Stress-strain curves from the consolidated undrained triaxial tests (after Milovic, 1971a) 60 ST-SIMON N E CLAY I 50 ....u NICOLET 2 ... CLAY Cl :ox:: 40 <, ai ::I 2 ...I ::I Cl 0 <, "- ::!; >I:::i 20 CD Ul Ul L>.I a: a. " ....... <, <, 30 to 1 BLOCK 2 PISTON 3 SHELBY 3 -_ - <, "' <, __ "- -_ __ -_ zyxwvutsrqponmlkjihgfedcbaZYXWVUTS <, _ ----...""::::.---....zyxwvutsrqponmlkjihgfedcbaZYXW ----.::: 3zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA ::!; 0 (J -_ ------- 0 0 0.2 0.1 PRESSURE. 0.3 6 . kg/cm2 Fig. 2.11 Compressibility modulus versus vertical stress curves (after Milovic, 1971a) 55 ,ec\ FAILURE ~y FOLLOWED RAPID DROP IN (0".-0"3) SAMPLING METHOD 200zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA ,LeER' Cl! ~ -, Z ....: 100 b I b- o'----"----"----..__---'zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONML o STRAIN 3 2 - PERCENT 4 Pig. 2.12 Stress-strain curves for tube and block samples (after Raymond. 1971) 1I0r---------------___, 100 '0 10 10 60 r:r,' , .e. 1. 10 20 10 010 to zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA er; •.•. 1. P·19. 2. 13 Stress paths for "perfect" and "ground" specimens (after Skempton and Sowa, 1963) 56 10zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA o o, -oX ::s: b ~ 12 (kPa) (CJ'~ • CJ'~) zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONML -50zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA (a) RlzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA cF_ to ('/.1 2.0 Axial 3.0 strain.Eo (.,.) -50 (b) Normally consolidated soil: AG - "in-situ" path, AB - "perfect" sampling path BC - triaxial compression after "perfect" sampling Overconsolidated soil (OCR = 7.4): DEF - "in-situ" path. DE - "perfect" sampling path EF - triaxial compression after "perfect" sampling Fig. 2.14 "Perfect" sampling behaviour of normally consolidated and heavily overconsolidated North Sea clays: (a) Stress paths (b) Stress-strain curves (after Hight et aI, 1985) 57 -0. Fig. 2.15 IS C Effect of stress history on "perfect" sampling stress path for low plasticity clay (after Hight and Burland, 1990)zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQ - E Ccmpresslon - E from in situ stress stcte Exl<:nsion C PS compr~$$ionl . Exll!:nS10n cuer perfect sampling 0.2 -0.2 Fig. 2.16 Effect of stress history on undrained triaxial compression and extension strength after "perfect" sampling in low plasticity clay (after Hight and Burland, 1990) 58 0.2 a.zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA Empire (PI=S3%) -ti'zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA N ..... ' ( eTa + O"r') /2CTap , Fig. 2.17 Effect of soil composition on "perfect" sampling stress path for normally consolidated soils (after Hight and Burland, 1990) ' , ( eTa • CTr')zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA 12eT cp Fig. 2.18 Effect of soil composition on undrained triaxial compression strength after "perfect" sampling of normally consolidated soils (after Hight and Burland, 1990) 59 2.80 -- 2.80 "1\" \\ 2.40 ~ .. c> c> I .. 2. 00 tlltJ.Olla&ll. II-SIN '_1" It 'UflCl Cl I, •. _ """' .• ,.,,.,, "'." ,., · c.llkIIUIDI 'al"'. -1\ 1\ 1.8 0 ~ [\ 1.6 0 1. 40 -, .- I~ 1.20 r 1.0 0 0.02 S~rLE 8-/1-3 0.05 0.1 0.5 0.2 VERTICAL Fig. 2.19 IzyxwvutsrqponmlkjihgfedcbaZYXW I 1IUIUll.. ,."" ... " 2.2zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA 0 t- oe ee '. -zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQP zyxwvutsrqponmlkjihgfedcbaZYXWVUTSR 1 EFfECTIVE 2 :; STRESS. k2'cm1 0 10 5o Influence of "perfect" sampling on compressibility of San Francisco Bay Mud (after Noorany and Poormand, 1973) .. e E -~ b -~ b ~ 0.33 ~ .....:r . -:r ezyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA -15" ~ b I - > b ...... :r b I ~ I. 80 (C1~. Fig. 2.20 ~) I (C1~ • C1~) mo. " Eo ('1.) Stress-strain characteristics of block samples of reconstituted Lower Cromer 1111in unconsolidated undrained triaxial compression tests (after Hight et al, 1985) 60 B --T-~- i-- I - IJ CD NzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA Z z c o ;: C u .9 -O.5r--r--t--+--+-+t--A=-_J_~ czyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA E tal tal -1.0 r-1r----t--t--+++-~:.-:4..:.::..:.:..J V2rticol Fig. 2.21 strain. E:zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA -i; zzzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA Strain paths for an element on the centreline of a tube sampler (after Baligh, 1985) 61 TubezyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA SompDng abc - compl1lssion cycle cd c - exlension cycle I zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA sce Fig. 2.21 tfgh - ulrusion ID-~'" N Fig. 2.22 Predicted stress paths for tube sampling of normally consolidated soil (after Hight, 1986) Tube Sampling d,h- I "ij]-". abc - compl1lsslon cycle cde _ eXlension cycle SIC Fig. 2.21 Fig. 2.23 Predicted stress paths for tube sampling of overconsolidated soil (after Hight, 1986)zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA 62 ..zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA N ,.. If I zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA e... (CTJ.CT~)/2 a b c d = UU = UU = UU = stress stress stress UU stress path path path path and and and and stress-strain stress-strain stress-strain stress-strain curve curve curve curve for for for for "perfect" sample tube sample "in-situ" young sample "in-situ" aged sample Fig. 2.24 Unconsolidated undrained (VU) stress paths .and stress-strain curves for "perfect", tube and "in-situ" samples 0.:> r---r--.,----r---, Solid S1mb." 0.4 - Pre, hear Condition.zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQ ·Undl,turbed· ·Undl'turbed" aehovlor a,havlor (Tilt') (T.. , I) 0.1 O~---L--~---~--~ o 2 € a 3 4 0 0.1 0.2 (%) 0.6 0.7 08 0.5 0.4 (OJ' +zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLK <T~ )/2<T:c 0.3 V Test 2 : Ideal sampling disturbance 0 Test 3 : Tube penetration disturbance Fig. 2.25 Tube penetration disturbance and ideal sampling disturbance effects in Boston Blue Clay (after Baligh et al, 1987) 63 10 ---_ ---_ 60zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA SO ;;zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA ~ 40 0.. N OCR=1.0 '.. 30 .. b Undisturbed Disturbed I E 20 vi V!zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA w ' 0:: 10 ..... V! 0:: -e w 0 J: !2 V! !4 !6 !8 'AXIAL STRAIN, -10 :!:10 :t12 Ea !14 (%1 EXTENSION -20 --- -- -30 SOzyxwvutsrqponmlkjihgfedcbaZYXWVUTS ---- 40 ii 0.. = 30 OCR=2.S N :::::: 20 c"- Undisturbed Disturbed .. I E10 vi'zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDC VI w 0:: ..... VI 0 0:: -e w J: -10 !1 !2 :t3 . !4 AXIAL STRAIN, Ea !S !6 %7 (%1 VI -20 ...... -----30 Fig. 2.26 Comparison of triaxial tests on undisturbed and disturbed specimens of Drammen Gay (after Lacasse and Berre, 1988) 64 Simple Shear ••zyxwvutsrqponmlkjih No.5 NoS A NoB • zyxwvutsrqponmlkjihgfedcbaZY 'Y No.IO Nol2 • No.13 No.1'I • Triaxial 2 •x 3 ~ I 0 CKoU - I \ \ 0 .;. \ • CKoPu- • A 2 \zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA • \ 3 e \ 4 TV 0 \ \ \zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGF ,.zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA ... \ \ ,, , ~ zyxwvutsrqponmlkjihg '.,,I .~ - Disturbance Ratio O~I(J.;. oeo: Fig. 2.27 Disturbed strength ratio versus disturbance ratio (after Okumura, 1971) 65 Fig. 2.28 Predicted deformation pattern during cone penetration in saturated clays (after Levadoux and Baligh, 1980)zyxwvutsrqponmlkji .zyxwvutsrqponmlkjihgfedcba ~h.~ rH aE r~. ~ .. ,.......... . " Fig. 2.29 Deformations during sampling of saturated clays: (a) Flat-ended wall (b) Simple sampler (after Baligh, 1985) 66 CHAPTER 3 FINITE ELEMENT ANALYSIS '3.1 THE MAIN OBJECTIVES The broad aim of the finite element analysis was to develop an approximate numericalmethod to predict the strain paths of soil elements due to the undrained penetration of a sampler. The numerical technique was then utilised to reach the following objectives: (i) To study the strain histories of soil elements at different positions within the sampling tube due to undrained penetration of some of the samplers which are widely used in sampling of soils. The cutting shoe designs of the following samplers were modelled: (a) Norwegian Geotechnical Institute (NGI) 54 mm diameter piston sampler, (b) Swedish Geotechnical Institute (SG1) 50 mm diameter piston sampler, and (c) two typical British Standard General Purpose 100 mm diameter open-drive samplers. (ii) A parametric study was performed in order to assess the effects of area ratio of sampler, inside clearance ratio of sampler and cutting edge taper angles on the strain histories of soil elements. (iii) Finally. the strain paths of some flat-ended samplers of different thickness andzyxwvutsrqpo BIt ratio were investigated.zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA 67 3.2 DEVELOPMENT OF AN ANALYTICAL TECHNIQUE FOR STRAIN PATH COMPUTATION 3.2.1 INTRODUCTION Deep steady penetration of a sampler is an axisymmetric problem. Considering undrained shearing of the soil and neglecting visco-elastic and. inertial effects, the process of penetration is reduced to a problem of irrotational steady flow of an incompressible, inviscid fluid around a sampler under conditions of axial symmetry (i.e., properties and flow characteristics are independent of the tangential coordinate). In this flow problem the soil particles move along streamlines around the sampler. A solution of such a problem, therefore, consists of obtaining strains and deformations of soil elements along different streamlines within the sampling tube. Prediction of strains and deformations along the streamlines at different locations within the sampler tube is not a straightforward task. However, it is possible to predict flow velocities around a sampler under conditions of axial symmetry using the Finite Element Method. For the prediction of flow velocities LUSAS Finite Element Package (version 86.07) was used. LUSAS is a general purpose engineering analysis system which incorporates facilities for linear and non-linear static stress analysis, step by step dynamic analysis, natural frequency analysis, linear buckling analysis, spectral response analysis, and steady and transient field (thermal) analysis. The system is based on the finite element displacement method of analysis and contains a comprehensive range of elements and solution procedures for the analysis of most types of engineering problems. Several computer programs (written in Fortran 77) have been developed for computing the strain paths from the results obtained from finite element analyses. A description of the problems analysed, the relevant theory incorporated in the LUSAS analysis system to predict flow velocities, technique for the computation of strains, errors in the analyses and further analyses to reduce errors are presented in the following sections. 68 3.2.2 PROBLEM DESCRIPTION The problem considered was irrotational steady flow of an incompressible, inviscid fluid around a thin walled sampler under conditions using a potential formulation. neglected. The frictional of axial symmetry drag at the sampler boundary has been The sampler has the following characteristics: External diameter of sampler tube, B = 54 mm Internal diameter of sampler tube, D. = 51 mm Internal diameter of the sampler cutting edge, D, = 50.5 mm Thickness of the sampler tube,zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIH t = 1.S mm Length of the sampler tube, L = 300 mm Area ratio = 14.34% Inside clearance ratio = 0.99% Inside cutting edge taper angle = 0.716° Outside cutting edge taper angle = 5° 3.2.3 THEORY FOR ANALYSIS The steady-state flow through a two-dimensional soil mass is governed by the following differential equation .... (3.1 ) in which K, and K, are the coefficients of permeability in x and y directions respectively, <I> is the fluid potential or piezometric head (<I> = ply + z; p ='pressure, 'Y = specific weight, z = elevation head), and Q is the internal flow source. With reference to Fig. 3.1, the types of boundary conditions applicable are as follows: Type (A): The value of the unknown to be specified at nodal points on the boundary .... (3.2) Mathematically, this is termed the Dirichlet boundary condition. Type (B): That a boundary loading exists of the form a; a; Kxox Lx+ Kv~Lv+q+a(;-;a) - 0 69 .... (3.3) in which q,zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA a and (jl. are constants and L", L, are direction cosines between the outward normal, n and x and y axes respectively. condition. The physical significance This is called Cauchy boundary of this second boundary condition is best = K, = K. The boundary condition type illustrated by considering isotropic case K,.zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPON (B) then reduces to .... (3.4) where, d(jl/dn is the velocity gradient in a direction normal to the surface at the point under consideration. By taking approximate values of q and ex well established physical boundary conditions can be recovered. Case I, = q ex = 0, then Equation 3.4 reduces to o~ ...0 on .... (3.5) which implies that the velocity or flow gradient in a direction normal to the surface is zero. In other words, the ponion of the surface is perfectly impermeable. Case 2, a = 0, then o~ on K- ..-q .... (3.6) This states that a specified quantity of fluid, q, flows into the body per unit area of the surface. Case 3, q This is well known as the flux boundary condition. = 0 In this case Equation (3.4) reduces to o~ on K---a(~-~ a ) .... (3.7) This states that the flow of fluid from any point on the surface proponional to the difference in pressure head or potential, is directly (jl and the ambient potential, (jl.. This is known as convection boundary condition. Now, for a two-dimensional situation the components of the hydraulic gradient are defined as 70 . 1 x Where . oh ohzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA 1 =- .. - .... (3.8) YzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA oy oXzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA h is the pressure incompressible, head at any point. Assuming that the fluid is the so-called continuity equation can be developed by considering the flow into and out of an elemental volume of material per unit o~, time. In order to balance the fluid input with the fluid output for the material element shown in Fig. 3.2, it is required, that OVx OVY)A uxuy= A ( --+-ox oy where and v'l represent Va respectively at a point. 0 .... (3.9) the velocity of the flow in the x and y directions Since by definition the material permeability is the amount of fluid flowing through unit area per unit time in the presence of a unit hydraulic gradient oh oh vzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA --K v "'-K.... (3.10) x xox Y oy where negative sign indicates that flow occurs in the direction of decreasing pressure head and h is the pressure head. -o ox (K Oh) xox 0 ( K Yoy Oh) - +- oy Substituting from Equation (3.10) in (3.9)zyxwvutsrqponmlkjihgfedcbaZYX .... (3.11) =0 If there is a change in volume/unit volume at the rate dV/dt in the element during flow, then continuity demands that e ox (K xoxOh) oy0 (K YoyOh) + - dV - dt - 0 .... (3.12) Further if fluid is being injected into the element at a rate of S per unit volume then .!_(K Oh)+'!_(K ox xox oy Oh)+s_dV =0 Yoy dt .... (3.13) The solution procedure for Equation (3.13) can be formulated in terms of potential function $(x,y). Defining a potential function $(x,y) such that 71 0; zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA .... (3.14) vzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFED --K xzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA xox Then from Equation (3.10) we have that 0; oh --oy oy 0; oh --ox ox .... (3.15) Substituting in Equation (3.13) provides !_(K a;)+!_(K ax xox oy a;)+S_dV Yoy dt =0 .... (3.16) Equations (3.15) can be integrated to give .... (3.17) ;(x.y)"'h(x.y)+C Where C is a constant of integration. to (j) (x.y) a family of curves. equipotential defined lines, will be obtained. constant. equal to h., hl' h, etc. If a series of values (j)lt (j);u (j), etc. are assigned by Equation (3.17), which are called Along each of these lines the head h(x,y) is Thus lines on which pressure head is known to be constant. must have a constant value of potential (j), prescribed for finite element analysis. At the impermeable boundary, it is required that the flow velocity in the direction normal to the boundary be zero. From Equation (3.14) this condition becomes KzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA o;L +K o;L -0 xox x voy v .... (3.18) Comparing Equations (3.1) and (3.3) respectively with Equations (3.16) and (3.18) it is seen that a potential analysis of a flow problem reduces to solution of the differential Equation (3.1) with dV S--=Q .... (3.19) dt and subject to boundary condition type (B) with q = a = O. Also at points where the pressure head is known, (j) must be prescribed as boundary condition type (A). 72 The finite element solution of the differential Equation (3.1) subject to the boundary conditions, Equations (3.2) and (3.3) can be derived either from variational or Galerkin approach (Hinton and Owen, 1979; Rao, 1989). Once the nodal values of the potential ~ have been evaluated, the flow velocity in each element can be calculated according to Equation (3.14) 3.2.4 FINITE ELEMENT MODEL The art of a finite element analysis of this kind of problem lies in the development of a suitable idealisation of a flow domain. The fineness of the mesh must be varied to provide results of acceptable accuracy at required points in the flow domain. However, for a problem as described earlier, it is not possible to know the likely distribution of flow velocities around the sampler before an analysis has been done. As a result, it was necessary to run a pilot coarse mesh idealisation in the first instance. The mesh was then refined to reduce errors and to obtain acceptable results. The finite element mesh used (Fig. 3.4) contained 56 four-noded quadrilateral axisymmetric field (QXF4) elements. QXF4 is an isoparametric element (Fig. 3.3). The variation of fluid potential (~) within the element is linear. The elements are numerically integrated and the number of integration points is four. The y-axis is taken as the axis of symmetry.zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA 3.2.S MATERIAL PROPERTIES AND BOUNDARY CONDITIONS The permeabilities,zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA K, and K,. were assumed to be the same and were 1 m/day. The values of potential (~) along the boundaries AE and ID (see Fig. 3.4) were the same and were equal to zero. ~ = 1.0 was assumed for the boundary BC. Along CD and both sides of the sampler, i.e., EFG and GHI,zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCB d$/on = 0, where n is the outward normal drawn to the surface of the sampler. The boundary CD and the sampler are thus represented by a boundary where no transverse flow occurs or in other words the surface of the sampler and the boundary CD are perfectly impermeable. Also symmetry conditions along AB (centreline of the sampler) require o~/on = 0 on this section. The different boundary conditions are shown in Fig. 3.4. 73 3.2.6 COMPUTATION OF STRAINS AND DEFORMATIONS In order to calculate radial and axial strains along the streamlines at different locations within the sampling tube detailed computation was carried out in several steps. The steps followed are described below. Step I: Once a decision was made on the type of element, "the element mesh, material properties, boundary conditions and loading cases for the problem, a descriptive data file (DATA1.DAT) was prepared for LUSAS. This input data file was run on the batch system of the PRIME main-frame computer. The output file (DATAl.OUT) from the LUSAS analysis contains results under headings of element topology, node coordinates, material properties, support nodes, load case, summary of data, field gradients, flow velocities, field value at nodes and reactions to earth. Step II: For further analysis, only the element topology, node coordinates and flow velocity in y-direction at each node are needed. A computer program, written in Fortran 77, was developed to read the entire output file from LUSAS analysis and to print only the required data. The computer program was named PROG 1. The input parameters for PROG I were the name of the LUSAS output file (OATAI.OUT) and the number of nodes in each element. PROG I first reads the whole file DATA1.0UT by searching different headings in the file DATA1.0UT and once the appropriate heading has been found, the contents under that particular heading are printed in a separate output file (NEWI-A). The new output file contains a listing of element topology, i.e., the element numbers and the corresponding node numbers connected to each element, the node coordinates and velocity of flow in y-direction, v'/ at each node. A listing of PROGI is given in Appendix-A. Step III: An input data file (NODE!) was prepared which is essentially an array containing all the node numbers of the finite element mesh. PROG3, was written in Fortran 77. A program, named The general input for the program are as follows: (i) the name of file containing an array of node numbers of the finite element mesh, i.e., NODEI; (ii) the name of file containing element topology, node coordinates and nodalzyxwvutsrq 74 velocities, i.e., NEW1-A; (iii) total number of columns in file NODE!; (iv) total number of nodes in each element; (v) number of columns in file NODEl counting from the centreline to the inside edge of the sampler; (vi) internal radius of the cutting shoe of the sampler; (vii) the row number in file NODEI that corresponds to the bottom of the sampler; and; (viii) the name of the new output file. PROG3 first reads the files NODEI and NEW1-A and then calculates the values of stream functions corresponding points to ten streamlines within the sampling tube. passing through ten prescribed These points lie at a depth corresponding to the bottom of the sampler and are located at a distance of 10% to 100% (10%, 20%, 30%, 40%, 50%, 60%, 70%, 80%, 90% and 100%) of the cutting shoe radius ~ from the centreline of the sampler. Finally, the complete position of each streamline is located by estimating its position at different depths along the streamline. The procedures followed to establish the position of the streamlines are outlined below. (a) The average flow velocity in the y-direction (Vy) was calculated for each node of the mesh. (b) A linear variation of Vy between two successive nodes (e.g., a node on the centreline of the sampler and the node next to it, i.e., the node on the inside edge of the sampler) at all depths was assumed. (c) The value of stream function at all the nodes lying on the centreline of the sampler, l.e., on the axis of symmetry was assumed to be zero, since the centreline of the sampler constitutes a streamline. (d) The flow increment between two consecutive nodes was computed by integrating the velocity profile over the corresponding area between these two nodes. The value of the stream function at any node was obtained by adding the value of stream function of the previous node and the incremental value of stream function between the two nodes considered. Following this procedure the magnitudes of the stream function at all the nodes of the mesh were estimated. (e) At a depth corresponding to the bottom of the sampler, a parabolic variation of stream function was assumed between the node on the centreline of the sampler and 75 the next node lying on the inside edge of the sampler. Stream functions at ten locations (10%, 20%, 30%, 40%, 50%, 60%, 70%, 80%, 90% and 100% of cutting shoe radius from the centreline of the sampler) were calculated. In this way, the location of ten streamlines for ten stream function values was defined at a depth corresponding to the bottom of the sampler.zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQP (f) At all other depths, a similar parabolic variation of stream function between two successive nodes was assumed. Using one of the stream function values as calculated previously the positions of the corresponding streamline were estimated at all other depths. The procedure was repeated for the rest of the stream function values to locate the complete position of all the respective streamlines. All the results of analyses are printed in a separate output file (NEWt-B) which contains a listing of the magnitudes of the stream functions and the position of the respective streamlines. A listing of PROG3 is presented in Appendix-A. StepzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA N: Another program was written in Fortran 77. The input parameters for this program, named PROG4, are: (a) the name of the me containing a listing of stream function values for the streamlines and their respective positions, i.e., NEWt-B; (b) the total number of streamlines; (c) the total number of rows in the file NODEI; and; (d) the name of the new output file. PROG4 first reads the file NEWI-B and then calculates the magnitudes of strains, both radial and axial at different depths along the streamlines. The estimation of strains was carried out in accordance with the following procedure: (a) The radial strain at any point along the boundary BC (Fig. 3.4) which is situated at a depth far below the bottom of the sampler was assumed to be zero. (b). For any streamline, the radial strains at all other depths were estimated by comparing its position at the depth considered with that on the boundary BC (c) It is assumed that the volumetric strain for an axisymmetric undrained penetration problem is zero. Therefore, the magnitude of axial strain (e.) is twice the negative value of the radial strainta). It should be mentioned that Ea is not equal to all points in the flow domain, but in effect this is assumed to be so. -2Er at This interpretation has been made because the data were produced to give input for stress path triaxial tests. It is probably reasonable for the middle 50% of the sample, where Er is approximately constant. Beyond this it is not correct, but it is necessary since radial variations in strain can not be applied to a triaxial specimen.zyxwvutsrqponmlkjihgfedcbaZ 76 The results are printed in a separate output file (NEWt-C) that contains a listing of radial and axial strains along the streamlines together with the values of stream functions for the streamlines and their respective positions. A listing of PROG4 is given in Appendix-A. Step V: Finally another program, named PROGS was written in Fortran 77. The general input for this program are: (a) the file NEWt-C; (b) the total number of streamlines; (c) the total number of rows in file NODE; (d) the depth at the bottom of sampler; (e) the external diameter of the sampler tube; and; (t) the name of the new output file. This program first reads the entire file NEWt-C. It then calculates the cumulative axial deformation at different depths along the streamlines. The cumulative axial deformation at different depths along each streamline was calculated according to the following procedure: (a) A linear variation of axial strain was assumed between two successive nodes (e.g., a node on the boundary BC and the node immediately above it) at all depths was assumed. (b) The axial deformation increment between the two consecutive nodes was computed by finding the area below the axial strain profile between these two nodes. (c) The value of the cumulative axial deformation at any node was obtained by adding the value of the axial deformation of node below it and the incremental value of axial deformation between the nodes considered. Following this procedure the magnitudes of the cumulative axial deformation at all depths for each streamline were estimated. The results are printed in a separate output file (NEWt-D). contains a listing of cumulative axial deformations at different The file NEWt-D depths along the streamlines, depth to external diameter ratio at different depths for all the streamlineszyxwvutsrqp 77 together with all the results printed in the output file NEWt-C. A listing of PROG5 is given in Appendix-A. 3.2.7 ERRORS IN THE ANALYSES In order to check the accuracy of the analyses, firstly, the 10th streamline which passes through the tip of the bottom of the cutting shoe has been plotted as shown in Fig 3.5. Secondly, the strain paths of soil elements at three different locations within the sampling tube have also been plotted. These strain paths are shown in Fig 3.6. Three different types of error have been noticed from Figs. 3.5 and 3.6. These are categorised as follows: Error Type I: A streamline touching the tip of the bottom of the cutting shoe (point C of Fig. 3.5) must follow the inside edge (line ABC of Fig. 3.5) of the sampler tube. However from Fig 3.5 it is quite obvious that a large discrepancy occurs. In order to assess the magnitude of this error the difference between the position of the inside edge of the sampler and the actual position of the streamline obtained from analyses has been expressed as a percentage of the actual position of the streamline. The maximum error was found to be as high as 2.05 per cent. It was noticed that the error was highest at a depth corresponding to the point F (Fig. 3.4) lying on the inside edge of the cutting shoe. This error was thought to result because of the variation of stream function values at nodes along the inside edge of the sampler. This was checked by calculating the stream function at all the nodes along the inside edge of the sampler. To achieve this PROG3 was modified so that it computed the values of average flow velocity, v., and the stream function at each and every node and printed the results together with the node coordinates in a separate output file. The modified program was named PROG2 and has been listed in Appendix-A. Error Type II: In Fig 3.6 the strain paths at three different locations within the sampler have been shown. From Fig 3.6 it is evident that the strain histories are In other words, soil elements near the inside edge and independent of their position.zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFED the centreline of the sampler have suffered identical strains. With the mesh shown in Fig 3.4, this result was inevitable where only one element has been used from the centreline to the inside edge of the sampler. However, intuitively the degree of disturbance would be expected to vary across the diameter of the sampling tube. Soil 78 elements near the inside edge of the sampler are expected to strain more than those near the centreline of the sampler (Burmister, 1936; Hvorslev, 1949). Error Type III: The strain paths in Fig 3.6 also indicate that during penetration soil elements are subjected to two phases of undrained shearing. Firstly, an extension phase (ab) below and near the vicinity of cutting edge of sampler and secondly a compression phase (be) inside the sampler. However, it has previously been found from closed form analytical solution of strains on centreline (Baligh, 1985) that undrained penetration of simple samplers with flat-ended and round-ended walls produce three distinct phases of triaxial shearing. These are: (a) an initial compression phase ahead of the sampler; (b) an extension phase in the vicinity of cutting edge; and (c) a second compression phase inside the sampler. It is, in fact, difficult to ascertain the reasons for the above mentioned errors without performing further analyses with a refined mesh, changing the element type or modifying boundary conditions. It is well established that the accuracy of results from finite element analyses are very much dependent on type of element used, boundary conditions and refinement of mesh. Nevertheless, it can be stated that the errors in the analyses perhaps resulted because of adopting a relatively coarse finite element mesh, with only one element from the centreline to the inside edge of the sampler. 3.2.8 MINIMISATION OF ERRORS In order to minimise the errors in the analyses the following points were considered: (i) Increasing the number of elements in the mesh. (ii) Refining the mesh in areas near the edges and bottom of the sampler. (iii) Fixing the boundary DC (Fig 3.4) further from the sampler so that it did not interfere with the actual area of interest, l.e., the inside of the sampling tube. (iv) Changing type of element. Several analyses were carried out to assess the implementation of the above mentioned points to minimise errors. An analysis was performed with a sampler of thickness 1.25 mm. The area ratio and inside clearance ratio of the sampler were 79 12.23% and 0.99% respectively. Theoutside and inside cutting edge taper angles were respectively 4.3° and 0.72°. 400 four-noded quadrilateral asymmetric field elements (QXF4) were used in the analysis. The sampler lengthzyxwvutsrqponmlkjihgfedcba (L) was 150 mm. The finite element mesh was refined only near the inside edge of the sampler. The boundary DC (Fig 3.7) was fixed at a distance of four times the internal radius (R) of the sampler from the centreline of the sampler. Other boundary conditions were similar to those shown in Fig 3.4. A schematic diagram of the mesh is shown in Fig. 3.7. From the results of analyses it was found that error type I was reduced to a maximum value of 0.91%. An error of this magnitude was still considered significant. Strain paths at five different locations within the sampler were plotted and are shown in Fig 3.8. From Fig. 3.8, it is evident that both error types II and III are eliminated. Another analysis was carried out for the same sampler, but with 480 QXF4 elements. In this case, the mesh with 400 elements was refined to some extent near the inside edge of the sampler. It was found that error type I was reduced to a maximum value of 0.795%. In order further to reduce type I errors, an analysis was carried out with 2156 QXF4 elements. In this case the length (L) and thickness of the sampler tube were 120 mm and 1.25 mm respectively. In this analysis, the mesh was refined near both the inside and outside edges of the sampler and the boundary DC was fixed at a distance of 3.9 times the internal radius of the sampler. The geometric characteristics of the sampler, the material properties and boundary conditions were the same as those used in the previous analyses. From the results of the analysis it was found that the type I error was decreased to a maximum value of 0.54%. Although the type I error was reduced to some extent by increasing the number of elements and refining the mesh near the edges of the sampler, the abrupt increase in error at the depth corresponding to point F (Fig. 3.7) was not eliminated. It was then decided to investigate the effect of changing element type on type I error. First an analysis was carried out with the same sampler and mesh shown in Fig. 3.4 but using 9-noded quadrilateral axisymmetric field element. Such an element is designated as QXF9. QXF9 (Fig. 3.9) is an isoparametric element which is capable of modelling curved boundaries. The variation of field value within the element is quadratic and the number of integration points is five. equidistant from the two adjacent corner nodes. The mid-side node is From the results of the above analysis. the type I error was found to be 1.28%. The error this time was, however, 80 identical in magnitude throughout the depth of the sampler. Finally, another analysis was performed for the same sampler with a mesh consisting of 2156 QXF9 elements. The material properties and boundary conditions were similar to those for the analysis done with 2156 QXF4 elements. 0.3%. It was found that the type I error was reduced to An error of such a magnitude was considered eliminate error type I completely, Vy acceptable. However, to was corrected for some of the nodes lying on the inside edge of the sampler so that the magnitude of the stream function at all the nodes along the inside edge of sampler were approximately equal. A comparison of the type I error for all the analyses carried out is shown in Fig. 3.10. In order to model piston sampling, it was required to change the boundary conditions along the boundary AE so that the total axial deformation along the boundary AE was as minimum as possible and preferably less than 0.1 mm. To achieve this, two types of distribution of cI>-valuealong the boundary AE was considered. Firstly, a triangular variation with a positive cI>-valueat the centreline and an equal negative value at the inside edge of the sampler and secondly, a constant positive value of cl> along the boundary AE. Several numerical experiments were carried out with these two types of boundary conditions using the mesh (Fig. 3.7) consisting of 400 QXF4 elements. It was found that minimisation of total axial deformations along the boundary AE was only achieved by adopting the second type of distribution of cI>value. 3.2.9 CONCLUDING REMARKS From the analyses described in the previous articles, it is evident that significant improvement of errors can be achieved by using 9-noded elements. It is also apparent that an increased number of elements and refinement of the mesh near the inside and outside edges of sampler contribute to a reduced error. Therefore, it was concluded that for each analysis proposed for this research at least 2000 9-noded axisymmetric field elements should be used. The mesh should be refined near both the inside and outside edges of the sampler. The boundary DC (Fig. 3.4) should be located at a distance of at least 3 to 4 times the internal radius of the sampler tube (R) from the centreline of the sampler. To model piston sampling a constant positive value of cl> should be assigned for the boundary AE that minimises the value of total axial deformation along the boundary AE. 81 Finally, in order to eliminate the type I error, the average Vy for the required nodes lying on the inside edge of the sampler should be corrected. 3.3 DETAILS OF COMPUTATIONAL PROGRAMME All the analyses included in the computational programme were carried out using In all the analyses QXF9 elements with the newly developed analytical technique.zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONML aspect ratios less than 5:1 were used. The material properties K,. and K, for all analyses were assumed to be the same and were 0.01 mm/sec. The computational programme consists of five categories of analysis. These are as follows: (a) In the first series of analyses, the cutting shoe geometries of NGI, SGI and Ul00 samplers were analysed. Although Ul00 samplers are open-drive samplers, these were modelled as piston samplers for the sake of simplicity and comparisons. (b) The second series of analyses were conducted on four samplers with area ratios 10.14%, 29.64%, 50.73% and 100.46%. The inside clearance ratio of the samplers were 0.99%. Inside and outside cutting edge taper angles of the sampler were respectively 0.716° and 9.9°. These analyses were performed to understand the effect of area ratio on strain paths of soil elements. (c) The third series of analyses were carried out to investigate the effect of inside clearance ratio on strain histories of soil elements. For this purpose, three analyses were done with samplers having inside clearance ratios 0.495%, 1.98% and 4.96%. The area ratio of all the samplers were 29.64% while the inside and outside cutting edge taper angles were 0.7160 and 9.90 respectively. (d) The fourth series of analyses were performed to investigate the influence of inside and outside cutting edge taper angles on the strain paths of soil elements. Firstly, two samplers with inside cutting edge taper angles 0.3580 and 1.43° were studied. The outside cutting edge taper angle of both the samplers were 9.90 Secondly, another two samplers with outside cutting edge taper angles equal to 5° and 19.29° were analysed. The inside cutting edge taper angle for these samplers was 0.716°. The area ratio and inside clearance ratio of the four samplers studied under this category were 29.64% and 0.99% respectively. (e) The fifth series of analyses were conducted on four flat ended samplers. For three of these samplers, the external diameter and thickness were equal to those of NGI, SGI and Ul00 samplers. The external diameter to thickness ratios (BIt) of 82 these samplers were 45.6, 12.2 and 19.9. The external diameter to thickness ratiozyxwvutsrq (BIt) of the other sampler studied was 23. This series of analyses was carried out to understand how cutting shoe designs of two samplers having identical diameter and thickness affect strain paths of soil elements. The analyses also enabled assessment BIt ratio on strain paths of soil elements for flatof the effect of thickness andzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA ended samplers. 3.4 ANALVSIS OF NGI, SGI AND UIOO SAMPLERS The characteristics of the NOI piston sampler, SOl piston sampler and the two typical Ul00 samplers are presented in Table 3.1. In the analysis a sampler tube length (L) of 120 mm has been used for both the NOI and SOl ·samplers. The length of the sampler tube for the analysis of UlDO (type I) and UlDO (type II) samplers was taken as 204 mm. 3.4.1 CUTTING SHOE DESIGNS OF THE SAMPLERS The cutting shoe designs of the NOI and SOl samplerszyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLK are shown in Figs. 3.11(a) and 3.11(b) respectively. For the NOI sampler, the inside cutting edge taper angle is 1.43°, while the outside cutting edge taper angle is 12° up to a tube thickness of 0.47 mm and 7° thereafter. For the SOl sampler, the inside and outside cutting edge taper angles are 0.106° and 5° respectively. In the original design, (Kallstenius, 1961) the outside cutting edge taper angle was specified as 45° up to a thickness of 0.3 mm and 5° thereafter. The initial outside cutting edge taper angle of 5° up to a thickness of 0.3 mm was not modelled in the analyses to avoid complexity in generating node coordinates and to restrict the total number of elements in the finite element mesh. Details of the cutting shoe designs of the Ul00 (type I) and UIOO (type II) are presented in Figs 3.12(a) and 3.12(b) respectively. For the Ul00 (type I) sampler, the inside clearance is provided by tapering the inside diameter of the cutting shoe at an angle of 1.1° to meet the 105.5 mm internal diameter of the sampler tube. The outside cutting edge taper angle is 2fP up to a thickness of 2.585 mm and 7° thereafter. 83 In case of UlDO (type II) sampler, the inside clearance is created by providing a uniform internal diameter (l05.7 mm) to the cutting shoe and thus stepping out abruptly at the junction of the shoe and the sampler tube. The outside cutting edge In the analysis the inside clearance taper angle is 300 up to a thickness of3A-6 mm.zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPON has been provided by an inclined (45°) step instead of a flat step at the junction of the cutting shoe and the sampler tube. For both Ul00 (type I) and Ul00 (type II) samplers, the outside clearance were neglected. Therefore, in the analyses, thickness of the sampler tube for UIOO (type I) and UIOO (type II) samplers were taken as 5.785 mm and 5.9 mm respectively throughout the full length of the sampler. Consequently,zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA Bit ratio of the Ul00 samplers studied were 19.9. 3.4.2 FINITE ELEMENT MODELS AND BOUNDARY CONDITIONS Figs. 3.13 and 3.14 show the schematic diagrams of the element meshes used for analysing NGI and SGI samplers respectively. QXF9 elements. Both the meshes consisted of 2254 Total number of nodes in each mesh was 9357. The schematic diagrams of the meshes used for analysing Ul00 (type I) and Ul00 (type II) samplers are shown in Figs. 3.15 and 3.16 respectively. (type I) and Ul00 elements respectively. The meshes used for analysing Ul00 (type II) consisted of 2550 QXF9 elements and 2436 QXF9 The respective total number of elements in the meshes were 10517 and 10065. All the meshes, shown in Figs. 3.13 - 3.16, were refined near the inside and outside edges of the samplers and also to some extent above and below the bottom of the samplers. In all the meshes the boundary BC was fixed at a distance of twice the length of sampler tube from the top of the sampler. locations of the boundary DC are shown in Table 3.2. The In order to model piston sampling several analyses for each sampler were carried out to minimise total axial deformations along the boundary AB and preferably to bring them to less than 0.1 mm. In all the analyses the boundary conditions along the boundaries AB, DC, EPG, GID, BC and ID were unchanged and these are as follows: (a)zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA a<l>/an = 0 along the boundaries AB, DC, EPG and GID; (b) <I> = 1000 along the boundary BC; and; (c) <I> = 0 along the boundary ID. 84 For each sampler different tj)-valueswere assigned along the boundary AE to simulate piston sampling and these values are listed in Table 3.2. In Table 3.2. the magnitudes of maximum total axial deformation along the boundary AB for all the analyses are also shown. In Table 3.2 positive axial deformation indicates compression while negative axial deformation means extension. The variation of total axial deformations along the boundary AE for the analyses which modelled piston sampling is shown in Fig. 3.17. 3.4.3 STRAIN PATHS OF SOIL ELEMENTS The strain paths of soil elements at six locations within the sampler for the NOI. SOl. UI00 (type I) and Ul00 (type IT) samplers are presented in Figs. 3.18. 3.19. 3.20 and 3.21 respectively. It is evident from Figs. 3.18 - 3.21 that for all strain paths the peak strains in compression and extension are not equal. The strain paths for these samplers also show that soil elements near the inside edge of the sampler are strained much more than those near the centreline of the sampler. For strain paths located at 10% and 30% of cutting shoe radius (RJ from the centreline of NOI sampler the peak compressive strain is higher than peak extension strain. Atzyxwvutsrqponmlkjihg 0.5 R, from the centreline the peak strains are approximately the same while at 0.7 R, and 0.9zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA R, from the centreline the peak strain in extension is greater than peak compressive strain. In case of SOl sampler. for all strain paths. the peak axial strain in compression is greater than that in extension. However. for the both Ul00 (type I) and Ul00 (type IT) samplers, the peak axial strain in extension is greater than that in compression for all the strain paths. For all the samplers the minimum peak axial strains in compression and extension (which occur at the centreline of the samplers) were determined by extrapolating the curves shown in Fig. 3.22. 3.5 PARAMETRIC STUDY OF CU'ITING SHOE DESIGNS In order to study the effect of cutting shoe design on strain histories of soil elements several analyses were carried Out. Four parameters have been considered and they are: (a) the area ratio of sampler; (b) the inside clearance ratio of sampler;zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA 85 (c) the inside cutting edge taper angle; and; (d) the outside cutting edge taper angle. 3.5.1 ANALYSES WITH DIFFERENT AREA RATIOS 3.5.1.1 DIMENSIONS AND CHARACTERISTICS OF THE SAMPLERS Four samplers with area ratios 10.14%, 29.64%, 50.73% and 100.46% were studied. The area ratios were varied by changing the thickness of the sampler or in other words by changing the external diameter of the sampler tube. Thicknesses of the sampler tubes were 1.00 mm, 3.25 mm, 5.5 mm and 10.25 mm and the respective external diameters were 53 mm, 57.5 mm, 62 mm and 71.5 mm. The internal diameter of the sampling tubes and the internal diameter at cutting shoe of all the samplers were 51 mm and 50.5 mm respectively. Thus the inside clearance ratio of all the samplers was 0.99%. Inside and outside cutting edge taper angles were also kept fixed for all samplers and their values were respectively 0.716° and 9.9°. Dimensions and designs of the cutting shoe of the samplers are shown in Fig. 3.23. Length of the sampler tube for the analysis of each sampler was 120 mm. 3.5.1.2 FINITE ELEMENT MODEL AND BOUNDARY CONDITIONS For all analyses a finite element mesh consisting of 2156 QXF9 elements was used. The total number of nodes in the mesh was 8963. A schematic diagram of the mesh is shown in Fig. 3.24. The mesh was refined near the inside and outside edges of the sampler and also in the vicinity of the cutting edge. The boundary BC was fixed at a distance of twice the length of sampler tube from the top of the sampler, while the boundary DC was located at a distance of 3.89R from the centreline of the sampler. For each sampler, analyses were carried out with different «I>-valuesalong the boundaryzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA AB in order to model piston sampling. In all the analyses the boundary conditions along the boundaries AB, DC, EFO, GHI, BC and ID were unchanged. These boundary conditions were similar to those reported for the previous analyses (see section 3.4.2). Different «I>-valuesassigned along the boundary AE for each sampler are listed in Table 3.3. Table 3.3 also shows the magnitudes of maximum total axial deformation along the boundary AE for all the analyses. Variation of the total axial deformation along the boundary AB for the analyses simulating piston 86 sampling is shown in Fig. 3.25. 3.5.1.3 STRAIN PATHS OF SOIL ELEMENTS The strain paths of soil elements at six locations within the sampler tube for samplers with different area ratios are shown in Figs. 3.26 - 3.29. It is evident from Fig. 3.26 that at small area ratio, for all the strain paths peak axial strain in compression is less than that in extension. For a sampler of moderate area ratio (see Fig. 3.27) the peak axial strain in compression is larger than that in extension for soil elements located near and around the centreline of sampler (i.e., 10% to 50% ofzyxwvutsrqponmlkjihgfedcbaZY R, from the centreline of sampler). However, for soil elements moving near and along the inside edge of the sampler, the peak axial strain in extension is greater than that in compression. Fig. 3.28 indicates that for a ~ampler of high area ratio, for strain paths located at 10% to 70% of R, from the centreline of sampler, the peak axial strains in compression are larger than those in extension. Whereas, for other strain paths the peak axial strains in extension are larger. However, for a sampler of very high area ratio, as evident from Fig. 3.29, for all the strain paths except the one moving along the inside edge of sampler, the peak axial strains in compression are larger than those in extension. For all the samplers, the minimum peak axial strains in compression and extension at the centreline were found by extrapolating the curves shown in Fig. 3.30. 3.5.2 ANALYSES WITH DIFFERENT INSIDE CLEARANCE RATIOS 3.5.2.1 DIMENSIONS AND CUTTING SHOE DESIGNS Three samplers with inside clearance ratios of 0.495%, 1.98% and 3.96% were analysed. Inside clearance ratios were varied by changing the inside diameter of the sampler tube while the internal diameter at the cutting shoe and external diameter of the sampler tube were kept fixed. The external diameter of the sampling tube and the internal diameter at the cutting shoe of the samplers were 57.5 mm and 50.5 mm respectively. So, the area ratios of all the samplers were 29.64%. Inside and outside cutting edge taper angles were unchanged for the samplers and their values were respectively 0.716° and 9.9°. Cutting shoe designs of the samplers are presented in Fig. 3.31. Length of the sampler tube for analysis of each sampler was 120 mm. 87 3.5.2.2 FINITE ELEMENT MODEL AND BOUNDARY CONDITIONS The finite element mesh shown in Fig. 3.24 consisting of 2156 elements of QXF9 type was used for each analysis. The boundary BC was fixed at a distance of twice the length of the sampler tube from the top of the sampler. The locations of the boundary DC are shown in Table 3.4. For each sampler, analyses were carried out with various q,-values along the boundary AE to simulate piston sampling. In all the analyses the boundary conditions along the boundaries AB, DC, EFO, OHI, BC and ID were unchanged and these boundary conditions were identical to those reported earlier (see section 3.4.2). Different values of q, assigned along the boundary AE for each sampler are shown in Table 3.4. The magnitudes of the maximum total axial deformations along the boundary AB for all the analyses are also listed in Table 3.4. The variation of the total axial deformation along the boundary AB for the analyses modelling piston sampling is presented in Fig. 3.32. 3.5.2.3 STRAIN PATHS OF SOIL ELEMENTS The strain paths of soil elements at various locations within the sampler tube for the samplers are presented in Figs. 3.33 - 3.35. Fig. 3.33 indicates that for a sampler of small inside clearance ratio, the peak:axial strains in compression are much greater than those in extension for all strain paths, while the strain paths in Fig. 3.34 show that for a sampler of high inside clearance ratio the peak: axial strain in extension is significantly higher than that in compression for all strain paths. Strain paths (Fig. 3.35) for the sampler of very high inside clearance ratio, however, demonstrate the following: (a) During the initial compression phase ahead of the sampler, soil elements suffer very small strain (up to maximum 0.32%). (b) During the extension phase in the vicinity of the cutting shoe, soil elements are subjected to considerable strains (up to maximum -5%). (c) During the second compression phase inside the sampler tube, soil elements are again subjected to significant strains (up to 2.6%). From the strain paths presented in Figs. 3.33 - 3.35, it can be concluded that variation in the inside clearance ratio of a sampler affects all the three phases ofzyxwvutsrqpon 88 strains, especially the extension phase in the vicinity of the sampler and the second compression phase inside the sampler. For all the samplers the minimum peak axial strains in compression and extension (which occur at the centreline of the sampler) in Fig. 3.36. were determined by extrapolating the curves shownzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPO 3.5.3 ANALYSES WITH DIFFERENT INSIDE AND OUTSIDE CUTTING EDGE TAPER ANGLES 3.5.3.1 DIMENSIONS AND CHARACTERISTICS OF THE SAMPLERS Two samplers with inside cutting edge taper angles of 0.358° and 1.432° were analysed. Outside cutting edge taper angles of these samplers were the same and were equal to 9.9°. Another two samplers with outside cutting edge taper angles of 5° and 19.29° were analysed. The inside cutting edge taper angles of these two samplers were kept fixed and their values were 0.716°. The other dimensions and characteristics of all the four samplers studied were as follows: External diameter of the sampler tube, B = 57.5 mm Internal diameter of the sampler tube, D. = 51.0 mm Internal diameter at cutting shoe, D, = 50.5 mm Thickness of the sampler tube, t = 3.25 mm Length of the sampler tube, L = 120 mm Inside clearance ratio = 29.64%zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA Bit ratio = 17.69 Cutting shoe geometries and dimensions of the samplers with different inside and outside cutting edge taper angles are shown in Figs. 3.37 and 3.38 respectively. 3.5.3.2 FINITE ELEMENT MODEL AND BOUNDARY CONDITIONS For all analyses a finite element mesh shown in Fig. 3.24 consisting of 2156 QXF9 elements was used. The boundary BC was fixed at a distance of 240 mm from the top of the sampler while the boundary DC was located at a distance of 3.89R from the centreline of the sampler. In order to model piston sampling, analyses were carried out with different values of cl» along the boundary AB for each sampler. The 89 boundary conditions along the boundaries AB. DC. EFG. GHI. BC and ID for all analyses were similar to those reported previously (see section 3.4.2). Table 3.5 shows a listing of the different cj)-valuesassigned along the boundary AE and the corresponding maximum total axial deformations along the boundary AE for all the analyses. Variation of total axial deformation along the boundary AE for the analyses modelling piston sampling is shown in Fig. 3.39. 3.5.3.3 STRAIN PATHS OF SOIL ELEMENTS The strain paths at different locations within the sampler tube for the samplers studied are shown in Figs. 3.40 - 3.43. From the strain paths shown in Fig. 3.40 it is evident that for a sampler with small inside cutting edge taper angle. the peak axial strains in extension are greater than those in compression. This is true for all the strain paths. However. for the sampler with a large inside cutting edge taper angle, the peak axial strains in compression are greater than those in extension for the strain paths located at 10-70% of cutting shoe radius from the centreline (see Fig. 3.41). For soil elements passing near the inside edge (O.9~ from the centreline) of the sampler, the peak strain in compression is slightly less than that in extension whereas for soil elements moving along the inside edge of the sampler. the peak axial strain in extension is considerably higher (approximately twice) than that in compression. Fig. 3.42 indicates that for a sampler of small outside cutting edge taper angle, the peak axial strains in extension are significantly higher than that in compression for all the strain paths. However, from Fig. 3.43zyxwvutsrqponmlkjihgfedcbaZ it is evident that for the sampler with large outside edge taper angle, the peak axial strain in compression are greater than that in extension for all the strain paths e~cept for the one where the soil elements move along the inside edge of the sampler. For this strain path, however, the peak axial strain in compression is slightly less than that in extension. For all the samplers the minimum peak axial strains in compression and extension (which occur at the centreline of the sampler) were computed by extrapolating the curves shown in Fig. 3.44. 90 3.6 STUDY OF FLAT·ENDED SAMPLERS Four samplers have been analysed under this category. The dimensions and characteristics of all the samplers are presented in Table 3.6. A schematic diagram of the mesh used for analysing samplers I and TI is shown in Fig. 3.45. The mesh consisted of 2303 QXF9 elements. Total number of nodes in the mesh was 9555. Figs. 3.46 and 3.47 show the schematic diagrams of the meshes used for analysing samplers TIl and IV respectively. The meshes shown in Figs. 3.46 and 3.47 consisted of 2352 QXF9 and 2520 QXF9 elements respectively. respective total number of nodes in the meshes were 9753 and 10405. The All the meshes (Figs. 3.45 - 3.47) were refined near the inside and outside edges of the sampler and also to some extent near the bottom of the sampler. In all the meshes the boundary BC was fixed at a distance of twice the length of the sampler tube from the top of the sampler. The locations of the boundary DC are shown in Table In order to simulate piston sampling, analyses were carried out with different 3.7. e-values along the boundary AE for each sampler. For all analyses boundary conditions for boundaries AB, CD, EF, FO, OH, BC and HO were unchanged and these are: (i)zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA aCl>/anzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA = 0 for the boundaries AB, CD, EF, FO and OH; (ii) (iii) Cl> Cl> = = 1000 for the boundary BC; and; 0 for the boundary HO. The different values of Cl> assigned for the boundary AE and the respective maximum total axial deformation along the boundary AE for each sampler are shown in Table 3.7. Variation of total axial deformation along the boundary AE for the analyses modelling piston sampling is shown in Fig. 3.48. The strain paths at various locations within the sampler tube are presented in Figs. 3.49 - 3.52. The characteristics of the strain paths are discussed in chapter 6 and a comparison of these samplers is also presented in the same chapter. For all the samplers the minimum peak axial strain in compression (at the centreline of the samplers) were determined by extrapolating the curves shown in Fig. 3.53. 91 Table 3.1 Characteristics of NOI, SOl and U 100 Samplers A.R. I.C.R.zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHG BIt Reference t B D.zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA Sampler (mm) (%) (%) 54 1.25 11.4 0.93 45.6 VoId, 1956 50 4.9 44.0 0.4 12.2 Kallstenius, (mm) (mm) (mm) NOI 54.5 57 SOl 50.2 60 type ratio 1961 UIDO 105.5 117.25 104 4.5 27.1 1.44 26.1 Clayton et al, 1982 (TypeI) UIDO 105.7 117.5 104.5 4.5 26.4 1.1 26.1 Clayton et al, 1982 (TypelI) Note: D. = Internal diameter of the sampler tube B = External diameter of the sampler tube D, = t = Thickness of the sampler tube A.R. = Area ratio I.C.R. = Inside clearance ratio Internal diameter at cutting shoe 92 Table 3.2 Boundary Conditions for analysing NOI, Sal and U100 samplers <J>-value Max. total along axial (from the the deformation centreline boundary along the of sampler) AE boundary AE Sampler Location of type Boundary DC Analysis No. (mm) o -0.85 1zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA NOI Sal UIOO 3.70R 3.96R 3.59R (Type I) UIOO 3.55R (Type I!) Note: 2 10 1.367 3 4.2934 0.099 4 3.8434 -0.012 1 o -2.59 2 12.25 1.27 3 12.18 0.104 4 11.7297 0.014 1 o 2 17.2388 2.62 3 10.3863 0.051 1 o 2 11.517 R = Internal radius of the sampler tube 93 -3.81 -4.26 0.098 Table 3.3 Boundary conditions for analysing samplers of different area ratios Area Analysis ratio No. (%) e-value Max. Total along the axial boundary AE deformation along the boundary AE (mm)zyxwvutsrqponmlkjih 10.14 1 o -0.66 2 2.743 -0.063 3 3.03 0.01 1 o -2.11 29.64zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA 2 0.22 10.5 50.73 100.46 3 9.502 -0.0107 1 o -3.63 2 16.91 0.383 3 15.187 0.02 1 o 2 26.506 0.58 3 25.0543 -0.023 -5.33 94 Table 3.4 Boundary conditions for analysing samplers of different inside clearance ratios Inside Location of clearance boundary DC ratio (from the (%) Analysis No. cjl-value Max. total along the axial boundary AE deformationzyxwvutsrqponm centreline along the of sampler boundary AE (mm) 0.495 1.98 3.96 3.91RzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA -2.187 1 o 3.85R 3.78R 0.017 2 9.8656 1 o -1.96 2 8.8606 -0.02 1 o -1.76 2 7.938 95 0.04 Table 3.5 Boundary conditions for analysing samplers of different inside and outside cutting edge taper angles AnalysiszyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFED <j>-value Max. total Sampler No. along the axial boundary AE deformation along the boundary AE (mm) Inside 0.358 cutting edge -2.12 1 0 2 9.5655 1 0 2 9.4745 1 0 2 8.4276 1 0 -2.25 2 10.1254 -0.02 0.015 taper angle (degree) 1.432 Outside 5.000 cutting edge -2.1 0.013 -1.86 0.02 taper angle (degree) 19.290 96 Table 3.6 Sampler Dimensions and characteristics D. B of flat-ended samplers L t A.R. BIt (mm) (mm)zyxwvutsrqponmlkjihgfedcbaZYXWVUTS (mm) (mm) (%)zyxwvutsrqponmlkjihgfedcbaZYXWVUT ratio No.zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA I 54.5 57.0 1.25 120 9.38 45.6 II 52.5 57.5 2.50 120 19.95 23.0 III 50.2 60.0 4.90 120 42.85 12.2 IV 105.7 117.5 5.90 204 23.57 19.9 Note:' L = length of the sampler tube 97 Table 3.7 Sampler No. Boundary conditions for analysing various flat-ended samplers Location of boundary DC Analysis cl> Value ofzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA Max. total No. (from the along the axial boundary AE deformation centreline along the of sampler) boundary AE (mm) I II III IV 3.70R 3.85R 4.02R 3.55R 1 o 2 7.3353 0.72 3 4.1977 0.03 1 o 2 8.225 1 0 -3.58 2 12.24 -0.93 3 16.1965 0.17 4 16.4 0.22 1 o 2 12.187 98 -0.93 -1.8 0.08 -4.5 0.19 Boundary condition type (A)zyxwvutsrqponml <I> zyxwvutsrqponmlkjih = <I>p Surface, SA L x Region,zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA A Surface, Ss Boundary condition type (B) 0<1> K;:c-L,.+ ox K"-iJ iJ<I> y L.,+ q + a(<I> - <1>..)=0 Fig. 3.1 Two dimensional region with permissible boundary conditions for quasi-harmonic problems (after Hinton and Owen, 1979) x Fig. 3.2 Flow through a porous medium showing equipotential and flow lines (after Hinton and Owen, 1979) 99 1zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA Fig. 3.3 Four-noded quadrilateral axisymmetric field elementzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHG rzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA 25S mm ~ ~ 255 oW ~ ,=0 A EI O---} .... I'-- ~1.5 ---4 Oc!I/iln = 0 ~/iln mm = 0 280 mm Oc!IJ()n-O n=O I.zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA Cenb'e hne H E of sam pier j 20mm I (j ISO mm C • ,. 1 ,.z..~__ B k25.25 mm">l< 27.25 mm ~ Fig. 3.4 Finite element mesh with 56 QXF4 elements (not to scale) 100 6 ~-------------------------------------------------------------------------------, 5 A E --------10th STREAMLINE ,.... 4 zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA CD <, N .... z o ...... ~ 3 < U o ....J 2 ~ ZzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA W :E:zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA W ....J 8 W ....J < ~ o D C U ffi > -1 -2 -30L----------.....J.5------------~lO--------------....J1~5------------~2~O------------~2=5~--------~30 DISTANCE FROM THE CENTRELINE OF SAMPLER (mm) Fig. 3.5 Position of 10th streamline 6 . 5 " CD "N S.P. AT 10% OF C.S.R. FROM C.L. I- ................ S.P. AT 50% OF C.S.R. FROM C.L. -------- S.P. AT 90% OF C.S.R• FROM C.L. 4 f- .... z 0 ...... ~ . EXTENSION 3 S.P. • STRAIN PATH C.S.R. • CUTTING SHOE RADIUS C.L. • CENTRELINE OF SAMPLER < u 0 ....J 2 ~ Z W c f- :E: W _J W ....J < u b 0 ...... ~ ~ __I . ~ w -1 f>zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA a COMPRESSION -2 -3 -4 -3 I I -2 -1 I o 2 3 AXIAL STRAIN (X) Fig. 3.6 Strain paths of soil elements at three different locations 101 4 E I D AzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA I- n ", Centreline of sampler c B Fig. 3.7 Finite element mesh withzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFED 400 QXF4 elements (not to scale)zyxwvutsrqponmlkjihgfedc 102 3 S.P. AT 10% OF C.S.R. FROM C.L.zyxwvutsrqponmlkjihgfedcb ................ S.P• AT 30% OF C.S.R. FROM C.L. 2 ,..., --S.P. AT 50% OF C.S.R. FROM C.L. CDzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA <, EXTENSION N ~ AT 70% AT C.S.R. FROM C.L. -·-S.P. Z 0 -------- ...... I- < U 0 _J l- z D -----~ ·,,::;:- _____ ""u -, -- - -- S.P. AT 90% OF C.S.R. FROM C.L. , , ----- zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA ... ~ LU :::E LU ~: _J LU _J < u ...... -1 I- COMPRESSION a::: LU >zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA -2 r -3 -2.0 . -1. 5 -1. 0 -.5 O. 0 .5 1.0 1.5 2. 0 AXI AL STRA INzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJ (Z) Fig. 3.8 Strain paths of soil elements for the analysis with 400 QXF4 elements 1 Fig. 3.9 Nine-noded quadrilateral axisymmetric field element 103 IJ) IJ) IJ) I-- I-- Z Z LaJ Z UJ X LaJzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA X LaJ X UJ _J _J LaJ _J LaJ UJ LaJ I-- -.t' -.t' U. >< u, C3 Cl CD • U. >< C3 >< C3 to to If) If) "¢ .... .... N N CD <, N '-' ~ Z Cl _., ~ -< w .,_ - w 1+-0 .,_ Cl _J (T) I I I I I I I I I I I I IJ) IJ) IJ) I-- I-- I-- LaJ ::E LaJ Z LaJ :E UJ Z LaJ z _J _J LaJ UJ • U. en U. >< C3 C3 to to If) If) >< I X LaJ I _J LaJ -.t' u. >< Cl Cl Cl • z :::E W _J w _J < w _., N 0 -a ,....,. en til til ~ c<:I c<:I 1-0 1-0 g 0 0 ~ 0 ~ 0 .~ ell P. 8 I-- (3 0:: W 0 :> - I M I ~ bb ....... I I I I ( j zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA I I a U1 a U1 a (T) N N ....... ....... . . . GO I 3dAl ~O~~3 104 U1 a a azyxwvutsrqponmlkjihgfedcb . 7.5 mm 57.5 mm 2.5zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA mm 5° taper taper taper (a) (b) Fig. 3.11 Cutting shoe designs of two typical piston samplers: (a) NOI sampler (after Vold, 1956) (b) SOl sampler (after Kallstenius, 1961) 105 25.5zyxwvutsrqpon mm 34 mm 6.5 mm 31 mm 31.5 mm 11.25 mm 7.5 mm 6mm (a) (b) Fig. 3.12 Cutting shoe designs of two typical British Standard General Purpose open-drive samplers: (a) Type I (b) Type II (after Clayton et al, 1982) 106 A_El _D Centreline of samplerzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA c B Fig. 3.13 Finite element mesh with 2254 QXF9 elements for analysing NO! sampler (not to scale) 107 E I D AzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA . Centreline of sampIer c B Fig. 3.14 Finite element mesh with 2254 QXF9 elements for analysing SOl sampler (not to scale) 108 E I A D , I Centre line of sampIer zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA c B Fig. 3.15 Finite element mesh with 2550 QXF9 elements for analysing Ul00 (Type I) sampler (not to scale) 109 A El D , Centreline ofzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA sampler BzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA c Fig. 3.16 Finite element mesh with 2436 QXF9 elements for analysing UlDO (Type II) sampler (not to scale) 110 r-------------------------~r_------------------------~~zyxwvutsrqponmlkjihg zyxwvutsrqpo in _ Cl -- tn - ~ ~ ,..... E E _ Cl ~ ~ W _J a... :E -e lJ1 lJ... Cl _ Cl -I- (T) W Z ....... _J W ~ I~ I.LI _J n, :E -c ~ _ 111 W N U a, w :c :E -e 1\ lJ1 - ..... I- (J') ~ ~ ....... ....... I.LI WJ I.LI I.LI a, o, :E :E >- >- lJ1 tJ') _J _J n, I- Z I.LI _J _ Cl \ -I- -e <: ................ ~ Cl ~ lJ... \ n, I- N :-1- _111 ~ W U Z < IlJ1 ....... _ Cl ~ ) -111 ~--~I~--~I-----~I----L-I----~--~I~--~I-----~.-----~.-----Cl 00 ID ~ N Cl N ~ ID 00 r-1 Cl Cl Cl Cl Cl Cl Cl Cl Cl Cl . . Cl (WW) I . I . I 3V A~VONn08 ~NOlV NOI1VW~O~30 lVIXV lV10l 111 . I Cl . I Cl 2.5 2.0zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA S. P. AT lOr. OF C. S. R. FROM C. L. --S.P, AT 30% OF C, S.R, FROM C, L.zyxwvutsrqponmlkjihg ,... CD <, EXTENSION ~ ::z 1.0 0 ..... I- < u S. P. AT sor. OF -------- 1.5 c. S. R. FROM C. L. 70r. OF C.S.R. FROM C. L. ATzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPON -·-S.P. ................ S. P. AT 90% OF C. S. R• FROM C. L. ---- S,P, AT lNS10E EDGE OF SAMPLER .5 0 _J I- ::z w ::E w _J w 0.0 -.5 _J < u ..... -1. 0 COMPRESSION I0:: w > -1. 5 -2.0 -2.5 -2.0 -1. 5 -1. 0 -.5 AXIAL STRAIN Fig. 3.18 .5 0.0 1.0 1.5 2.0 (iD Strain paths of soil elements due to penetration of NOl sampler 2.5 2.0 ,..... CD <, N ....., ::z ..... l- 1.5 LEGENDI SAME AS ABOVE EXTENSION 1.0 0 < u .5 0 _J I- ::z w ::E w _J w 0.0 -.5 _J < u ..... I- -1. 0 COMPRESSION 0:: W > -1. 5 -2.0 -2.5 -2.0 -1. 5 -1.0 -.5 AXIAL STRAIN Fig. 3.19 .5 0.0 1.0 1.5 2.0 00 Strain paths of soil elements due to penetration of SOl sampler 112 2.5 2.0zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA ,.... cc <, .!j --- S.P. --- S. P. AT 30% OF C. S. R. FROM C. L. AT lOX OF C.S.R. -------S.P. AT 501. OF C.S.R. 1.5zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA ___ . S.P. AT 701. OF C.S.R. EXTENSION z 1.0 CJ l- < u FROM C.L. FROM C.L. FROM C.L. ................ S. P. AT 90X OF C. S. R. FROM C. L. ---- S.P. AT INSIDE EDGE OF SAMPLER .5 CJ _J IZ W :::E w _J w 0.0 -.5 _J - < u -1.0 I- a::: w > COMPRESSION -1.5 -2.0 -2.5 -3 o -1 -2 AXIAL STRAIN 2 3 (7.) Fig. 3.20 Strain paths of soil elements due to penetration of UIOO (Type I) sampler 2.5 2.0 ,.... cc 1.5 <, N ...., z EXTENSION LEGEND I SAME AS ABOVE 1.0 CJ I- < u .5 CJ _J IZ W :::E W _J w 0.0 -.5 _J < u -1. 0 I- COMPRESSION a::: > -1.5 L1.J -2.0 -2.5 -3 -2 o -1 AXIAL STRAIN 2 3 (7.) Fig. 3.21 Strain paths of soil elements due to penetration of UIOO (Type IT) sampler 113 3.0 +--+ " N ..., NGI SAMPLERzyxwvutsrqponmlkjihgfedcbaZYXWVU *_* SGI SAMPLER 2.5zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPON 0---0 UIOO (TYPE 1) SAMPLER zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIH ...... le--le UIOO (TYPE II) SAMPLER UJ 2.0 et: Z CJ t/) t/) /)( n, :E CJ w z ......zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA z ...... < et: le---- I- le t/) ..J < ...... se < 1.0 ___ ___ UJ n, .--- _+--- + + + • •__ * .5 0----0 0 o c ~ < .i.->" 1.5 0.00l_-_..JIO---2~0-----3~0-----4LO----~50~---6=0~--~7~0~--~8=0----~g~0--~100 LOCATION FROM CENTRELINE OF SAMPLER (~ OF CUTTING SHOE RADIUS) (0) 3.0 ~-----------------------------------------------------. LEGEND 2.5 I SAME AS ASOVE z CJ )(~ t/) Z UJ lX UJ Z ..... z ..... -> zyxwvutsrqponmlkjihgfe 2.0 1.5 ___»>: 1----)(1_----- < / /" 0/ et: lt/) ..J < ...... x < ~ 1.0 O~ ~-~.-------c---------- < UJ n, ~--+,-----------+-------~zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFED 0.0 ~ __ o _..J~ 10 __ ~ ~ 20 ~ 30 ~ 40 50 ~ ~ 60 70 L_ 80 __ ~~ __ so ~ 100 LOCATION FROM CENTRELINE OF SAMPLER (~ OF CUTTING SHOE RADIUS) (b) Fig. 3.22 Peak axial strains at various locations for NO!, SOl and UlOO samplers: (a) in compression (b) in extension 114 A.R. A.R. A.R. = 100.46% = 50.73% = 29.64% A.R.zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA = 10.14% e e ~ 0 0 ('I ('I El e V"I taper f'\ e- Fig. 3.23 Cutting shoe designs for samplers of different area ratios 115 , I zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA Centreline of sampler c 8 Fig. 3.24 Finite element mesh with 2156 QXF9 elements for analysing samplers of different area ratios (not to scale) 116 ~ -r ~Cl zyxwvutsrqponmlkjihgfe ("f") ,, ,, ,, ,, ,, , 111 N ,..... E E - ,, , et: ,,~'. lJ.J ,,'.~ _J a... ,~zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA~-e Cl N ,: :, , : , ,, : U") I.L.. Cl lJ.J Z \ ..... _J lJ.J 111 ..... et: I- Z lJ.J U N . en. m r-- . c:5 Cl Cl Ul __, II II II Cl -. Cl -. Cl -. Cl I- I- I- I- 0::: 0::: Cl __, N II -e < < 0::: 0::: -e < -e < -e IJ.J 0::: IJ.J 0::: UJ 0::: -e lJ.J ::r: I I- I I I I I ~ Q ,, Cl ..... ,, ,, , ,,, , -. < < IJ.J 0::: lJ.J U Z < IU') ..... ,, < et: I.L.. Cl Ul ("f") N ..... Cl Cl Cl (WW) __, Cl Cl Cl Cl I . . N Cl . ("f") . I I IzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONM 3V A~VONn08 ~NOlV NOI1VW~O~30 lVIXV lV10l 117 . Clzyxwvutsrqponmlkjihgfedcba 2.5 S.P. 2.0zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA ,.. CD -~~~~~~--- 1.5 ...... z -·-S.P. ................ EXTENSION N ...., 1.0 ---- 0 l- < u S. P. AT 30% OF S. P. AT 50% OF AT 70% OF S. P. AT 90% OF S.P. AT INSIDE L. FROM C.zyxwvutsrqponmlkji C. S. R. FROM C. L. C. S. R. FROM C. L. C.S.R. FROM C. L. C. S. R• FROM C. L. EDGE OF SAMPLER .5 _- Cl ...J I- 0.0 w __j w -.5 z w ~ AT )0% OF C.S.R. __j < u ........ -1.0 COMPRESSION I0::: w > -1. 5 -2.0 -2.5 -2.5 -2.0 -1. 5 -1.0 -.5 AXIAL 0.0 1.0 .5 1.5 2.0 2.5 CiD STRAINzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA Fig. 3.26 Strain paths due to penetration of sampler of area ratio 10.14% 2.5 r----------------------------,,----------------------------. 2.0 ,.. CD 1.5 <, N z LEGEND : SAME AS' ABOVE EXTENSION ....... 1.0 Cl ........ I- < U .5 Cl ...J IZ W ~ W __j W -.5 ...J i5 -1.0 COMPRESSION I0::: W > -1.5 -2.0 -2.5 ~----~----~----~----~----~----~----~----~----~--~ 1.0 .5 0.0 -.5 -2. 5 -2. 0 -1. 5 -1. 0 AXIAL STRAIN 1.5 2.0 GO Fig. 3.27 Strain paths due to penetration of sampler of area ratio 29.64% 118 . 2.5 ~-------------------------r-------------------------' 2.5 --S.P. AT ------- S.P. AT -------- S.P. AT ----.---- S.P. AT ................ S.P. AT ---S.P. AT 2.0 1.5 EXTENSION 1.0 10% OF 30% OF 50% OF 70% OF 90% OF INSIDE C.S.R. FROM C.L. C.S.R. FROM C.L. C.S.R. FROM C.L. C.S.R. FROM C.L. C. S.R. FROM C.L. EDGE OF SAMPLERzyxwvutsrqponmlk zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA _J < w -1. 0 COMPRESSION I- 0::: L1J > -1. 5 -2.0 -2.5 L-__ ~ -2. 5 -2. 0 ~ ~ -1. 5 -1. 0 ~ -. 5 ~ __ ~ O. 0 AXIAL STRAIN .5 ~ ~ 1. 0 1. 5 ~ __ ~ 2. 0 2. 5 00 Fig. 3.28 Strain paths due to penetration of sampler of area ratio 50.73% 2.5 ~------------------------.---------------------------. 2.0 ,.... 1. 5 CD <, N ....., LEGEND : SAME AS ABOVE EXTENSION 1.0 COMPRESSION -2.0 -2.5 ~---~--~--~~-~~---~--~-~~-~~--~~-~ -2. 5 -2. 0 -1. 5 -1. 0 -.5 0.0 .5 1.0 1.5 2.0 AXI AL STRA I N GO Fig. 3.29 Strain paths due to penetration of sampler of area ratio 100.46% 119 2.5 2.50 2.25zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA ~. AREA RATIO - 10.14% ,...zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGF 0------0 AREA RATIO • 29.64% N ._, z ...... 2.00 +------+ AREA RATIO· 50.73% x-_x AREA RATIO· 100.46% Cl 1Il 1Il LIJ 0:: CL % Cl u z ...... z ...... < 0:: I1Il 1. 75 1. 50 1. 25 1. 00 ....J < ...... x .75 ~ .50 < < LLJ CL .25 0.00 0 10 20 30 40 50 LOCATION FROM CENTRELINE OF SAMPLER 60 (1. 70 80 90 100 OF CUTTING SHOE RADIUS) (a) 2.50 LEGENDzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIH I SAME AS ABOVE 2.25 ,.... N ._, 2.00 z Cl ...... 1Il 1. 75 z LaJ lX LaJ .... ....z< 1. 50 Z 0:: l1Il 1. 25 I. 00 ....J < .... x .75 < ~ < LaJ CL .50 .25 0.00 0 10 20 30 40 SO 60 70 BD 90 LOCATION FROM CENTRELINE OF SAMPLER (X OF CUTTING SHOE RADIUS) (b) Fig. 3.30 Peak axial strains at various locations for samplers of different area ratios: (a) in compression (b) in extension 120 100 ICR = 3.96% ICR:: 1.98% ~ o 00 ICRzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA = 0.495% e e e ~ o N e Fig. 3.31 Cutting shoe designs for samplers of different inside clearance ratios 121 .04zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA ,..., E zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA ezyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA INSIDE CLEARANCE RATIO· 0.495% >oJ .03 lLJ < >~ < 0 .02 z INSIDE CLEARANCE RATIO· ].geO% -------- INSIDE CLEARANCE RATIO· 3.960% =:J 0 CD L:l .. •......... ' .01 z ........... / •.••... 0 _J < z 0.00 .... 0 I- < ~ ~ ,.. -.-.-.:-...... -.01 0 U. lLJ 0 _J < .... x -.02 < _J < -.03 I0 I- -.04 20 15 10 5 0 25 DISTANCE FROM THE CENTRELINE OF SAMPLER 30 (mm) F·Ig. 3 .32 Variation of total axial deformation along boundary AB for samplers of various inside clearance ratios 2.5 a 2.0 --- s. P. --- S.P. AT 30% OF C.S.R. FROM C.L. 1.5 -------- S.P. AT 50% OF C.S.R. FROM C.L. AT 70%·OF C.S.R. FROM C.L. <, EXTENSION N ._, z 1.0 .... 0 I- < w AT 10% OF C. S. R. FROM C. L. ----.---- S.P. ................ S. P. AT 90Y. OF C. S. R. FROM C. L. - - - - S. P. AT I NSJOE EDGE OF SAMPLER .5 0 _J IZ 0.0 lLJ ~ lLJ _J lLJ -.5 _J < w -1. 0 .... COMPRESSION I- ~ lLJ > -1. 5 -2.0 -2.5 -2.0 -1. 5 -1. 0 -.5 0.0 .5 1.0 1.5 2.0 AXIAL STRAINzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPON (X) Fig. 3.33 Strain paths due to penetration of sampler of inside clearance ratio 0.495% 122 2.5 2.0zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA ,.... 1.5 zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFED CD <, N ._, z ...... EXTENSION 1.0 Cl I- < u .5 Cl _J IZ LI.J :::E: LI.J _J LI.J 0.0 S.P. AT 10% OF --S.P. AT 30% OF -------- S.P. AT 50% OF -·-S.P. AT 701.OF ................ S.P. AT 90% OF ---S.P. AT INSIDE -.5 _J < u ...... -1. 0 I0::: LI.J > -1. 5 C.S.R. FROM C.L. C.S.R. FROM C.L. C.S.R. FROM C.L. C.S.R. FROM C.L. C.S.R• FROM C.L. EDGE OF SAMPLER COMPRESSION -2.0 -2.5 -4 -3 -2 -1 AXIAL 0 STRAIN 2 3 4 (7.) Fig. 3.34 Strain paths due to penetration of sampler of inside clearance ratio 1.98% 2.5 2.0 ,.... CD 1.5 <, N ._, EXTENSION z 1.0 -e .5 Cl ..... I- u Cl _J IZ LI.J :::E: LLJ _J l.LJ 0.0 -.5 _J < ...... -1. 0 u I0::: l.LJ LEGEND I SAME AS ABOVE COMPRESSION > -1. 5 -2.0 -2.5 -6.0 -4.5 -3.0 -1. 5 AXIAL 0.0 STRAIN 1.5 3.0 4.5 6.0 (7.) Fig. 3.35 Strain paths due to penetration of sampler of inside clearance ratio 4.96% 123 3.0 + + + .....zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA tot 'oJ Z c:J ..... til til LLI 0:: • • 2.0 0... ::::E c:J + + 2. S INSIDE CLEARANCE RATIO· 0.495% 0------0 INSIDE CLEARANCE RATIO· +------+ INSIDE CLEARANCE RATIO· 3.96% (2nd compression phose) lC--II INSIDE CLEARANCE RATIO· 3.96% (1st compression phose)zyxwvutsrqponmlkjihg 1.98% U z z ..... I.S < til _J < ..... 1.0 __ ----.----* -> _- 0:: ~ r x < ~ < LLI 0... .5 _o ___ - J"I. o~ - II 0.0 IC II le I I 70 so 60 30 20 40 10 ozyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA 80 100 90 LOCATION FROM CENTRELINE OF SAMPLER (% OF CUTTING SHOE RADIUS) (a) 6 ~--------------------------------~---------------------, ------* 5 ""'0 --0 Z LLI ~ INSIDE CLEARANCE RATIO - 1.98% +------+ INSIDE CLEARANCE RATIO· z c:J ..... til INSIDE CLEARANCE RATIO - 0.495% ----------.> ._/ 4 X L&J Z ..... z ..... / 3.96% 31---·+----+----zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGF < 0:: ~ til 0____ _J < ..... x < ~ o D 2 ~----~-----------~----------~----------- < L&J 0... 1 ___ / ~----------------------------------~--------~---r o~ ~ o 10 l~ 20 ~ ~ '~ ~ 3D 40 SO 60 '~ 70 __ ~~ __ BD ~i~ __ 90 ~ 100 LOCATION FROM CENRTELINE OF SAMPLER (% OF CUTTING SHOE RADIUS) (b) Fig. 3.36 Peak axial strains at various locations for samplers of different inside clearance ratios: (a) in compression (b) in extension 124 Fig. 3.37 Cutting shoe designs for samplers of different inside cutting edge taper angles 125 5° taper ~ o - Fig. 3.38 Cutting shoe designs for samplers of different outside cutting edge taper angles 126zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIH ~----------------------~~----------------------~~zyxwvutsrqponmlkjihgfed m 4) '~ in N '. \, "' ....'. \ . . . t3 t.!) t.!) UJ Cl UJ Cl t.!) UJ Cl lJ.J Cl en N \ \ If) ('T) ('T) Cl ~ II II UJ Cl UJ (J') (J') ~ \ Cl II II lJ.J Cl lJ.J Cl ~ (J') (J') l.!) t.!) Z Z \ ---t.!) l.!) Z Z ~ u, Cl u, 0 lJ.J lJ.J U .....J U U .....J t.!) t.!) Z Z lJ.J .....J Z Z < < UJ UJ l.!) lJ.J t.!) t.!) t.!) t.!) Cl lJ.J Cl lJ.J Cl lJJ Cl UJ lJ.J Cl lJ.J \ ,, , , Cl ..... ..... (J') (J') Z Z ..... ..... ,, , ,, ,, , ._ ._ (J') ~ 0 (J') ~ 0 ,, N Cl ..... Cl ..... o Cl Cl ,• d (WW) ..... ~~ ~a ~t t:: 0.. u, .J:lO SS UJ ~i UJ 0 ..... ..... ::s .,_ ~ Z U UJ :z:: .,_ 2: Cl ~ u, UJ u Z < .,_ (J) ..... Cl tn ~ Co) E-8 0· .... c,.;;;fl .gS -a~ ..... t:: ~= --.5 -a.g..... o 0 ..... c:: t:: '5~ >"S 0\ oh Cl ,• ("I") ,• Cl "lit Cl ,• e .g~ it N rn Co!-< ~ ~ ,/ 3V A~VONn08 ~N01V NOrlVH~O~30 lVrXV lV10l 127 '"' ~ -e (J) UJ ,) \/ , ("I") Cl Ln I, \, . . m 0 ..... -abJ) t:: _J C::'Jj Cl ..... ,, ) , ,zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA \ ed Z \, . \ lJJ Cl Cl. 2: Cl \ -- lJJ Cl N \, \, \, , , < < lJ.J , , ,,, ,, ,, , Cl 0 \, \, , ,, Cl t.!) , I u, u, .....J _J \ U lJ.J UJ \ ._ ._ ._ ._ ._ ._ ~._ ._ ~ ~ ~ , 0 ~ J: E I I I I I I If) ,.... - ,, ,, ,, , N 0.. E \ . . en ..... . . ..... CC -E '"'zyxwvutsrqpo 2.5 S. P. AT )0% OF C.S.R. FROM C.zyxwvutsrqponmlk L. 2.0 S.P. AT 30% OF C.S.R. FROM C. L. --,... ee -------. S. P. AT 50% OF C.S.R. FROM C. L.zyxwvutsrqponml -·-S.P. AT 70% OF C.S.R. FROM C. L. ................ S. P. AT 901. OF C.S.R • FROM C.L. ---S.P • AT lNSIOE EDGE DF SAMPLER 1.5 <, N EXTENS10N '-J z 1.0 Cl ...... ..... -e u .5 Cl ...J ..... 0.0 W ::::E W ...J -.5 z w ...J < u ...... -1. 0 ..... COMPRESS1ON ~ w > -1. 5 -2.0 -2.5 -2.5 -2.0 -1. 5 -1. 0 -.5 AXIAL Fig. 3.40 0.0 .5 1.0 1.5 2.0 2.5 (X) STRAINzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJI Strain paths due to penetration of sampler of inside cutting edge taper angle 0.358° 2.5 2.0 ,... ee 1.S <, N EXTENSlON LEGEND : SAME AS ABOVE '-J Z Cl 1.0 ...... ..... < u .5 ..... 0.0 Cl ...J z --- W ::::E w ...J w -.5 ...J < u ...... -1. 0 ..... ~ w > -1.5 COMPRESS10N -2.0 AX 1AL STRA 1NzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA (X) Fig. 3.41 Strain paths due to penetration of sampler of inside cutting edge taper angle 1.432° 128 2.5 r-----------------------------~--------------------------~ --- 2.0 @ <, !:! z S. P. AT lOX OF C. S. R. FROM C. L. --S. P. AT 301. OF C. S. R. FROM C. L. -------S.P. AT 501. OF C.S.R. FROM C.L. 1.5 zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFED _._ S. P. AT 701. OF C. S. R. FROM C. L. EXTENSIONzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA . S. P. AT 901. OF C. S. R. FROM C. L. 1.0 ClzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA ...... l-c .5 ---- S. P. AT INSIDE EDGE OF SAMPLER Ll Cl _J IZ W ~ W _J W -.5 _J t5 ..... -1.0 ~ -1. 5 COMPRESSION I0::: -2.0 -2.5 ~----~----~----~----~----~----~----~----~----~--~ -2.5 -2.0 -1.5 -1.0 -.5 0.0 AXIAL . P·Ig. 342 .5 1.0 1.5 2.0 2.5 STRAINzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFED (1.) Strain paths due to penetration of sampler of outside cutting edge taper angle 5° 2.5 ------------------------------~--------------------------~ 2.0 @ 1.5 LEGEND : SAME AS ABOVE EXTENSIONzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA <, N -..; :z 1.0 ..... Cl l- -e w .5 Cl _J I- :z w ~ w ....J W -.5 _J t5 ..... -1.0 COMPRESSION l- a::: g; -1. 5 -2.0 -2.5 ~ __ ~ -2.5 -2.0 ~ -1.5 ~ -1.0 J_ -.5 AXIAL L_ __ __J~ 0.0 STRAIN .5 __ ~-- 1.0 __ ~ 1.5 ~ __ ~ 2.0 (1.) Pig. 3.43 Strain paths due to penetration of sampler of outside cutting edge taper angle 19.29° 129 2.5 2.50 0------0 INSIDE CUTTING EDGE TAPER ANGLE· 2.25zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA ,..... N --z 0.358 DEG. +------+ INSIDE CUTTING EDGE TAPER ANGLE • 1.432 DEG. zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA 2.00 *---* OUTSIDE CUTTING EDGE TAPER ANGLE· 5.0 DEG. 0 ...... U) U) _--le 1. 7S OUTSIDE CUTTING EDGE TAPER ANGLE· / 19.29 DEG. x UJ ~ n, :x S z ...... 1. 25 z ...... < ~ 1. 00 ~ U) __J < ...... x .75 < .50 < ::s:: _X----x------- ----- 1.50 l- UJ e, .25 0.00 x/ ~----x----------- --..-----II-----~ r----.----------.--------~ ~--~---------------------.----------_..----- ~~ _----I 0 20 10 30 40 SO 60 70 80 90 100 LOCATION FROM CENTRELINE OF SAMPLER (7. OF CUTTING SHOE RADIUS) (c) 2.50 LEGEND 2.25 .-. N --z I SAME AS ABOVE 2.00 Cl ...... 1. 75 z UJ ~ x 1. SO UJ U) Z 1. 25 z ...... < ~ ~ 1. 00 Ul __J < ...... x < ::s:: < UJ a.. .75 .50 .25 0.00 0 10 20 30 40 SO LOCATION FROM CENTRELINE OF SAMPLER 60 (1. 70 BD 90 OF CUTTING SHOE RADIUS) (b) Fig. 3.44 Peak axial strains at various locations for samplers of different inside and outside cutting edge taper angles: (a) in compression (b) in extension 130 100 Centreline of sampler c B Fig. 3.45 Finite element mesh with 2303 QXF9 elements for analysing flat-ended sampler I and sampler II (not to scale) 131 ...,zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA Centreline of sampler c 8 Fig. 3.46 Finite element mesh with 2352 QXF9 elements for analysing flat-ended sampler ITI (not to scale) 132 EH AzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA D ~ , I Centre line of sampIer c B Fig. 3.47 Finite element mesh with 2520 QXF9 elements for analysing flat-ended sampler IV (not to scale) 133 t.n t.n ...zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA '". enen". , N . . Cl N (D ..... ..... (T) Ul N ~ ... II \, \ II +J <, +J <, CD +J <, CD CD ....... ....... ....... ....... ....... ....... Cl:!: W et: et: _J -' CL W W _J o, :.£ CL ~ ~ < tJ') < tn < Ul - Cl -I"- - in in -,... II II -I- +J <, CD ~ E E ..._, 1 \- > ....... Cl I- ~ \ Cl:!: W -' CL - ~I :.£ -~ < tn in (T) - "',,, I"- ,, \ !\ , .: w \ Cl N t.n ..... - I I: Cl ..... - Ul I~ ,: I: I (T) t.n N (WW) I I I I Cl N t.n Cl Ul Cl ..... ..... 3V A~VONn08 ~NolV I I _l _l I Cl . . -. ...... . . . Cl Ul CJ CJ C) I Cl Ul I I NO IlVWtlo.:l30 lVIXV 134 ro § tB 0 '0 ._ ro 0 ~ >< ....... roo.. ::J: I- "''0 :.£ C) o Cl) zyxwvutsrqpo 4-<0 0'0 c:: = o ? Z .~: >tB I- 00 u < ....... I '= 0 W U) 1 - f-~~ ._...= ....... 8zyxwvutsrqponm ro ro w - ~ ,: W U =0 ta .= Ct1 lJ..zyxwvutsrqponmlkjihgfe \, 1 Cl u, Cl:!: \, 1 1 \ b.O < Z \ \ :.£ U) _J - ::I 0 o, Cl:!: I- \, \, \, -l- ............. = .L:J Cl Ul N ~ _J (T) 1 ,, ,, W Z ......, \ ......... Cl:!: w -r- \ 1 ~ '0 C) 1 ...... ...... ,....., Cl N Ul N CJ I I I lVlol (T) Cl ~ ~ oil ...... (.l:.. 2.5 L. AT 10% OF C. S. R. FROM C.zyxwvutsrqponmlk S.P.zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFE L. FROM C. AT 30% OF C.S.R. S.P. 2.0 --- ,.... CD " ~ z EXTENSION 1.0 ---- ..... c::l I- < w S. P. AT 50% OF C.S.R. FROM C. L. AT 70% OF C.S.R. FROM C. L. S.P. AT 90% OF C. S. R• FROM C. L. S.P • AT lNSlOE EDGE OF SAMPLER --------·-S.P. ............•... 1.5 .5 c::l ..J I- Z 0.0 lLJ ::E lLJ ..J -.5 lLJ ..J < -1. 0 w COMPRESSION I- ~ lLJ > -1. 5 -2.0 -2.5 -4 -3 -2 2 0 -1 AXIAL STRAIN 3 4 (7.) Fig. 3.49 Strain paths due to penetration of flat-ended sampler I 2.5 2.0 ,.... CD 1.5 <, LEGEND : SAME AS ABOVE EXTENSION N "-J Z 1.0 c::l ..... I- <: w .5 c::l ..J IZ 0.0 lLJ ::E lLJ ..J lLJ -.5 ..J - < w -1. 0 I- COMPRESSION ~ w > -1. 5 -2.0 -2.5 -8 -6 -4 -2 0 AXIAL STRAIN 2 4 6 (;0 Fig. 3.50 Strain paths due to penetration of flat-ended sampler IT 135 8 2.5 S.P. S.P. 2.0zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA --- ,.., ID 1.5 <, EXTENSION ~ z --------·-S.P. ................ 1.0 L. AT 10% OF C. S. R. FROM C.zyxwvutsrqponm FROM C. L. AT 301. OF C.S.R. S. P. AT 501. OF C.S.R. FROM C. L. 70r. OF C.S.R. FROM C. L. ATzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQ S.P • AT 901. OF C.S.R. FROM C. L. Cl I-- -c w .5 t·:~ Cl _J I-Z W ::£ w _J w _- I .....:.......... 0.0 -, -.5 _J -c w -1.0 COMPRESSION I-0:: w > -1.5 -2.0 -2.5 -12 -9 -6 3 0 -3 AXIAL STRAlN 6 12 9 (in Fig. 3.51 Strain paths due to penetration of flat-ended sampler ITl 2.5 2.0 ,.., CD <, N ...., z Cl ...... I-- -e w 1.5 LEGEND : SAME AS A80VE EXTENSION 1.0 .5 _- Cl _J I-Z W ::£ w _J w 0.0 -.5 _J < w -1. 0 COMPRESSION I-0:: w > -1. 5 -2.0 -2.5 -8 -6 -4 -2 o AXIAL STRAIN 2 4 6 8 GO Fig. 3.52 Strain paths due to penetration of flat-ended samplerzyxwvutsrqponml N 136 Cl Cl ...... U) Clzyxwvutsrqponmlk ::J Cl Cl < 0:: Cl CD UJ Cl :c U) ...... CIl ...... u u.. (0 e Z ::J Cl CIl --- L!) Cl E' ~ ~ 0 0 ~ o 0 ::s 0 -5 ....>~ - -e N c 0 0:: UJ _J . . . - . N co 11') Cl + 0 I/") Cl 0 N (T) N Cl _, u.. •zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA ...., II II II U ...., +J <, en <, CD ~ -- --- ..J <, en UJ n, ::E < Ul + ~ ..J ..J ..J ::E ::E ::E UJ a... UJ a, < < Ul • 0 Ul ~ en - ~ Cl +J ..... Cl (T) ~ UJ n, < Cl N Ul x L- - N • ~ - Cl - )( _L ~~ __ ~ __ ~ ~~ __ ~ N (0 __ ~Cl Cl ex) NOISS3~dHOJzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA NI NIV~lS lVIXV ~V3d 137 u ....c CIl - -~ Z ~Q) UJ Z ..... CIl b t:I CIl 0.. e ~ CIl ]~ ~-g UJ u .!. Q) ~ ~ti:::: ::E 0 0:: ~ "1 ~ Z ....~bh u.. ..... 0 e0 Clzyxwvuts c- Cl x + 0.. ...Jzyxwvutsr UJ 0:: )(. > a.. ::E < U) CIl CIl < U Cl ....J CHAPTER 4 LABORATORY INVESTIGATIONS, EQUIPMENTS AND INSTRUMENTATION 4.1 THE MAIN OBJECTIVES The main objectives of the investigations in the laboratory were as follows: (1) The development of an axial strain measuring device which could monitor strains up to 9 to 10 per cent directly on triaxial specimens (102 mm dia. x 204 mm high). (2) To find a suitable device to measure porewater pressure locally at the mid-height of triaxial specimens during Ko-consolidation, application of strain paths and undrained shearing. to run stress and strain path tests on (3) To develop a computer programzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIH reconstituted soft London Clay specimens in a triaxial apparatus. (4) Finally, to carry out oedometer tests on reconstituted soft London Clay specimens to determine its one-dimensional consolidation and permeability characteristics. 4.2 PREPARATION OF RECONSTITUTED SOIL 4.2.1 INTRODUCTION Reconstituted soils are those which are prepared by breaking down natural soils, mixing them as slurry and reconsolidating them. Reconstituted soils are distinguished from both remoulded soils and from resedimented soils which are mixed as a suspension and allowed to settle from that state. Jardine (1985) discussed the difficulties of implementing detailed investigations of general stress-strain and strength properties using intact samples and it was found that the most comprehensive studies invariably employed reconstituted soil. Reconstituted soil enables a general pattern of behaviour to be established and comparisons with the response of intact samples may be used to identify any special features associated with fabric, stress history or 138 bonding. The major advantages of using data from reconstituted soils are that the ambiguous and substantial effects of sampling of natural soils and inhomogeneity can be eliminated, represented. while the essential history and composition The disadvantages of in-situ soils can be are that the important effect of post-depositional process, such as ageing, leaching, etc. and of variations of composition and fabric are not included. So the pattern of behaviour for reconstituted soils discussed in the following chapters will be taken to represent that of young or unaged resedimented soils where no post-depositional processes have operated. 4.2.2 SOIL USED The soil used in the study was brown London Clay collected from Stag Hill site, University of Surrey, Guildford. The liquid and plastic limits, specific gravity and grain size distribution were determined according to the procedures described in BS 1377: 1975. The liquid and plastic limits were determined starting with soil at its natural water content and not using dried material. The cone penetrometer method and density bottle method were used for determining liquid limit and specific gravity respectively. The grain size distribution was determined by pipette analysis. grain size distribution curve of the clay is shown in Fig. 4.1. The The index properties of the clay were as follows: Specific gravity = 2.74 Liquid limit, LL = 69 Plastic limit, PL = 24 Plasticity index, PI = 45 Clay fraction = 54% = 0.83 ActivityzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA A comprehensive description of the London Clay has been reported by Apted (1977). This includes the geological history of the deposit, its present morphological its mineralogy, lithology, structure and its geotechnical characteristics. 139 setting, 4.2.3zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA PREPARATION OF SLURRY Clay slurry with an initial water content well beyond the liquid limit is commonly used as an initial state for sample preparation. Higher initial water contents provide higher degrees of saturation and higher freedom of particle orientation larger initial volumes and longer consolidation periods. was required for preparing enough samples but require Since a large volume of clay and also in order to reduce the consolidation time it was essential to use an initial water content which was sufficient to yield a uniform and homogeneous slurry. To prepare the clay slurry the following procedure was adopted. (1) Large blocks of clay were broken down into medium-size pieces with a wooden hammer. The clay pieces were then put into the container of the clay mixing machine shown in Fig. 4.2 (2) The clay mixer was run at a high speed in order to break down the pieces into smaller and finer particles using the thick blade of the mixer shown in FigA.3(a) (3) The thick blade of the mixer was replaced by the medium-thick 4.3(b)]. blade [Fig. Initially a small quantity of water was added to the clay pieces and the mixer was run at two speeds, initially at a slow speed and then at fast speed, until the clay was thoroughly mixed with water. Then water was added in intervals and the mixer was run at high speeds until a slurry was formed. (4) The medium-thick blade of the mixer was replaced by the thin blade [Fig. 4.3(c)] and the slurry was mixed thoroughly by running the mixer at a high speed until the slurry appeared uniform and homogeneous. The slurry was sieved through a B.S. 2 mm sieve in order to remove small lumps of clay which were not mixed properly. (5) Finally, the slurry was stored in a large plastic container Approximately (94 litre capacity). 30 litres of slurry were prepared from each batch. Repeating the steps from 1 to 5, altogether about 376 litres of slurry, stored in four large plastic containers, were prepared. (LI) of the slurry were respectively The initial water content and liquidity index 103% and 1.75. 140 4.2.4 CONSOLIDATION OF SLURRY The consolidation of the slurry was carried out in large cylindrical steel tank, 1000 mm in diameter and 490 mm deep. The tank was rigidly fixed to its base by 36 bolts. The tank has a lid which can also be fixed to the top of the tank with bolts. Four klinger valves are fitted with the lid. There is a hole at the centre of the base of the tank for bottom drainage. There are also eight holes on the side of the tank, four holes near the base and four slightly below the top of the tank. A schematic in Fig. diagram showing the arrangements for the consolidation of slurry is presentedzyxwvutsrqponml 4.4. Consolidation of slurry was carried out in accordance with the following procedure. (1) All the eight holes on the side of the tank were sealed by placing silicone rubber in them. (2) An approximately 30 mm thick layer of clean sand was placed at the base of the tank. The sand layer was saturated with water and the top surface of the sand layer was levelled properly. Terram 1000, a thermally bonded non-woven geotextile, was placed on the top of sand layer. Terram 1000 functions as a drainage material. The Terram was also saturated with water. (3) Clay slurry was poured into the tank and the tank was filled with slurry up to its rim. The top surface of the slurry was flush with the rim of the tank. A 4 ft x 4 ft x 2 mm thick neoprene rubber sheet was placed on the top of the slurry ensuring that all the holes, cut in the rubber sheet previously, were aligned properly with the holes on the flange at the top of the tank. Silicon rubber was applied on the neoprene rubber sheet along the periphery of the flange of the tank. (4) A rubber spacer was placed on the neoprene rubber sheet along the periphery of the flange of the tank. Four holes in the rubber spacer were properly aligned with those in the neoprene rubber sheet. Silicon rubber was applied along the centreline of the rubber spacer. (5) The lid of the tank was placed on the top of the tank with the help of the lifting rig ensuring that all the holes on the flange of the tank were in alignment with those on the lid of the tank. The lid was then rigidly fixed to the flange of the tank by screwing bolts through the holes using a torque wrench. (6) The space between the bottom of the lid of the tank and the top of the neoprene rubber sheet was filled with de-aired water by closing valve d and opening valves 141 a, b, c and e (see Fig. 4.4). Water was bled through the other three valves attached to the top of the lid of the tank in order to remove all the entrapped air. Valves b and e were then closed, and valve d was opened. A pressure of 100 kPa was set by using air regulator f and applied to the slurry by opening valve b. Consolidation started immediately and the volume of water expelled from the slurry through the valve g was measured by a burette connected to valve g through a rubber tube. At it was found that the bladder within the air-water the early stages of consolidation,zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIH cylinder was completely inflated very quickly (within an hour) and as a result the pressure dropped below 100 kPa. In that situation, the air-water cylinder was refilled with water until the bladder collapsed. However, it was found that after two to three days of consolidation the complete inflation of the bladder took several hours. (7) After 16 days of consolidation, the lid of the tank was removed using the lifting rig. The rubber spacer and the neoprene rubber sheet were also removed. It was found that the slurry has compressed by about 50 mm. Four small holes were dug in the consolidated slurry at points near the four holes on the upper side of the tank. These holes were then unsealed by removing the cured silicone rubber. Four plastic tubes were put through these holes into the tank. One end of each tube was placed near the centre of the tank while the other end was connected with a burette through a rubber tube. Terram 1000 was placed around the surfaces of the holes dug in the consolidated slurry and also on the top surface of the consolidated slurry. Saturated sand was put on the top of Terram 1000 and filled up to the rim of the tank. Top surface of sand flush with the rim of the tank and the neoprene rubber sheet was placed on the top of sand layer. Steps (4) to (6) were followed to apply hydrostatic pressure of 100 kPa to the partially consolidated slurry. (8) By providing drainage through the top and bottom of the slurry, the consolidation rate was accelerated. The volume of water expelled through the slurry was measured periodically. Fig. 4.5 shows the photograph of the detail arrangements while the consolidation was under progress with drainage through top and bottom of the soil. (9) After every 10 to 15 days of consolidation, the very wet sand at the top of the consolidated clay was replaced with saturated sand. Also the volume of sand replaced every time was increased because of decrease in height of the clay layer due to consolidation. It took approximately 133 days for complete consolidation of the clay, which was indicated by no further change in volume of the clay undergoing consolidation. The 142 volume change curve of the clay during the whole consolidation period is presented in Fig. 4.6 4.3 SAMPLING OF CLAY After the completion of consolidation, the bolts used to fix the lid of the tank with the top flange of the tank were unscrewed using torque wrench. The lid of the tank was then lifted using the lifting rig and was placed on the floor. The neoprene rubber gasket, the neoprene rubber sheet, the sand layer and Terram 1000 sheet at the top of the consolidated clay were also removed. A groove of approximately 40 mm width and a depth equal to that of the consolidated clay layer was cut along the periphery of the clay layer using a clay knife. The bolts fixing the bottom flange of the tank with the base of the tank were unscrewed with torque wrench. The tank was then lifted by the lifting rig and was placed on the floor. Fig. 4.7 shows the consolidated soft clay layer resting on the base of the tank. Block samples (approximately 180 mm x 180 mm x 280 mm high) were cut by hand using piano wires and a thin sharp stainless steel plate. The samples were put into polythene bags which were sealed and stored in the laboratory. 4.4 AUTOMATED STRESS/STRAIN PATH TEST EQUIPMENT 4.4.1 INTRODUCTION The triaxial apparatus (Bishop and Henkel, 1962) is the most common testing device for routine examination of soils for geotechnical design and for much current research. This is because the device is simple in design and cylindrical samples are relatively simply prepared by extrusion from sampling tubes or by trimming in a soil lathe. Soil, unlike many materials, is history dependent and path dependent, meaning that its behaviour is governed by the recent stress and strain history and by the current stress and strain changes. The stress history and the stress path applied in a conventional triaxial test are unlikely to be the same as those relevant to soil in the ground with the result that the soil properties measured in the laboratory may not apply to soil behaviour in the ground. AlsozyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQP it is difficult to use the conventional triaxial apparatus to impose the stress or strain paths normally encountered in various geotechnical problems. The difficulty arises because performing such complex stress 143 or strain paths requires accurate and rapid control of the applied stresses or strains in order to follow the desired stress or strain paths. Besides, computer control is required because complex stress or strain paths may require longer test periods which may be beyond the capability of the operator. However, there have been major developments in microcomputers, in electronic instrumentation and in control systems which have allowed automatic control and monitoring of stress paths in the triaxial apparatus with high accuracy and at reasonable cost. Hight (1983) developed a control system which is capable of applying precisely specified sequences of stresses or deformations to soil samples in various pieces of testing equipment, for example, in the hydraulic triaxial apparatus (Bishop and Wesley, 1975), the plain strain apparatus (Atkinson, 1973), and the hollow cylinder apparatus (Hight, 1983). The closed loop control is based on a digital computer which both analyses data from transducers, monitoring the test's progress and soil's response and directs servo-systems operating pressure and displacement regulators. Microcomputer controlled stress path equipment developed at the City University, London has been described by Atkinson (1985) and Atkinson et al (1985). Based on a Bishop and Wesley cell (Bishop and Wesley, 1975), it is able to apply a radial stress to the sample by the cell pressure and an axial stress independently through a hydraulically operated ram while a back pressure can be applied at the base of the sample. All these pressures may be changed independently to follow any triaxial stress path. Pressures are controlled by electromanostats operated by stepper motors. The microcomputer provides facilities not only for control but also for data logging, data analysis, printing and ploning. Automated stress path system has also been developed at the University of Surrey (Khatrush, 1987; Clayton and Khatrush, 1988). The system has been programmed utilising a microcomputer to take full control of the applied stresses so that any desired stress path can be closely followed in the conventional triaxial cell for testing 4 inch diameter specimens. The stress path apparatus was instrumented with a internal load cell, cell and back pressure transducers, local axial and lateral strain measuring devices. an external axial strain gauge and a volume measuring device. The system has proved accurate and reliable in applying a wide range of stress paths in compression and extension and a combination of both.zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQP 144 The automated stress path system developed at Surrey University was used for carrying out the strain path tests to model undrained tube penetration disturbances due to sampling in clay. and also modified. device. However, some of the instruments of the system were changed The new system used the newly developed local strain measuring The local lateral strain measuring device was also modified and a miniature transducer for measuring local porewater pressure at the mid-height of the specimen has been provided. The programming of the system was considerably modified to meet the requirements of the testing programme for the present research. 4.4.2 BASIC FEATURES OF THE STRESS/STRAIN PATH TEST EQUIPMENT The equipment consists of four major parts, the microcomputer, controllers, the signal conditioning unit and the loading system. is monitored by means of transducers monitoring the pressure The data from a test axial load, cell pressure, back pressure, pore pressures at the base and mid-height of specimen and, axial and lateral deformations. The system also contains some peripheral devices interfaced to the computer, including a disc drive for storing data and loading programs, a printer for obtaining hard copies of the data. A general view of the system is shown in Fig. 4.8 and the basic features of the system are also schematically illustrated in Fig. 4.9. The major parts of the system are described below. THE MICROCOMPUTER The microcomputer is the central controller of the whole system. Its basic function is to receive and store initial test information as assigned by the operator, receive digital output from the transducers, do the necessary calculations to convert these to engineering units; command the pressure controllers to supply the required loading pressures; print, plot and store data at the end of required intervals of time; and finally to process the data at the end of each test. B microcomputer was used. A standard Hewlett-Packard Through an IEEE-488 interface, the computer 86 was linked to the signal conditioning unit and to other peripherals such as the disc drive and printer. 145 THE PRESSURE CONTROLLERS Three pressure controllers were used, each to control one pressure unit, namely back pressure, cell pressure and axial pressure. Each pressure controller consists of a small stepper motor operating through a reduction gear box and a flexible coupling to provide the required mechanical rotations in order to drive a manostat air pressure regulator. Making the motor step in either direction increase or decrease. causes the air pressure to The stepper motor requires a 12 Volt DC power supply to operate and therefore a special voltage converter box was used for this purpose. The controllers were interfaced to the computer through an HP 8294A General Purpose Input and Output (GPIO) interface. configurations This interface provides eight different hardware for four 8 bit ports. In the system only one output 8 bit port was used to control the three stepper motor driven air pressure regulators (for cell and back pressure, switching and deviatoric from triaxial load) and two relay driven compression to extension solenoid and vice versa). valves (for A typical relationship between the number of steps and the generated pressure for the pressure controllers is shown in Fig. 4.10. It can be seen that the relationship is linear up to a pressure of 500 kPa, above which a slight diversion appears with further increase in pressure. Each step by the motor was found to cause a change of pressure of approximately 0.07 kPa and the time taken to perform one step is approximately 0.1 sec. Thus an application of an increment ofzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIH 100 kPa cell pressure, for example, will take a minimum of approximately THE SIGNAL CONDITIONING The signal conditioning 143 seconds. UNIT unit receives the output analogue signals from various measuring devices, amplifies and converts them to digital form, and then passes this information to the computer. The unit incorporated gauge amplifier system (SGA 11(0) manufactured within the system is a strain by CIL Electronics Ltd. signed 12 bit instrument (i.e., only ±4096 bits digital output can be obtained). logged device was calibrated within this range. It is a Each A total of nine channels were used to monitor the signals from nine transducers within the system; three transducers to detect the back, cell and local pore pressures, an internal load cell to measure the axial deviatoric load, three local strain measurement devices, an external displacement transducer to measure the overall axial strain of the specimen and the volume change 146 device (see Fig. 4.9). Although the volume change device was incorporated in the system, it was not used in the actual tests. Each device was energised with 10 volts DC power supply. Any desired channel can be selected and to read the data from the corresponding device. A listing of the basic algorithm to scan data from various measuring devices is given in Appendix-B. THE LOADING SYSTEM The loading system consists of a loading frame, an air actuator and a triaxial cell. The loading frame is an ordinary two-post frame with adjustable cross bar and a flat plate base, used to provide reaction for vertical load application and to accommodate the 102 mm dia. triaxial cell. The actuator is a double acting Bellofram diaphragm air cylinder of 10 bar maximum pressure capacity. stress. It is used to apply deviatoric The two chambers behind the upper and lower Belloframs are filled with pressurised air and each of them is connected to one of the solenoid valves, which is in tum connected to the axial pressure controller. By controlling the opening and closing of the two valves both compression and extension stress controlled tests can be performed. it is necessary to apply an upward force on During an extension testzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPON the top cap which is transmitted to the sample. the load cell to be screwed to the top cap. Measurement of this load required The top cap arrangement developed by Bishop and Henkel (1962) was used. 4.4.3 THE MEASURING DEVICES The measuring devices incorporated in the system include a load cell, cell and back pressure transducers, a pore pressure transducer to measure pore pressure at the base of the specimen, an external measuring and a miniature devices deformation discussed measuring in separate deformation characteristics and pore displacement devices sections transducer, local pore pressure and the miniature measuring transducer. pore pressure with a brief introduction pressure three local deformation devices. transducer of the available Types of the other measuring devices are stated below. 147 The local and are local calibration LOAD CELL The load cell used to measure deviator stress was manufactured by MIL to the specification of the University of Surrey, commissioned by Dr. C.R.I. Clayton and Mr. M.C. Matthews. This load cell is now supplied by Wykeham Farrance. It has a 5000 N capacity. The load cell was calibrated using a Budenberg dead load tester. Fig.4.11(a) shows a typical calibration curve of the load cell. It showed excellent linearity with negligible hysteresis (0.075%) during unloading from a maximum calibrating load of 3982 N. The resolution of the load cell was 1 Nlbit. Linear regression analysis was carried out and the regression error plot is shown in Fig. 4.11(b). It can be seen that the maximum linear regression error is ±O.06% of the full-scale output (at 3982 N). A calibration of the load cell against increase and decrease of cell pressure was also carried out while the load cell was connected to a 102 mm dia. soil sample. Fig. 4.12 shows the calibration characteristics of the load cell for two cycles of loading and unloading. An approximate linear relationship between the increase and decrease in cell pressure and the developed negative tensile force were obtained as shown in Fig. 4.12. During unloading large hysteresis (a maximum of 10%) was obtained. A change of cell pressure of 100 kPa developed a tensile of approximately 28 N which was transmitted to the sample. This tensile force is equivalent to approximately 3.5 kPa per 100 kPa increase in cell pressure for a 102 mm dia. sample. CELL PRESSURE, BACK PRESSURE AND BASE PORE PRESSURE TRANSDUCERS The measurement of cell and back pressure was achieved by means of Druck PDCR 10 pressure transducers having an operating range of 0 to 150 psi (i.e., 0 to 1034 kPa). They were calibrated using the Budenberg dead load tester for a working range of 0 to 1000 kPa. The calibrations of the cell and back pressure transducers are shown in Figs. 4.13(a) and 4.14(a) respectively. The resolution of each transducer is 0.25 kPa/bit. The transducers showed excellent linearity and negligible hysteresis (0.05%) during unloading from a maximum calibrating pressure of 1000 kPa. Linear regression analyses were carried out for both the transducers. 148 The maximum linear regression error was found to be ±O.08% of the full-scale output (at 1000 kPa) for both the cell and back pressure transducers. The regression error plots for the cell and back pressure transducers are presented in Figs. 4.13(b) and 4.14(b) respectively. A Bell and Howell fluid pressure transducer (operating range of 0 to 150 psi) was used to monitor porewater pressure at the base of the sample. It was calibrated for a working range of 0 to 1000 kPa using the Budenberg dead load tester. A typical calibration of the transducer is shown in Fig. 4.15(a). This transducer also showed very good linearity and negligible hysteresis (0.1%) during unloading from full working pressure. Linear regression analysis was performed and the regression error plot is shown in Fig. 4.15(b). Maximum linear regression error was ±O.I % of the full-scale output (at 1000 kPa). EXTERNAL AXIAL DEFORMATION MEASURING DEVICE Overall deformation of a specimen was measured by means of a linear strain conversion displacement transducer (LSCDT) manufactured by MPE transducers Ltd. The transducer was calibrated over a travel range of 1 inch (25.4 mm). A typical calibration curve is shown in Fig. 4.16(a). approximately 6.35 urn/bit, The resolution of the transducer is Linear regression analysis was also carried out on the linear portion of the output and the regression error plot is presented in Fig. 4.16(b). It can be seen that the maximum linear regression error is ±O.077% of the full-scale output in the linear range. 4.5 LOCAL DEFORMATION MEASUREMENT 4.5.1 INTRODUCTION Conventional measurement of the axial deformation of triaxial specimens, made outside the triaxial cell, introduces significant errors in the computation of strains. The errors mainly result from the effects of compliance of the apparatus and the bedding on the end platens (Daramola, 1978; Burland and Symes, 1982; Costa Filho, 1985). Error due to compliance of the apparatus can partially be minimised by modifying the testing equipment in order to increase its stiffness (Atkinson and Evans, 1985), and partially by careful calibrations of various components. However, the bedding errors are difficult to ascertain and correct since their magnitude depends 149 on the way in which the ends of the specimens are prepared. Therefore, the only way to obtain accurate determination of axial deformation is to carry out the measurement directly on the surface of the triaxial specimen. This has been recognised by many previous investigators, who have suggested different methods to determine local strains (Daramola, 1978; Yuen et al, 1978; Brown and Snaith, 1974; Brown et al, 1980; Jardine et al, 1984; Symes and Burland, 1984; Costa Filho, 1985). The various techniques employed by the previous investigators have comprehensively been reviewed by Khatrush (1987). The various devices have been reported to be either costly or difficult to implement. Recently a new device for measuring local axial strains on triaxial specimens has been developed by Clayton and Khatrush (1986). This device makes use of a Hall effect semiconductor. A Hall effect semiconductor is typically direct current (DC) energised and delivers a DC output which varies linearly with magnetic flux over a specified range. The relationship between the output voltage and the relative displacement between the two ends of the device is linear over a range of 2.5 mm. This device, therefore, can measure a maximum of 3.5% axial strain (on a 70 mm gauge length) of the middle third portion of a 102 mm dia. x 204 mm high triaxial specimen. Thus this type of strain gauge is particularly suitable for measuring small it was proposed to axial strains. During the current research, as mentioned earlier,zyxwvutsrqponmlkjihgfedc perform strain path tests on soft London Clay specimens which were likely to undergo large axial deformations during testing. Therefore, in order to measure large local axial strains, a new strain device was required. 4.5.2 DEVELOPMENT OF A LOCAL AXIAL STRAIN MEASURING DEVICE The Hall effect semiconductor used for local axial gauge developed by Clayton and Khatrush (1986) was chosen as the sensing element for the new device because of a number of special features. These are as follows: (a) The semiconductor is light (0.35 g), so that it imposes negligible loads on the sides of the soil specimen, and it works equally well in air or pressurised water. (b) The sensor is compensated against changes in ambient temperature and DC voltage supply. (c) The sensor has a single DC output that varies linearly with magnetic flux density 150 from -40 mT to +40 mT, and can work with any DC voltage supply from 8 to 16zyxwvuts V. The sensor has proved 10 be very reliable and accurate (Khatrush, 1987). The problem was to configure a Hall effect sensor-magnet system to give an output voltage which is linear over a considerable range (6 to 7 mm) with respect to the displacements between its ends. This linear range was required for the measurement of axial strains up to 9 to 10 percent over a 70 mm gauge length. In order to achieve high linearity between voltage and displacement, attempts were made, for the first time, to use pole pieces with a magnet. A pole piece or flux concentrator is a magnetically soft material such as mild steel. When added to a magnetic system, pole pieces provide a lower resistance path to the lines of flux. As a result pole pieces tend to channel the magnetic field, thus, changing the flux densities in a magnetic circuit. 4.5.2.1 STAGES IN THE DEVELOPMENT Basically two different configurations of Hall effect sensor-magnet-pole piece system were investigated to achieve the desired degree of linearity between the output voltage and the relative displacement of device ends. In the first type of configuration two magnets and one pole piece were used. The magnets, separated by a distance and placed on a mild steel pole piece, were moved over the semiconductor's sensing face as shown in Fig. 4.17(a). This type of movement is If a second horizontal plane is drawn through the called slide-by movement.zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLK semiconductor (or sensor), the distance between the two planes is referred to as the gap. The sliding contact between the sensing face of the semiconductor and the magnets and pole piece was made with of a small polytetrafluoroethylene (PTFE) block. This PTFE block also controls the gap between the semiconductor's sensing face and the face of the magnets. Several calibrations were carried out by varying the distance between the magnets and keeping the gap fixed. All the calibrations were performed with the help of a micrometer (0.00254 mm minimum resolution) mounted in a specially built calibration jig. Both the micrometer and the jig were manufactured of non-magnetic materials, in this case aluminum alloy, in order to avoid any local influence on magnetic fields in the area of the semiconductor. The different types of magnet used under this configuration and their respective calibration characteristics are described below: 151 (i) Two 2 x 2 x 1 mm thick magnets were used. During calibrations the distance between the magnets was varied between 1.5 mm and 3.0 mm. The gap was fixed at 0.5 mm. From the calibrations curve it was found that the maximum linear range was about 2.5 mm and within this range the output voltage varied by about 2.68 V. In order toassess the effect of a pole piece on the linear range for this configuration calibrations without the pole piece were carried out and it was found that the calibration characteristics were identical. Because a small linear range was produced, these magnets were not accepted. However, this configuration without the pole piece was later used with a number of other improvements (Clayton et al, 1989) to replace the earlier version of the Hall effect axial gauge developed by Clayton and Khatrush (1986). (ii) Two 2 x 2 x 4 mm long magnets were used. During calibrations a gap of 1 mm was kept and the distance between the magnets were varied from 2 to 4 mm. The maximum linear range obtained was approximately 2.2 mm and over this linear range the output voltage varied by approximately 2.22 V. The magnets were discarded because they produced a small linear range. (iii) Calibrations were performed using 3 x 2 x 7 mm long magnets. The distance between the magnets varied from 3 to 7 mm while a gap of 1 mm was maintained. From the calibration curves it was found that the maximum linear range obtained was about 2.2 mm. Within this linear range the output voltage varied by about 3.21 V. These magnets were discarded because they produced a small linear range. The second type of configuration consisted of one magnet and two pole pieces. The pole pieces, made of mild steel, were rectangular in shape and were slightly tapered at the corners of one end. The magnet was held in between the inner faces of the pole pieces and was moved horizontally over the sensing face of the semiconductor as shown in Fig. 4.17(b). The sliding contact between the semiconductor's sensing face and the magnet and pole pieces was made with the help of a smallzyxwvutsrqponmlkjih P'IFE block. This PTFE block also serves to maintain the gap between the semiconductor's sensing face and the edges of the pole pieces. With the help of the same micrometer as used earlier,several calibrations were carried out using different types of magnets. 152 The different types of magnets which were tried under this configuration and their corresponding calibration characteristics are described in the following paragraphs. (i) A 6.2 x 6.2 x 8 mm long magnet was used. During calibration a gap of 0.5 mm was maintained between the semiconductor's sensing face and the edges of pole pieces. Although about 5.1 mm linear range was obtained, the magnet was not accepted because of very low output voltage (0.49 V) within the linear range. (ii) Calibrations were carried out using a 3 x 3 x 9 mm long magnet. The gap between the semiconductor's sensing face and the edges of the pole pieces was varied between 0.5 mm and 2.0 mm. The maximum linear range was found to be about 6.0 mm when the gap was fixed at 1.5 mm. For this linear range the output voltage varied by only 0.44 V. The magnet was, therefore, discarded because it provided a very low output voltage even though the linear range was acceptable. (iii) A 6.35 mm long x 6.35 mm dia. cylindrical magnet was adopted. A calibration was performed with a gap of 1.5 mm between the sensor's sensing face edges of the pole pieces. Only about 2.2 mm linear range was obtained and over this range the output voltage varied by approximately 0.53 V. So, this magnet was not accepted because it produced very small linear range and very low output voltage as well. (iv) A 7.25 mm long x 10.2 mm dia. cylindrical magnet was used. Several calibrations were carried out by varying the gap (0.5 to 2.0 mm) between the sensing face of the semiconductor and the edges of pole pieces. A maximum linear range of about 5.7 mm was obtained when the gap was kept at 1.5 mm. Over the linear range the output voltage varied approximately 1.51 V. Although this magnet produced a reasonable linear range and output voltage,zyxwvutsrqponmlkjihgfedcbaZYX it was not accepted because of the larger size and weight of the magnet. Some problems were also experienced in holding the magnet between the pole pieces. (v) A 6 x 6 x 8 mm long magnet was used. Several calibrations were performed by varying the gap between the semiconductor's sensing face and edges of the pole pieces. The gap was varied between 1.5 mm and 2.5 mm. The maximum linear range for this magnet was found to be about 7.0 mm when a gap of 2.0 mm was maintained. Over the linear range the output voltage varied by about 1.89 V. 153 A summary of the calibration performances of all the magnet systems is shown in Table 4.1. The second type of configuration, using a 6 x 6 x 8 mm long bar magnet held between the inner faces of two mild steel pole pieces, thus provides much better results regarding both the linear range and the output voltage. This configuration, therefore, was finally accepted for the Hall effect local axial strain device. Apart from increasing the linear range, a number of improvements over the earlier version of the gauge (Clayton and Khatrush, 1986) have been made. These include the following: (a) replacement of the metal spring strip at the top of the arm with two lengths of stainless steel spring wire, (b) the use of an arm consisting of bent brass strip, to reduce machining operations, (c) introduction of an adjustment system in the lower pad, to allow proper pad alignment, subsequent placement of the semiconductor, and final adjustment after fixing to achieve any desired positionzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIH in the linear range of the device, and (d) optional pinning. Design of this Hall effect local strain device is shown in Fig. 4.18. 4.5.2.2 GAUGE CHARACTERISTICS AND CALm RATIONS Two gauges were built. They were logged with the signal conditioning unit and were calibrated using the micrometer. A typical calibration characteristic is presented in Fig. 4.19(a). It shows that the relationship between the output (bits) and the relative displacement between the two ends of the device is linear over a range of about 7.0 mm. Within this range the output voltage varies by about 1.89 V DC. Very little magnification is, therefore, required before the signal can be offered to the analogue to digital convener and input to the computer. The minimum resolution of the gauge is approximately 1 J.l.m/bit,which is equivalent to an axial strain of less than 0.002% on a 70 mm gauge length. Linear regression analysis was carried out on the linear portion of the output, and a typical regression error plot is shown in Fig. 4.19(b). 154 4.5.3 THE LOCAL RADIAL STRAIN MEASURING CALIPER The mechanical parts of the caliper are essentially similar to those originally designed by Bishop and Henkel (1962), but with the original mercury indicator column replaced by a slide-by Hall effect sensor-magnet-pole pieces configuration [Fig. 4.l7(b)] used in the axial gauge. The design of the device is shown in Fig. 4.20. A 6 x 6 x 8 mm long bar magnet held between the inner faces of two mild steel pole pieces and attached to a brass block was placed inside an aluminium container. The position of the magnet can be adjusted with respect to the semiconductor by means of two screws moving through the sides of the container. The container housing the magnet, pole pieces and brass block is attached to one arm of the caliper while the Hall effect sensor, encapsulated inside a brass container, is attached to the other caliper arm. The gap between the edges of the pole pieces and the sensing face of the semiconductor was kept fixed at 2.0 mm by means of azyxwvutsrqponmlkjihgfe PTFE block. The two parts of the device are maintained in contact because of the use of uneven spring forces at the caliper hinge. The Hall effect caliper, logged with the signal conditioning unit, was calibrated using the same micrometer and the calibration jig used to calibrate the axial gauge. Initial alignment between the semiconductor and the magnet was adjusted during calibration and kept unchanged during any further use. This alignment should be made when the distance between caliper pads is equal to the nominal diameter of the test specimen (102 mm). A typical calibration curve showing the relationship between the output (bits) and the lateral movement is shown in Fig. 4.2l(a). A linear range of approximately 3.4 mm was obtained. The minimum resolution of the caliper is approximately 0.53 J.1m/bitwhich allow measurement of lateral strains of less than 0.001%. Linear regression analysis was also performed on the linear portion of the output and the regression error plot is shown in Fig. 4.21(b). The maximum linear regression error for the axial and radial strain gauges calculated from the linear regression error plots [Figs. 4.l9(b) and 4.2l(b)] are respectively 0.57% and 0.54% of the full-scale output in the linear range. This is a good achievement for a non-commercial transducer. The overall accuracy of these devices is rather more difficult to assess, since it is controlled not only by the performance of the device during calibration under the ideal conditions. Factors which may be 155 relevant are as follows: (1) ability to determine the gauge length over which the strains are calculated, (2) the vertical position of suitable reference points from which to calculate strains free of end effects, (3) the need for three devices per specimen (as opposed to the two currently commonly used), to avoid tilt effects, and (4) effect of barrelling and necking of the specimen at high strains. The influences of the above factors have been described by Clayton and Khatrush (1987) and Khatrush (1987). 4.6 VOLUME CHANGE MEASUREMENT Volume change can be measured in triaxial testing by means of three methods (Bishop and Henkel, 1962). The first measures the volume of fluid entering or leaving the triaxial cell to compensate for the change in volume of the sample. This method is used for partially saturated soils. Appropriate corrections are required for cell and tubing expansion shearing. and piston rod penetration into the chamber during The second measures the volume of fluid entering or leaving the pore space of the soil. This method is used only for saturated specimens. method permits calculation of volume from direct measurement length and diameter of the specimen. The third of the change in This method may be used for both saturated and unsaturated specimens. Various devices have been developed in order to record volume change. earliest is the burette system (Bishop and Donald, 1961). include servomechanism Perhaps the Other types of devices systems (Lewin, 1971; Irwin, 1972; Watts, 1980), mercury pot systems (Rowlands, 1972; Darley, 1973; Klementev, 1914) and, rolling diaphragm and displacement transducer systems (Menzies, 1975; Hodgson, types of devices have been reviewed by Alva-Hurtado 1976). All these and Selig, 1981. However, the accuracy of these volume change devices does not depend only on their electrical characteristics, since in addition there are some other errors which take place. The nature of these errors has been discussed by Khatrush (1987) from the calibration results of three typical devices. also emphasised Khatrush, from his various stress path test results the need for local measurement 156 of volume change in order to achieve better accuracy. All volume change measurements were, therefore, carried out from the Hall effect local axial strain and lateral strain devices, as described in the previous sections. 4.7 MEASUREMENT OF POREWATER PRESSURE 4.7.1 INTRODUCTION In undrained triaxial tests, porewater pressures are normally measured at the base of samples. In tests carried out with fixed-end samples, the shearing rate has to be selected to ensure that pore pressure non-uniformities arising from end restraint have equalised throughout the height of the sample either at failure, if only effective stress shear strength parameters are required, or at an early stage of the test if the effective stress path is to be derived. However, because of non-uniform distribution of pore pressure due to the effect of end restraint, the base pore pressures during shearing are not equal to those at the centre of the samples (Bishop et al, 1960; Blight, 1963). It is, therefore, necessary to measure pore pressures at locations where the applied In the central zone of a stresses are uniform and known, i.e., away from the ends.zyxwvutsrqponmlkjihgfedcbaZYX fixed-end triaxial sample, including its periphery, stress conditions are reasonably uniform and determinable (Hight, 1983). If the pore pressures are monitored with a piezometer probe in the central zone, where total stresses are known and uniform, the restriction of using a shearing rate which will allow full equalisation of pore pressures throughout the height of sample can be lifted. However, it is still essential to use a shearing rate which ensures that local pore pressure non-uniformities and the pore pressure set up byinterference between the probe and the sample have equalised. Piezometer probes positioned at mid-height have been used by several investigators (Bishop et al, 1960; Richardson and Whitman, 1963; Barden and McDennotl, 1965; Blight, 1965; Maguire, 1975). Hight (198:!, 1983) described a technique for measuring pore pressures using a piezometer probe. The probe is based on a miniature silicon diaphragm pressure transducer which is mounted with its porous ceramic disc flush with cylindrical surface of the sample at mid-height. This combination leads to a minimum of interference between the piezometer and the sample and a short response time for the piezometer-soil system. The response of this piezometer to undrained increases and decreases in cell pressure for saturated 157 sample of Lower Cromer till and London Clay was investigated. was found to be much less than one second. The transducer The response time has proved quite reliable and accurate in monitoring pore pressures in monotonic loading and cyclic loading tests (Hight, 1983). One of the principal advantages of using this transducer is that considerable savings can be made in testing time by increasing the rate of shearing, especially for large diameter samples. It was, therefore, decided to use the same miniature pore pressure transducer for monitoring pore pressures locally at the mid-height of 102 mm dia. x 204 mm high triaxial specimens. 4.7.2 THE MINIATURE PORE PRESSURE TRANSDUCER The miniature pore pressure transducer used, manufactured by Druck Ltd., has a diffused silicon strain gauge diaphragm as the sensing element placed immediately behind a ceramic high air-entry stone. For use as a piezometer The transducer was of the PDCR 81 type. in triaxial testing the transducer offered the following advantages: (a) Small overall size and weight which enable it to be placed on the side of the specimen. (b) A rapid response time even with soft London Clay of very low coefficient of consolidation (c,zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA = 0.25-0.5 m2/yr.). (c) High output which reduces loss of resolution in high speed logging. (d) Lack of hysteresis, making it especially suitable for measurements of pore pressure changes during unloading from compression to extension. (e) Considerable savings in testing time as it is unnecessary to wait for equalisation of pore pressures throughout the specimen height. Characteristics of the transducer are summarised below. 158 Overall length 11.4 mm Overall diameter 6.4 mm Weight (excluding cable) approximately 2 g Transduction principle 4 arm silicon strain gauge Excitation voltage 5 Y 6 rnA nominal (10 Y max.) Full-scale output (PSO) 75 mY/5 Y energisation Non-linearity ±O.I% of FSO Pressure range 0-700 kPa Operating temperature range -2(11C to +12(11C The transducer was calibrated for a working range of 600 kPa by placing it in the triaxial cell and applying known increments of pressure with the Budenberg dead load tester. The gain was adjusted so that for an applied pressure of 600 kPa, the output from the transducer was 2400 bits. presented in Fig. 4.22(a). The calibration curve of the transducer is The resolution of the transducer is 0.25 kPa/bit. The transducer shows excellent linearity and negligible hysteresis (0.125%) during unloading from the maximum calibrating pressure of 600 kPa. Linear regression analysis was also carried out for the calibration curve shown in Fig. 4.22(a). The regression error plot is shown in Fig. 4.22(b). It can be seen that the maximum linear regression error is ±O.23% of the full-scale output (at 600 kPa). A technique was devised for the installation of the transducer at the periphery of 102 mm dia. x 204 mm high soft London Clay specimen. The installation procedure will be outlined in the following chapter. 4.7.3 VALIDITY OF MEASURED PORE PRESSURE AT MID-HEIGHT There are a number of sources or error in the mid-height measurement of pore pressures in undrained monotonic loading, as discussed by Hight (1983). These include: (i) The existence of radial non-uniformities of pore pressure at mid-height (Balla, 1960; Barden and McDermott, 1965; Costa Filho, 1980). 159 (ii) Expansibility (or leakage) in the pore pressure measuring systems and drainage lines leading to the specimen. Where expansibility is large, significant volume of porewater transfer between the specimen and the measuring device takes place so that test becomes partially drained. (iii) Interference between measuring system and the specimen. (iv) Calibration of the electrical transducer. (v) The occurrence of air diffusion through the membrane, and, in some cases, osmosis. In the light of theoretical and experimental evidence, radial non-uniformities are relatively small. Expansibility has been minimised by using a transducer with a low volume-intake factor. Leakage has been eliminated. Interference between the transducer and the specimen was reduced by mounting the transducer at the periphery of the specimen. Calibration was carried out using the Budenberg dead load tester. Excellent linearity between the output (bits) and applied pressure was obtained and the transducer showed negligible hysteresis. There was no possibility of diffusion of air because de-aired water has been used. 4.8 SOFTWARE FOR STRESS AND STRAIN PATH TESTING 4.8.1 INTRODUCTION A computer program was available for stress path testing in the triaxial apparatus. This program was developed by Khatrush (1987). The computer program was written in BASIC. Additional special commands provided by the HP-input/output ROM were used to communicate with various peripheral devices. The program was constructed as a series of block sub-programs, each serving a particular function. It was developed in an interactive way to allow a dialogue to take place between the computer and the user via a video display. All information and data will then appear on the screen so that the user has a choice of either keeping or correcting them. This computer program, originally developed by Khatrush (1987), has been modified considerably to fulfil the requirements of the proposed experiments. For convenience, two programs have been used. The first one was used for performing approximate Ko-consolidation paths followed by stress controlled undrained conventional triaxial compression or extension test as a special stress path test with constant cell pressure. 160 The second program was used to carry out approximate K,,-consolidation paths followed by undrained strain paths simulating tube penetration disturbances and finally undrained shearing in compression. During the application of strain paths and subsequent shearing in compression, the cell pressure was kept constant. The general layout of both the programs were the same. The flow chart showing the layout of the programs is illustrated in Fig. 4.23. The other special features which have been incorporated in the programs are as follows: (a) Control of the rate of testing for each stress or strain path as desired by the user. (b) Reporting of test specimen's status at intervals of 15 seconds; current values of stresses and strains are displayed on the screen and stress path and stress-strain curve are also plotted on the graphics display as the test proceeds. (c) The test data for each stress or strain path is printed in a tabular form at regular intervals of time as specified. The test data are finally stored on disc at the end of each test. (d) Safety features have been included in the software, so that none of the transducers may exceed their maximum operating range. 4.8.2 STRESS AND STRAIN PATH CONTROL PROGRAMS 4.8.2.1 STRESS PATH CONTROL SUB-PROGRAM In order to perform a stress path test, axial and cell pressures must be applied simultaneously so that a constant ratio is always maintained between them. For such a stress path to be performed in this system, the user has to enter sets of coordinates from the keyboard of the computer for all the required stress paths. The method of application of stresses is controlled by having one stress component in control and adjusting the other to maintain the stress path. The controlling stress component is either the deviator stress or cell pressure or radial stress. The controlling stress component for each stress path is chosen by the computer depending on the absolute value of the incremental stress ratio (AK). For each stress path there are two sets of coordinates; one for the starting or initial point and the other for the target or final point. The computer does all the necessary calculations to determine the incremental stress ratio, AK which is given by the following expression: AKzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA = llq/Aa. •...(4.1) 161 where, L\q = change in deviator stress between the target and initial point of the stress path, and L\O'r= change in radial stress between the target and initial point or the stress path If the absolute value ofzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA L\K is less than or equal to unity then the controlling stress component in the radial stress; otherwise it is the deviator stress. The direction (i.e., increase or decrease) of the controlling stress component is chosen depending on the relative values of L\q or L\crr. For example, if the controlling stress component is the deviator stress (i.e., L\K>I) and the Bellofram is in compression then the stress path is obtained by applying a small increment of deviator stress. The required value of current radial stress (crzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA from the following r) to maintain the stress path is calculated equation: o, = (q - '!i)/L\K .... (4.2) + (crr)! current deviator stress where, q = deviator stress for the initial point of the stress path , and radial stress for the initial point of the stress path The current value of radial stress is checked and if it does not satisfy the Equation (4.2), the computer will instruct the radial stress controller to increase or to decrease the radial stress until the current stress value approaches the calculated value within ±O.5 kPa. Another increment of the controlling stress will then be applied, and so on until the desired target coordinate value is reached. A sequence of priorities is normally followed between each two successive increments of stress applications; this includes checking the back pressure, the intervals for data plotting and printing and rate of testing. Alternatively, for example, if the controlling stress component is the radial stress and L\crris positive, then the stress path is obtained by applying a small increment of radial stress. The appropriate value of the current deviator stress (q) to maintain the stress path is determined from the following equation: ....(4.3) The current value of deviator stress is checked and if it does not satisfy Equation (4.3), the computer commands the axial stress controller either to increase or decrease the deviator stress until the current deviator stress approaches the calculated value 162 within ±0.5 kPa. Another increment of radial stress will then be applied, and so on until the target coordinate value is reached. Fig. 4.24 illustrates the flow chart for the stress path control sub-program. In order to simulate closely the desired stress path and to avoid overshooting, the increments (or decrease) of applied controlling stresses were made very small ( approximately 0.14 kPa and 0.153 kPa for the radial and deviator stress control respectively). For adjusting the radial and deviator stresses to maintain the desired stress path, however, the respective increments (or decrease) in radial and deviator stresses are approximately 0.07 kPa ~d 0.1 kPa. By achieving this any stress path in the triaxial stress space can be followed to within ±1 kPa. The desired total time for performing each stress path should be prespecified by the user on input for each stress path. The minimum stress application per single step, as obtained from the pressure controllers calibration shown in Fig. 4.10, is 0.07 kPa. This will be less in the case of deviator stress application , depending on the ratio between the area of the actuator and the area of the specimen (found to be 0.73). Hence, the minimum deviator stress application per single step becomes 0.051 kPa. (AT) required to complete one successive increment of the Therefore, the timezyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA controlling pressure in order to maintain the total time of testing for each stress path is controlled by the following equation: AT = (TOTIM x NS x NF x 3600)/SPL where, TOTIM = NS = ....(4.4) total testing time in hour for the stress path number of steps in a minimum applied increment of controlling pressure NF = minimum pressure per single step SPL = length of the controlling stress component If the desired rate of testing is below the fastest rate as calculated from Equation (4.4), no consequent pressure increment is applied until the increment of time, AT has elapsed. The desired rate of testing was achieved to within ±4% for stress paths which took several days to complete, e.g., Ko-consolidation path. However, this figure might change to higher values in cases of stress or strain paths those took only a few hours for completion. 163 As outlined previously, during the increase of cell pressure a small negative force is developed. Therefore, the resulting stress path will not follow the isotropic axis unless an equivalent pressure is applied to the top chamber of the actuator in order to maintain zero deviator stress during increase of cell pressure only. When crossing the isotropic line, one of the two valves supplying pressure to the actuator is shut and the other is opened at the same time and, the applied pressure is transferred from one side to the other. This process is usually allowed to occur when the pressures on the two sides of the actuator are equal. In that case no force is applied to the ram which is connected to the load cell. As a result, the deviator stress in the specimen should be zero. However, this has not been found to be the case, since a small pressure difference between the two sides of the actuator does exist to compensate for the development of tensile force during the increase in cell pressure only. Crossing (from compression to extension and vice versa), therefore. has been allowed to take place at a negative deviator stress equal to the pressure equivalent of the developed negative tension force. This negative tension force was calculated from the calibration curve shown in Fig. 4.12. This correction factor was included in the software and consequently the change in deviator stress during crossing has been significantly reduced to less than 2 kPa. 4.8.2.2 STRAIN PATH CONTROL SUB·PROGRAM The strain path control sub-program is basically a modified version of the stress path control sub-program. The levels of strains which the specimen will be subjected to have to be prespecified by the user on input. These include the maximum strain during the initial compression phase, the maximum strain during the extension phase and, the minimum strain during the second compression phase of undrained triaxial shearing. These strains were not imposed directly on the specimen. Alternatively, the specimen was subjected to follow prespecified stress paths in compression and extension during which it suffered the desired levels of strains. Therefore, for each level of strain to occur, the user has to enter sets of stress path coordinates from the keyboard of the computer. As with the stress path control sub-program, the controlling stress component for each level of strain to occur is first chosen which will always be the deviator stress as the radial stress has been kept constant during the application of strain paths. After the application of deviator stress increment and necessary adjustment of radial stress, if required, the current overall strain (computed 164 from the reading of external displacement transducer) occurring in the specimen due to deviator stress increment is compared with the desired level of strain.zyxwvutsrqponm If the current strain is less than or equal to the desired value of strain, another increment of the deviator stress will then be applied and so on until the current strain just exceeds the desired level of strain. A sequence of priorities is normally followed between each two successive increments of deviator stress applications. This includes checking the intervals for data plotting and printing and rate of' testing. Fig. 4.25 illustrates the flow chart for the strain path control sub-program. 4.9 OEDOMETER TESTS Two tests were carried out using the standard oedometer, .The sample ring was 76.2 mm internal diameter and 19.1 mm high. Porous stones were used to provide drainage from both top and bottom of the specimen. To trim a specimen, initially a small slab of clay was obtained from a block sample. The sample ring, its internal surface well covered with silicon grease, was gradually and in stages pushed into the clay, which was continuously being trimmed away from the cutting edge of the ring with a knife. Following trimming, the specimen was weighed and then set up. The tests were carried out in accordance with the procedure standardized in B.S. 1377 : 1975. A stress increment ratio of 1 (i.e., a load ratio of 2) was used. The vertical consolidation stresses applied in each test were 50 kPa, 100 kPa, 200 kPa, 400 kPa and 800 kPa. The specimens were also allowed to swell under stresses of 400 kPa, 200 kPa, 100 kPa and 50 kPa. Duration of each load step was approximately 24 hours. For each load increment settlement was recorded by a dial gauge at specified intervals of time. 165 Table 4.1 Summary of calibration performance of different magnet system used in axial gauges Magnet Full Linear Resolu- size scale range tion output Gauge length Max. Max. strain linear at 1 mV regre- output ssion error (V) (mm) (%) (% linear (mm)zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHG (urn) range) 2x2x1mm 2.68 2.5 0.93 70 3.6 ±O.71 2 x 2 x 4 mm 2.22 2.2 0.99 70 3.1 ±O.62 2 x 3x 7 mm 3.21 2.2 0.69 70 3.1 ±O.74 6.2 x 6.2 x 8 mm 0.49 5.1 10.41 70 7.3 ±0.87 3 x 3 x 9.6 mm 0.44 6.0 13.64 70 8.6 ±O.72 6.35 x 6.35 mm 0.53 2.2 4.15 70 3.1 ±0.93 7.25 x 10.2 mm 1.51 5.7 3.77 70 8.1 ±O.57 6 x 6 x 8 mm 1.89 7.0 1.89 70 10 ±O.42 166 j zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLK 90zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPON / i I )zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJ 80zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA ,V ' . , i~ ~ au 70 60 I 50 1/ I I . I ,I I i ! /. J I : !, : , ! I i I 40 30 20 1 ! : ~ It! J I ~ %au • i 1 ~I II I I : ! I ! zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONM I I I 10 SIZE mm 0· 002 CLAY Fig. 4.1 0· 006 F I M .. I 0· 2 0-06 F C I MJ SAND SILT 2 0-6 C 1 I 6 F 1 M GRAVEL Grain size distribution curve of London clay at Stag Hill site 167 60 20 1 C J I FIG.zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA 4. 2 PHOTOGRAPH OF THE SO I L MIXER FOR THE PREPARATION OF CLAY SLURRY ,zyxwvutsrqponmlkjihg a FIG. 4. 3 DIFFERENT a) THICK b c TYPES OF BLADES USEDzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA 1 N THE SO 1 L M 1 XER b) MEDIUM THICK c) THlN 168 4) b.OzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA ::s tU b.O e::s Cl) Cl) J: -... ~ 8 CIl ... ~~ ...~ .!! - ~ .!:I -e ~ Cl) Cl) e e-a ~zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA 169 FIG. 4.5 PHOTOGRAPH SHOWING THE DETAIL ARRANGEMENTS DURING CONSOLIDATION OF CLAY SLURRYzyxwvutsrqponmlkjihgfedcbaZYXWVUTS no O~------------------------------------------Izyxwvutsrqponmlkjihgfe - lJJ 0:::zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONML -t- ....J lJJ l.!) z < :J: U lJJ ::E ::J ....J o >zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA TIME (MINUTE) Fig. 4.6 Volume change versus log. time curve during consolidation of slurry 171 zyxw t..:) z ._, _J Q.._ :::E -c U) ~ u a _J CD ~ a 1L.. >- CJ -c w ~ >- <: _J u z a CJ z o _J l- u, o (.f) o W I- -c o _J o (.f) z o u . '<:t t..:) ._, 1L.. 172 :3: W ......, > ~ w _J f-c U') 0::: >U') W Z w :r: L.:l f- w CL f- z ......, <: :r: <: L.:l z ........ 0::: :3: a :r: U') :r: CL -c fU') U') U') w 0::: fU') 0::: W L.:l a :r: ffa :r: LL CL a . . CD ~ L.:l ........ LL 173zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONML - ~ M caEo U 4)"'d -- zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA .... tU c:: ':::0 0 .... :::l~ 0 zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA 10 .... r::Q~ 1 1 I I "7 1 1 1 r+: 1----- , 1 I I CIlI ~: rl U >'Irrt ~I I L ~ --~-CIl I I l_ I I I I .J -'I' ~,'" 'JI,I-" I ", "',I':'-:""r" 10 EIO r r ca~ ,_ II Eo I ca '80 c:: bI:)._._ ._~:a3 .... 8 1 rI ~------------ ~ I f - I I ~ a ~ I I I I I I I I I I ~ ,---r CQ ~ ~~~ II I ~~ I'" 9op-eeel OD g;I; ~B:5.10 u tU L _____ ~ II= !r 8 I - 9 I ~....... 0 -a ~ I >0 I r---t I I I -.:t I ~--- I ~ 1 II I I I 00 I S?,,~ " ~ 0", 00 :U :HSo i " I It_ I '" ~ II L---....J~l ~~ ~ I = El o~ .;;; .~ = I I £ I I I -.:t ~~----------, \ I 1 .... 00 00 I ~ I -=bJ> =' r--&..""-'--'--;::n..L.wLl..L ~~ ~ ~ i-~c..: hr I !=::::::::::J "'l: -zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA 1III'I'I~III1""" V~~~ 'ri\~ ~~ __""~ If ~ I II a >l !;! ~1-1'0 '0 - ·2 ~ II =0 s: ~ I I I I 1 ~--J I --J 174 zyxwvutsrqponm ~oo zyxwvutsrqponm ~ ....o 0 0 CD ......... rIl (,) .c:: rIl ~ ..... (,) 0 - n, 0 0 to ..l::: Cl IJJ ._ < a:: IJJ Z IJJ (.!) IJJ 0 0 ~ a:: :::J U') U') IJJ a:: Q. z ..... c < c ...J a o o o .... N c: 0 '::1 ~ ....~~ .c 0 ~ ~ := g c: 0 (,) e = ePo rIl rIl ....~Pozyxwvu (,) t: t.:) t.:) ~ ..c: (,) z ..... - 0 c < c 0 0 N ...J z ::J ~ bil ~ .... + o o o c c c c o c o o o • c o o CD o to N ....zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA 0311ddV Sd31S o ~O ~38HnNzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONM 175 5000 DzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCB BEST FIT ECUN. ISzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGF LOADING y. J.308 + J. 000 X C.COEF.· .99999 + UNLOADING 4000 RESOLUTION· 1zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA H/8IT " 3000 1Il I- ..... CD ...... I- ::J Il.. I- ::J 0 2000 1000 o~----~----~------~----~----~~----~----~----~ 3000 4000 ]000 2000 o APPLIED LOAD (N) (a) Max. linea~ ~e9~ession e~~o~ (X Ot tull-scale output> • +/- 0.06 i D e5 -------- UNLOADING ~----'~,----~------------~------------+-------------~------------~--------~~--------~~,~----~ , ' ... __~ __ oO--'"' " lLI , ~ !!§ UJ LOADING , ~ ~Ul --- t.-- .. ---a---I( -1" z :J :' " ' , "'ls---a---o---~ -2 p- __ .. ",, , II ---e---"' 'I ,, I' " I 'es -3~----~---------~------~------~~------~---------~---------------~ ..000 3000 2000 1000 o LOAD (N) Cb) Fig. 4.11 Typical calibration characteristics of the load cell: (a) Calibration curve (b) Linear regression error plot 176 r---------------------------------------zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONM ~g zyxwvutsrqponmlkjihgf (0 ~ .....u QI .....u QI :>.. ,- U U +l "tJ CJ CJ ./ :>.. 4.) u .,.I / - -l!) l!) Z Z Cl Cl Cl Cl Cl Z _J D < Z ::J ~ X r rI ,I - X => (J) (J) CJ CJ ('T) lLJ 0:: a... ....J ....J lLJ U c:J lLJ , 'It 1 CJ CJ I ,» 0 a... 0:: ,..1 1 , - lLJ 11 ::J • + CJ CJ 1 _J _J ,; / < < _J ,., z l!) Z Cl < Cl - I zyxw xtI .....c:: NzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA 1 ..... Cl ~ p. ..... lfl c (I) l!) =' en en N I ........J a... a... < ~ c:: ~ ..s::: u ~en c:: ';$ b.O ~ ..... 4.) u 'i0 zyxwvutsrqponm 4.) oS '(g en - ..... U en '1:: s ~ ..s::: u 5 .:: ._e .0 1 ~ I CJ CJ 'f ..... a - M ot:i ch u: 4c U') c CJ CJ 0 0 lfl I ....., CJ &n ..... IzyxwvutsrqponmlkjihgfedcbaZ eN) 113J OVal NO 3J~O~zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONML 177 5000 4000 r-----------------------------------------------------~ BEST FIT ECUN. IS Cl LOADING y. 3.273zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA + 3.999 X + UNLOADING C.COEF.· .99999zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA RESOLUTION· 0.25 kPa/8IT '" ~ 3000 ..... ....., al ~ ::J CL. ~ 2000 o 1000 200 800 600 400 APPLlED PRESSURE 1000 CkPo) (0) 1.00r-------------------------------------------------------~ Max. linaar ragrassian arrar CX of full-scala output> • +/- 0.08 .75 ----- ""' cf • So @j ~ .25 o LOADING --------UNlOADING L&J ~ ..... U1 U1 L&J 13 ~ -.25 s ~ -.50 ...J -.75 -l.DD~--~----~----~----.-~----~--~-----~-----~------.----~ o 200 1000 400 BOO 600 PRESSURE (kPo) (b) Fig. 4.13 Typical calibration characteristics of the cell pressure transducer: (a) Calibration curve (b) Linear regression error plot 178 5000~----------------------------------------------~ BEST FIT ECUN. IS y- 3.273 + 4.001 X D LOAOING + UNLOADING C.COEF.- .99999 4000zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONM RESOLUTION - 0.25 kPa/BIT " ~ 3000 ..... CDzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA ...., ..... :::J n, ~ 2000 Cl 1000 0~0~---~------2~0-0---~~----4~0~0------~---~6=0=D------~---~B~D~0------~-- APPLI EO PRESSURE (kPo) (a) 1. 00 Max. linear regression error output) -zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQ +1- O.oB (X of full-scala .75 ----- LOADING I ,... I I 0 • SO I -------- UNLOADINGzyxwvutsrqponmlkjihgfedcbaZYXW o,zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA I , x ...., , I I I I 0:: Cl 0:: 0:: I .25 I I I UJ ,"Ji Z C ..... 0.00 tn \ tn UJ 0:: CJ UJ 0:: < _J ,~ \ \ -.25 0:: UJ Z \ \ ' / zyxwvutsrqponmlkjihgfedcba " 'tt------- .. " "''III., -. SO / & zyxwvutsrqponmlkjihgfedcbaZYXWVUT ----e" ',~-------.------- -.75 -1. DO 0 200 400 600 900 PRESSURE (kPo) (b) Fig. 4.14 Typical calibration characteristics of the back pressure transducer: (a) Calibration curve (b) Linear regression error plot 179 1000 1000zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA ><zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA BEST FIT EQUN. IS Cl y. -.205+ 1. 001 X CD C.COEF ••• 99999 I::J 800 Cl D LOADING + UNLOADING Cl < UJ 0:: UJ 0:: ::J CJl CJl UJ 600 0:: a.. UJ 0:: Cl n, >CD 400 Cl < UJ 0:: UJ 0:: ::J CJl CJl UJ 200 0:: n, BOO 600 400 200 1000 APPLIED PRESSURE (kPo) (0) 1.5---------------------------------------------------.zyxwvutsrqponmlkjihgfedc Max. linear regression error (X of full-scala output) • 1.0 !3 --- LOADING -------- UNLOADING +/- 0.10 .5 ffi % Cl ..... 0.0 ~----+-~~~,----~------~----+-----~-----+----~;---~~~-.M/ zyxwvu CJl CJl UJ "\ 0:: l.:I UJ 0:: 0:: , ", \ -.5 < UJ .er..... " .............. ............ ,......... '...... I " \,' I I " ,"zyxwvuts , '" ...J " ", , r ........ ............ ..... \ " '" Z ........,'........... ........ ......q,'zyxwvutsrqponmlkji , \zyxwvutsrqponm ,, , II I lit -1.0 -1.S0~---~-------2~OO------~------4~OO------~------6~O-D-----~----B~D-O------~----1~DDD PRESSURE (kPa) (b) Fig. 4.15 Typical calibration characteristics of the pore pressure readout box: (a) Calibration curve (b) Linear regression error plot 180 5000 ~------------------------------------------------------------------------, BEST FIT EQUN. IS y- 2.238 + JS7.SJ4 X 4000 C.COEF'.- .99999 LINEARRANGE - 25.4 Mm (1 inch) ,...zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA RESOLUTION - 0.00635 mm/BIT ~ 3000 .....zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA CD ..., I- ::J a, ;; 2000 Cl 1000 o~------------._---------~---------~--------~--------~---__~ 30 o 5 10 15 20 25 DISPLACEMENT (mm) (a) 30 ~------------------------------------------------------------. MaximUM linear regression error (X of lineor range) • +/- 0.077 ,... c 20 C L U .... E ..., 10 0:: Cl 0:: 0::zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA W Z Cl ..... 0 tn tn LLI 0:: ~ LLI 0:: -10 0:: < LLI Z ..... _, -20 -30~------~--------~------~---o 5 10 ~ 20 15 DISPLACEMENT ~ 25 __' 30 (mm) Cb) Fig. 4.16 Typical calibration characteristics of the external displacement transducer: (a) Calibration curve (b) Linear regression error plot 181 Hall effect sensor Electrical cable Magnet Pole piece ~ Motion __.---::, of magnet (a) Electrical cable Hall effect sensor Pole piece Motion of magnets (b) Fig. 4.17 Configurations ofzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA Hall effect sensor-magnet-pole piece system: (a) Double magnet, bi-polar slide-by, with one pole piece (b) Single magnet, bi-polar slide-by, with two pole pieces 182 Membrane c ....8zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA c.J 8- en Hall effect -H:--+--iI sensor .-- Bar magnet separator .._..__..t:~~h- PTFE Adjustment screw Fixing pin F------Electrlcal cable Fig. 4.18 Design of the Hall effect local axial strain measuring device 183 5000 BEST FJT EIJUN.JSzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCB D Y--Ss24.247 + 959.631zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA X C.COEF.- .99993 4000 3000 ,.... IJl ._ ..... ~ ._ 2000 LINEAR RANGE - 7.0 mm RESOLUTION· 0.00]045 mm/BIT 1000 ] BIT - 0.0002B VOLT 0 :::J a.. I- 5 -1000 -2000 -3000 D D D D D D -4000 -50000L----L----2~---3~--~4----~5----~6----L7----LB----~9----]~0~--]~]~--7]2 RELATIVE DISPLACEMENT OF GAUGE ENDS (mm) (a) 50 Maximum linear regression error ex of linear range) • +/- 0.57 40 ,.... c 30 0 L U .... .....E 20 a:: Cl a:: a:: ]0 z 0 UI Cl ...... IJl IJl I.LI a:: l!) -10 UI a:: a:: < UI Z ...... ....J -20 -30 -40 -50 0 2 3 4 5 6 DISPLACEMENT 7 B 9 10 11 12 (mm) (b) Fig. 4.19 Typical calibration characteristics of the Hall effect local axial strain gauge: (a) Calibration curve (b) Linear regression error plot 184 Spring-loaded hinge Polished hinge-pin Radiused pads 102zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHG mm A A Ca) Brass container Aluminium ring Hall effect sensor Sliding block Bar magnet Cb) Fig. 4.20 Design of the Hall effect radial strain measuring device: (a) Plan Cb) Section A-A 18S SOOO BEST FIT EQUN. IS 4000 Y·-5455.254 ...1904.238 X C.COEF.- .99995 3000 LINEAR RANGE· 3.4 mm RESOLUTION· 0.000525 mm/BITzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFED 2000 '"' tn 1000zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA IJ BIT • .000285 VOLT ..... ._, III l::J Il... l- D S -1000 -2000 -3000 00 000000 -4000 -5000 0 2 3 5 4 RELATIVE DISPLACEMENT s 7 (mm) (a) 30 ~~----------------------------------------------------, MaxIMum linear regres.ion error (% of linear range) • ... /- 0.54 ,... 20 c 0 L U ... .... E ID ~ c ~ ~ L&J z: c .... 0 til til L&J ~ l.!I L&J ~ ~ -10 < LIJ Z .... _, -20 -30 ~ o L- ~ -L 2 3 -L 4 DISPLACEMENT ~ L- 5 6 (mm) (b) Fig. 4.21 Typical calibration characteristics of the Hall effect Caliper: (a) Calibration curve (b) Linear regression error plot 186 ~ 7 3000zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA 0 LOADING BEST FIT EDUN. IS y. S.SS9zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA 3.999 XzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA + C.CDEF.· .99999 + UNLOADING 2500 RESOLUTION • 0.2S kPa/BIT 2000 ,... U1 I- ..... CD '-' I:::l 0... I:::l 1500 0 1000 APPLIED PRESSURE (kPa) Ca) 1.5 ~----------------------------------------------------~ Max. linear re9ression error CX of full-scale output> • +/- 0.23 1.0 \ \ ,.... \ \ a --LOADING -------- UNLOADING '.\ \ ~ ~ , \ \ \ \ .5 \ \ \ \ IE UJ \ \ \ \ z: .... o ,, ,, ,,, ,, ,,, , ,,, ,, ,, \ U1 U1 ~ ~ UJ a: ~ UJ z: ..... -.5 ...J -1.0 -1.5 o~------------~----------~----------------~------------~------------~--------------~----100 200 300 400 500 600 700 PRESSURE (kPa) Cb) Fig. 4.22 Typical calibration characteristics of the miniature pore pressure transducer: (a) Calibration curve (b) Linear regression error plot 187 ( START)zyxwvutsrqponmlkjihgfedcbaZYXWVU t INPUT INFORMATION DO INITIAL CALCULATIONS INITIALIZE CONTROLLERS MAKE DEVIATOR STRESS ZERO r---------II 'II READ DATA SET UP INITIAL STRESSES It-- ---, V ~ II zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHG STRESS/STRAIN PATH SPECIAL " TEST CONTROL FUNCTION SUBROUTINESzyxwvutsrqponmlkjihgfedc JI' " 1/ STORE DATA IN DISC \zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGF V BRING ALL STRESSES DOWN " TURN OFF CONTROLLERS ,,/ (END) Fig. 4.23 Flow chan showing general program layout 188 DEFINE STRESS PATH DIRECTION CORRECT CORRECT OTHER BACK STRESS COMPONENT PRESSURE TO MAINTAIN THE STRESS PATHzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA CALCULATE VALUE OF OTHER STRESS COMPONENT TO IE---...I MAINTAIN THE STRESS PATH N N N y Fig. 4.24 Flow chart for stress path control sub-program 189 SELECT CONTROLLING STRESS COMPONENT FOR STRAIN PATH tE;:o------- ... DEFINE STRESS PATH DIRECTION CORRECT OTHER CALCULATE VALUE OF OTHER STRESS COMPONENT STRESS COMPONENT TO TO MAINTAIN THE MAINTAIN THE STRESS PATH STRESS PATHzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA N N N y Fig. 4.25 Flow chart for strain path control sub-program 190 CHAPTER 5 STRESS AND STRAIN PATH TESTS 5.1 DETAILS OF TESTING PROGRAMME All the tests included in the testing programme were carried out with the use of the automated stress path test equipment, incorporating devices for monitoring local deformations and porewater pressures. All tests were conducted on specimens of reconstituted soft London Clay, 102 mm dia. x 203 mm high nominal dimensions, prepared by trimming from blocks on a soil lathe. densities of the specimens were respectively 45 ± The water contents and bulk 1% zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONM and 1.76 ± 0.01 Mg/m', The tests were all stress controlled. Due to considerable amount of time spent in carrying out analytical work (see Chapter 3) and during development of the instrumentation and software (see Chapter 4), it was essential to obtain the required test results with a minimum number of tests. The testing programme consisted of two series of tests as described in the following. (1) The first series of tests were conventional undrained triaxial compression and extension tests on Ko-normally consolidated specimens. These tests were carried out to determine the reference "undisturbed" normally consolidated behaviour of the soil which is not subjected to any disturbance. In these tests, the specimens were first brought back to their "in-situ" stresses from their initial set up stresses (which are isotropic) by performing a short undrained stress path at constant radial stresses and at reasonably fast rates (48 kPa/hr). During these paths excess pore pressures (25 to 28 kPa) were generated and in order to dissipate the excess pore pressure, the specimens were allowed to consolidate under the "in-situ" stresses. The "in-situ" stresses are, in fact, the effective vertical and horizontal stresses in the reconstituted sample in the tank at the end of the consolidation period. Consolidation, during the dissipation of excess pore pressure, was done under a back pressure of 250 kPa. The use of an elevated back pressure to produce complete saturation in various types of laboratory test specimens has been well established and widely used (Bishop and Henkel, 1962; Lowe and Johnson, 1960; Lowe et al, 1964; Black and Kenneth, 1973; 191 Brand, 1975). It took 20 to 24 hours for complete dissipation of excess pore pressure which was indicated when the mid-height pore pressure was equal to back pressure (250 kPa) at the base of the specimen. The specimens were then consolidated anisotropically under Ko-condition to effective vertical stresses equal to 1.75 to 2 times the previous maximum vertical effective stress to eliminate the effects of sampling disturbance. During Ko-consolidation, the back pressure was also set at 250 kPa and the vertical effective stress was increased at approximately 0.7 kPa/hr. Finally, at the end of consolidation, the specimens were sheared up to failure in undrained compression and extension by performing special stress path tests with constant radial stress. During shearing in compression and extension, the deviator stress changes were 10 kPa/hr. (2) The tests included in the second series were carried out to investigate the effects of tube penetration disturbances on the subsequent undrained shear stress-strain, stiffness and strength properties. The method adopted for these tests was basically similar to that proposed by Baligh et al (1987). Specimens were first brought back to their "in-situ" stresses, consolidated under "in-situ" stresses to dissipate excess pore pressures and reconsolidated under K,,-condition to effective vertical stresses of 180 kPa to 190 kPa. The specimens were then subjected to follow specified undrained stress paths in compression and extension so that the specimens suffered the prespecified tube penetration disturbances. These include the maximum strain during the first compression phase (eJ, the maximum strain during the extension phase (eJ and the minimum strain during the second compression phase (EmoJ. Finally, the specimens were sheared up to failure in undrained compression. Radial stresses were kept constant during the stress paths applied to model tube penetration disturbances and also during shearing to failure. The summary of the tests included in this series is presented in Table. 5.1. The rate of increase (or decrease) of deviator stresses for the stress paths applied to impose the maximum strain during the first compression phase, maximum strain during the extension phase, minimum strain during the second compression phase and shearing up to failure in compression were respectivelyzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA 4-5 kPa/hr, 50 kPa/hr, 50 kPa/hr and 10 kPa/hr. However, the rate of increase of deviator stress for the stress path applied to impose maximum strain during the first compression phase in case of test 3 (see Table 5.1) was 50 kPa/hr. 192 During the consolidation stage of each series of test described above, drainage was restricted from the base of the specimens and also from the radial boundary. Drainage from the radial boundary of the specimens was effected by providing vertical filter paper strips placed between the surface of the specimens and the impermeable membranes. Filter strips used for radial drainage reduce the time required for full dissipation of the pore pressure during the consolidation stage in case of soils of low permeability (Bishop and Henkel, 1962; "Bishop and Gibson, 1963). The equalisation of pore pressures within an undrained specimen is also accelerated by the use of filter strips (Bishop and Henkel, 1962). Atkinson et al (1985). however, have shown that radial drainage produces some non-uniformity in the specimens. If radial drainage is necessary. then consolidation should be carried out at a slower rate of loading to reduce the non-uniform conditions. 5.2 THEORETICAL INVESTIGATIONS OF THE TEST RATES FOR Ko-CONSOLIDATION Some theoretical analyses were carried out in order to find out an approximate rate of testing during the Ko-consolidation stage under continuous loading. Terzaghi's (1948) one-dimensional consolidation theory can be solved numerically by the method of finite differences (Scott, 1963). The method is based on the depth-time grid as shown in Fig. 5.1. The oedometer sample is modelled by dividing the depth into equal parts of thickness,zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA Az and dividing time into equal intervals, AT (time factor). The value of excess pore pressure at any depth after any time is denoted by u1J' where i and j are subscripts denoting depth and time respectively. This reasonably accurate approximation was used to model the theoretical consolidation behaviour of a soil subjected to transient continuous loading. A computer program (written in Fortran 77), developed by Hopper (1988), was used to compute the theoretical percent dissipation of pore pressure at the mid-plane as a function of time factor, for continuous loading. Using these data a theoretical relationship was derived between the percent dissipation of pore pressure at mid-plane and testing time, for a London Clay specimen (102 mm dia. x 203 mm high, Cv = 0.25 m2/yr) subjected to consolidate under continuous loading with drainage permitted from base and radial boundary of the specimen. This relationship. presented in Fig. 5.2, was subsequently used to compute the theoretical pore pressures at the mid-plane as a function of effective vertical consolidation stresses. The excess pore pressures at the mid-plane 193 of the specimen were estimated for four different rates of increase of effective vertical stress. The theoretical relationships between the excess pore pressure at midplane of the specimen and the applied vertical effective consolidation stress for various rates are presented in Fig. 5.3. The curves in Fig. 5.3 show that the excess pore pressure generated at the mid-plane of the specimen increases with the increase of rate of loading. These curves also demonstrate that at the initial stages of loading there is a sharp increase in excess pore pressure which· eventually decays and becomes approximately constant at higher consolidation stresses. Fig. 5.3 provided Koa valuable basis for the selection of an appropriate rate of loading duringzyxwvutsrqponmlkjihgfedcb consolidation of the specimen. 5.3 TEST PROCEDURE 5.3.1 PREPARATION AND SET-UP OF SPECIMEN Each test specimen was prepared from a block sample (approximately 180 mm x 180 mm x 280 mm high) by trimming it in a hand operated soil lathe. The soil lathe. specially designed for preparing 102 mm (4 inch) diameter specimens from large block samples. was manufactured by Soiltech Ltd. A photograph of the soil lathe together with the wire saw and two-part split former used for specimen preparation is shown in Fig. 5.4. Before mounting in the soil lathe. a block sample was first trimmed roughly to shape. using a sharp blade and wire saw. The length was somewhat longer than the required specimen height (Le.•203 mm), and the end faces trimmed flat and parallel. These faces were mounted between the platens of the lathe, and the upper platen was brought firmly into contact with the upper surface and locked into position so that the sample was securely held. Surplus material was removed progressively from the sample by means of a series of fine vertical cuts. rotating it slightly between each cut. Tendency to cause distortion by dragging at the sample was avoided. Very small stones, if any. were removed carefully, and resulting cavities were filled with material from the parings. The specimen was trimmed accurately to the final diameter by using the frame of the lathe as a guide for the wire saw while making the last few cuts. The specimen was rotated slightly between each cut, until a smooth cylindrical surface was obtained. It took approximately two hours to prepare 194 a specimen by this method. A photograph of a specimen trimmed to its final diameter in the soil lathe is shown in Fig. 5.5. The specimen was then carefully removed from the soil lathe. It was placed in a two-part split former and the ends were trimmed to give the correct height, using the wire saw. The split former was removed and the height and diameter of the specimen were measured. The weight of the specimen was also determined. At this stage, all system controllers and the peripheral devices were switched on and the program for scanning all the measuring devices (see Appendix-B) was loaded and run. Water was flushed through the tubes supplying cell and back pressures in order to get rid of all the entrapped air in these tubes. The readings monitored by the back and cell pressure transducers when left open to atmosphere were made zero bits, if required. This was done by adjusting the zero knob in the signal conditioning unit for the respective devices. pedestal. A saturated porous stone was placed on the bottom A soaked filter paper was then placed on top of the porous stone and water was allowed to flood over the filter paper. The specimen was placed very carefully on the filter paper. Another soaked filter paper was placed on top of the specimen and a saturated porous stone was placed on top of the filter paper. Vertical filter strips were soaked with water and placed gently around the surface of the specimen, overlapping the bottom and top porous stones. The rubber membrane, stretched along a cylinder, was released around the specimen in the usual way. The membrane was sealed perfectly by placing three O-rings around the bottom pedestal and three O-rings around the top cap. 5.3.2 INSTALLATION PROCEDURE FOR THE MINIATURE PORE PRESSURE TRANSDUCER The porous stone and the cavity between the stone and sensor were initially deaired. This was achieved by immersing the transducer in de-aired water and applying vacuum to the water for several hours. Complete de-airing, however, can notzyxwvutsrqponmlkjih be guaranteed as there is no facility for flushing water through the porous stone or cavity. In view of the small volume of water in the cavity, and hence the limit to the volume of the gas which can be dissolved, the transducer was always kept saturated by immersing in a pot of de-aired water when not in use. 195 A special three-part brass mould was designed to prepare a rubber grommet for housing the transducer. The rubber grommet was prepared from a mixture of Dow Coring JRTV curing agent and Dow Coring JRTV silicone rubberzyxwvutsrqponmlkjihgfedcba (1 to 3 parts by weight) in the brass mould. For installation, a very small hole (4 to 5 mm diameter) was cut in the latex rubber membrane at its mid-height with a pair of scissors. The flat base of the rubber grommet was inserted through the hole by stretching the hole with the help of fingers and the base was kept in contact with the side of the specimen. In order to ensure intimate contact between the specimen periphery and the porous stone of the transducer, a pad of soft saturated kaolin was placed on the specimen periphery prior to the installation of the transducer. The transducer was pushed through the annular tube of the rubber grommet and put into intimate contact with the kaolin pad on the specimen. Any penetration was avoided in order to reduce the interference effects. The whole assembly was sealed with two O-rings placed around the annular tube of the rubber grommet. For additional security against leakage, latex rubber solution was painted around the installed transducer. Air is inevitably trapped during installation of the transducer. Furthermore, in specimens with relatively high suction, cavitation may occur in the cavity behind the porous stone. Gases in the cavity reduce response time and, if the volume is sufficiently large, can cause the measured pore pressure to be higher than the actual porewater pressure. As gases can not be flushed out, it is necessary to use a high back pressure when working with the transducer. Fig. 5.6 shows the photograph of the assembled pore pressure transducer on a 102 mm dia. x 203 mm high soft London Clay specimen. 5.3.3 MOUNTING PROCEDURE OF LOCAL AXIAL STRAIN DEVICES AND CALIPER After the installation of the local miniature pore pressure transducer, the two Hall effect local axial strain devices and the lateral caliper were mounted according to the procedure described below. With the help of a square, two vertical lines were drawn along the height of the specimen. Each line is diametrically opposite to the other and defined the longitudinal position of the two axial gauges. A gauge length of 70 mm was measured along the middle third of each side and cross marks were drawn to the 196 exact position of the pin holes on the lower and upper pads of the gauges. A thin layer of contact adhesive was applied to the lower pads and their positions on the membrane surface. The contact adhesive was allowed to become tacky for two minutes. Each lower pad was then pushed carefully towards the specimen in order to ensure proper contact of the pad with the rubber membrane. A thin layer of contact adhesive was then applied to top pads, holding the arms and magnets of the gauges, and their respective positions on the rubber membrane surface. After allowing appropriate time for the adhesive to become tacky, each upper pad was placed carefully on its position. While placing the lower and upper pads care was taken so that the pin holes on the lower and upper pads were in alignment with the cross marks drawn previously. For each pad, two sharp pins were inserted into the specimen through the rubber membrane in order to ensure intimate contact of the pads with the specimen. Hall effect semiconductors were placed in their positions provided in the lower pads. Some time (approximately half an hour) was allowed for the adhesive to set properly. Thin layers of latex solution were applied around the upper and lower pads to seal against leakage. After placing both the gauges, the caliper was placed slightly above the mid-height of the specimen. In order to avoid slippage of the caliper during the test, pads of the caliper were put into intimate contact with the specimen by placing strips of adhesive tapes over the pads. The latex solution was left several hours for setting. During all the tests carried out in the present research, the latex was allowed to cure for at least twelve hours. A photograph showing the set-up of all the local devices on a specimen is shown in Fig. 5.7. 5.3.4 TEST SET -UP AND EXECUTION The disc containing the programs for performing stress or strain path test and the special utility binary program used for fast labelling was put inside drive "0" of the disc drive. The computer program for carrying out stress or strain path test was loaded and the "RUN" key on the keyboard of the computer was pressed. The binary program was loaded automatically and a space in the electronic disc of the computer was allocated for later storing of the graphics, at the end of the test. The user was then directed to perform the required operations by pressing the two special function keys ("K8" and "KI2") on the keyboard of the computer. Key "K8" was 197 first pressed to read and display the signals from all the measuring devices. The Hall effect local strain measuring gauges were adjusted to achieve the desired position in the linear range of the devices. The cell top was then placed carefully in position and the load cell was screwed into the top cap. The external displacement transducer was positioned and the cell was filled with de-aired water. Finally, the test wa~ started by pressing the key "K12". The steps followed during the execution of the test are summarised in Fig. 5.8. 5.4 PROCESSING AND PLOTTING OF TEST DATA The typical printed output obtained during performing a test (Test no. 7 in Table 5.1) in the automated stress/strain path equipment is presented in Fig. 5.9. At the end of the test these data are also stored in the same format in a disc. It can be seen that in the first four columns, the data are given in engineering units while in the following four columns they are given in digital units (bits). The radial and axial stresses in the first two columns are calculated by the computer program from the signals of cell and back pressure transducers and the load cell using calibration factors prespecified in the program. Local pore pressure and external axial strain are also computed by the computer from the signals of the respective devices using appropriate calibration factors prespecified in the program. Two separate programs were written. One for the processing and analysing the data (Data Processing Program) of each test and the other for plotting graphs (Plotting Program) with the processed data. Both the programs were written in BASIC. The Data Processing Program converts the data into different parameters and these parameters can then be stored into various two dimensional arrays. Any of the following parameters can be stored against each other, selecting either as the ordinate or abscissa which can be plotted later on using the plotting program. 1 Radial effective stress, ctzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIH r (kPa) 2 Axial effective stress, ct. (kPa) 3 External axial strain (%) 4 Local mid-height porewater pressure (kPa) 5 Local radial strain (%) 6 Local axial strain for gauge 13 (%) 198 7 Local axial strain for gauge 12 (%) 8 Deviator stress, q (kPa) 9 Average local axial strain (%) 10 Mean normal effective stress, p' (kPa) 11 Stress ratio (a'Ja',) zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA 12 Shear strain (%) 13 Stress ratio (q'/p') 14 Local volumetric strain (%) 15 MIT effective stress parameter, s' (kPa) 16 MIT effective stress parameter, t' (kPa) 17 Change in mid-height pore pressure,zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFED dU, (kPa) 18 Skempton's pore pressure parameter A Radial and axial effective stresses have been calculated on the basis of mid-height pore pressure values. Deviator stresses in undrained stages were corrected using the conventional method (Bishop and Henkel, 1962) and also using actual diameter at mid-height monitored by the caliper. The sign convention used for strains is that all strains leading to a reduction in height, diameter and volume were considered to be positive. Following this convention, the shear strain (e.) and volumetric strain (I;,) were calculated according to the following equations: £.=2/3(£.-£,) ....(5.1) £.-=£.+2£, ....(5.2) where, Ea = average local axial strain (%), and e, = local radial strain (%) The Cambridge effective stress parameters (q', p') and the MIT effective stress parameters (t',s') were computed using the following equations: . q' = a' - a', .... (5.3) p' = 1/3 (a'. + 20",) ....(5.4) t' = 1/2 (a'. -zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA a/,) ....(5.5) s' = 1/2 (0'. + a',) ....(5.6) 199 Table 5.1 Test Summary of tests included in the testing programme Applied tube penetration disturbance number e, (%) Eo (%) Emln (%) 1• " UNDISTURBED"zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONML 2 •• " UNDISTURBED 3 2.0 -1.0 0 4 1.0 -1.15 0 5 0.5 -1.44 0 6 0.5 -1.15 -0.5 7 0.5 -0.6 0 8 0.25 -0.27 0 " Note: • sample sheared in compression up to failure •• sample sheared in extension up to failure Ec = Eo maximum strain imposed during the first compression phase = maximum strain imposed during the extension phase Emln = minimum strain imposed during the second compression phase All the reported strains were measured externally 200 I I IzyxwvutsrqponmlkjihgfedcbaZYX t.1T zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA = t/n 1 I "1 ~.----.----. i : .1z = Him I LJ 1-- : : I 1 U I I I I I il o ,:_. _ I Depth: i (0 I Time: j ( 0 sism ) s j s n )zyxwvutsrqponmlkjihgfedcbaZ -!- - - -1-I I I I I I I I I I lUi, I I I I I I I I I I lUi, I --j--- -j-- --j-- - -i-- _ -1I I j j+1 ! _-1----\-u,..~ I I I Finite difference approximation of the one-dimensional consolidation equation: Fig. 5.1 Numerical solution of one-dimensional consolidation theory using finite differences 201 LJJ zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA ~ _J a, I C ..... % Ba ._ < ~ ::J ~ 60 LJJ ~ DLJJ ~ C D- zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA 40 V) V) LJJ U X LJJ u, C ~ ..... ._ 20 < ..... ..... D- V) V) c TIME (HOUR) Fig. 5.2 Excess pore pressure dissipation - time relationship 35 "0 D- ~ '-' w _J 30 D% < V) u, c _____ 25 3.278 kPa/hr. 1.667 kPa/hr. w z < -' D- I Cl 1. 111 kPa/hr. 20 ._ < w ~ ::J en en w ~ n, zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA ------------ 15 10 0.833 kPa/hr. __ - __ % 1/' "'...--------- - -- -- - ---------------------zyxwvutsrqponmlkjihgfe W t/..."....---_.------------------- C D- I~"" ~ U') U') 5 w u >< w o o~--~----~----._--~~--~----~----~--~----~----._--~ 100 20 40 60 zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA BD 120 HO 160 IBO 200 VERTICAL EFFECTIVE CONSOLIDATION STRESS (kPa) Fig. 5.3 Excess pore pressure generation for different rates of loading 202 220 0 Wzyxwvutsrqponmlkji ::::E: ::::E: ........ a::: I- z: W ,....._ E E ::::E: ...... U W Q_ N tn 0 _j -.....- ....... -c u ,......., a::: Q_ I- w >- W I- ::::E: -< -c ...... 0 L1.. 0 _j -c Z ......, I CL -c L1.. 0 l- a::: L:)zyxwvutsrqponmlk c.n ......, I- 0 :r: 0 CL l- . in in L:) ....... LI.. 0 W c.n :::l a::: Wzyxwv L a::: Z 0 0 ....... lI- < ....... a::: _j < CL CL c.n w a::: Li; 0 CL < 0 :3t < en t..:J W ....... Z Z < Z a::: ....... L L ....... :3t a::: I- C\ W :r: < _j I- W _j CL L <: c: _j ....... a::: 0 0 c.n LI.. . ..q- in 203 . t..:J ....... u, FIG. 5.6 PHOTOGRAPH SHOWING THE LOCAL PORE PRESSURE TRANSDUCER INSTALLED AT THE MID-HEIGHT OF A SOFT LONDON CLAY SPECIMENzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQ 204 FIG. zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA 5.7 PHOTOGRAPH SHOWING THE FlNAL SET-UP OF THE LOCAL PORE PRESSURE TRANSDUCER AND LOCAL STRAIN MEASURING DEVICESzyxwvutsrqponmlkjihgfedcbaZYXWVU 205 ~ ENTER ~ INITIAL INFORMATION NAME, DATE .....,. TEST I TEST INFORM ATIONzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPO SOIL TYPE, LEVELS I OF STRAINS FOR STRAIN PATHS ~ SPECIMEN DATA ..., ~ INITIAL STRESS STATE ... STRESS PATH DATA ...., COORDINATES STAGE TEST TIME I I INTERVAL FOR DATA PRINTING ~ ~ MAKE DEVIATOR STRESS EQUAL TO ZERO ~ INCREASE CELL AND BACK PRESSURES INCREMENT ALL Y TOzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA REQUIRED VALUES ~ 'V EXECUTE STRESS PATHS OR IMPOSE LEVELS OF STRAINS FOR ~~ STRAIN PATHS \~ BACK PRESSURE '" EFFECTIVE STRESS ADJUST ACTUATOR PISTON IN DESIRED RANGE BEFORE CONNECTING IT TO RAM " CHECK AND SET-UP THE INITIAL TEST CONDITIONS HEIGHT, DIAMETER GAUGE LENGTHS FOR LOCAL AXIAL STRAIN DEVICES ~ ...,. ~ USE COMPRESSION AND EXTENSION MANUAL PRESSURE REGULATORS I-- PRINT INITIAL INFORMATION AND TEST DATA ""PRINTER DISPLA Y TEST INFORMATION AND PLOT TEST DATA ... ., -, COMPUTER MONITOR STORE DATA IN DISCzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDC \I END OF TEST Fig. 5.8 Stress/strain path test procedurezyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJ 206 ................................................................................................................................................zyxwvutsrqponmlkjihgf PIIlIl TEST STRESSzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLK .................... - - TEST NAME/HO.: SAMPLE zyxwvutsrqponmlkjihgfedcbaZYXWVUTS *"* II II T13 HEIGIH: TEST DATE: 203 SAMPLE MM HE12 GAUGE LENGTH: 69 AREA or SRMPLE: SOIL TYPE: London 0196.937 MM : 69.5 MM Clay 250 kP8 SET AT: ErfECTIUE PRESSURE: SPECIAL COMMEHT: TESTED WIlli SIDE DRAINS TEST TIME 61 kPa 171.B5 ••.•• STRESS 1 2 3 'I 5 6 7 AHD OASE ORAIHAGE DHlY. 1i0URS PATH COORDINATE EffECTIUE AHIAl STRESS (kP..> COORDINATE HUMBER 102.16 Sq.MM IHITIAL ESTIMATED . 13/5/09 OIAMETER: HE13 GAUGE LENGTH MM H-SECTIONAL BACK PRESSURE »1 UALUES ....• EffECTIUE RADIAL STRESS (kP,,) 61 61 100 200 260 30 2BO 350 61 132 132 132 132 132 STAGE TEST TIME ("OURS) 0.00 .50 112.75zyxwvutsrqponmlkjihgfedcbaZY 15.00 1.60 5.00 7.00 ••••.•....••.• TEST DATA ..•.••...•....••• EffECTlOCAL ErrECTIUE RADIAL IUE AHIAL PORE STRESS PRESSURE STRESS (kP.. > (kPa) (kPa> DATA 110. HE HE liE EXTEREHTECALIP. 613 G12 HAL AHIAl RIIAl STRAIII LSCDTzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQ (BITS) (BITS) (BITS) (X) (BITS) 2515 2511 I 366 63.75 0.000 63.75 258.25 422 64.00 .175 69.98 260.00 8" • • 63:50'" • • S·7."I7 •.. 265.25' .. '.792'•••• IiIi, .... 635 .841 64.00 99.50 265.50 9 Tll1E TAKEN BY STRESS PIITH NO. I IS .62 HOURS BIB 249.25 1.414 64.00 10 100.11 825 249.75 1.436 64.75 101.23 II , , •••••• , 1""" 1.,.1 • 2582 6.932 259.25 130.50 197.84 02 2604 7.001 199.07 259.25 1~1.00 B3 2604 259.25 7.001 199.57 131.50 B4 TIME TAKEN BY STRESS PATH NO. 2 IS 146.07 HOURS -.003 1313 199.94 259.50 131.60 B5 1314 0.000 259.75 199.57 131.50 B6 1314 0.000 199.45 260.00 131.50 87 1315 .003 260.00 131.50 199.70 S8 .003 1315 260.25 200.31 131.75 89 1316 .006 260.50 201.05 132.00 90 .463 1462 274.00 131.75 213.85 169 1467 .479 213.98 274.50 131.75 169 .500 1474 274.75 131.75 214.46 170 TIME TAKEN BY STRESS PATH NO. 3 15 3.96 HOURS 1474 .500 214.46 275.00 131.75 171 .510 1477 274.75 212.88 131.75 172 2 .. ... • • t· , 359 ••••• I. ••• 132.25 132.25 132.00 132.00 1.1 1939 1982 IOB9 2614 2622 2622 I t -3135 -3110 '-2743' -2704 -1879 -1966 -2324 -2310 1651 1598 159B 1390 1434 1434 2522 1600 1437 2622 1601 1437 1438 2621zyxwvutsrqponmlkjihgfedcbaZYXWVUTS 1601 2622 1603 1439 1604 1441 2621 2622 1443 1606 ••• t 1"926 1918 2193 IB32 2174 1937 IB48 2152 1951 . .... 2150 2139 1951 1958 1850 195B -.378 -.425 -.500 HOURS -.507 -.604 -.622 1193 1178 1154 3153 3205 3242 1387 1359 1335 1196 1178 1152 1121 1115 324B 32B2 3305 1333 131B 1307 1174 1147 1133 282.50 -.031 191. 17 -.009 282.75 191.8S .003 282.75 192.63 PATH NO. 5 IS 1.66 HOURS .003 283.00 193.25 .025 283.50 183.60 283.75 .044 194.46 .053 193.60 283.50 .066 293.75 194.21 1304 1311 1315 2097 2872 1639 1654 1662 1545 1561 1570 33 40 46 49 53 2833 2810 2789 2779 2765 1677 1693 1706 1713 1589 1605 1620 1722 1627 1637 3.744 4.307 4.911 5.694 1229 :j4IS' 3734 . 3657' -2003 -2547 -3009 3945 3475 3133 2591 135.78 262.00 133.70 262.00 132.85 PATH NO. 4 IS 2.17 262.00 132.85 262.00 132.49 262.00 131.87 t ••• 132.00 282 131.50 283 131.75 284 TIME TAKEN BY STRESS 132.25 285 131. 75 286 132.00 287 131.75 288 131.75 289 355 356 357 ...262.00. . •• 132.00 216 131.50 217 131.75 218 TIME TAKEN BY STRESS 131.75 219 132.00 220 131.50 221 I' , ........................... ..... ... . •• i992' . -2763 -2727 :23 i5 •• -2214 ., • I " .... 223.02 223.26 222.52 aer.ee 223.01 301.00 ...................................... 299.50 300.50 1409 1602 1949 ••• 2861 OATil STORED ON FILE TI3 NO. OF OATA READINGS 360 Fig. 5.9 Typical OUtput of test data 207 1228 3915 .3962 ........... CHAPTER 6 RESUL TS AND DISCUSSIONS 6.1 INTRODUCTION In this chapter, results from finite element analysis and from laboratory experiments will be presented and discussed. Firstly, the predicted strain paths obtained due to undrained penetration of samplers will be presented, and the strain paths obtained from different series of analyses will be compared. Secondly, the one-dimensional consolidation and permeability characteristics of soft London Clay are presented and discussed. Finally, stress and strain path test results will be shown. Undrained stress-strain properties observed for "undisturbed" specimens in compression and extension are discussed and a comparison is made between the two responses. The effect of varying degrees of tube penetration disturbances on the resulting undrained . stress-strain, stiffness, strength and pore pressure characteristics are discussed together with a comparison with previous investigations. The complete set of results from the stress and strain path tests are shown graphically in Appendix-C. 6.2 PREDICTED STRAIN PATHS DUE TO UNDRAINED PENETRATION OF SAMPLERS During axisymmetric undrained penetration of the samplers, strains in the soil elements have been predicted in terms of vertical (axial) strain only. This is because in conventional triaxial tests, vertical strains can be imposed to model the predicted strain paths. No attempt was made to calculate the components of deviatoric strains, e.g., tangential strain, meridional shear strain. Figs. 6.1 and 6.2 show two typical illustrations of the predicted strain paths followed by soil elements at four locations within the sampler tube. In Fig. 6.1, strain paths due to undrained penetration of the NOI 54 mm dia. sampler (ARzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA = 11.4% and ICR = 0.93%) are presented while strain paths due to undrained penetration of a flat-ended sampler (t = 2.5 mm and BIt 23) are presented in Fig. 6.2. The strain paths in Fig. 6.1 show the following: 208 = (i) A soil element is subjected to three distinct phases of undrained triaxial shearing; namely, an initial compression phase ahead of the sampler where axial strain increases from zero to a maximum value; an extension phase near the cutting edge of the sampler where the axial strain reverses from compression to extension and attains a maximum value in extension; and a second compression phase inside the sampler tube where axial strain decreases and attains a constant value. This finding agrees with those reported by Baligh (1985) and Baligh et al (i987). (ii) Peak axial strain in compression ahead of the sampler and peak axial strain in extension inside the sampler are not equal. For example, for the soil element located at a distance of 0.7R, from the centreline of the sampler, the peak axial strainszyxwvutsrqponmlkji in compression and extension are respectively 0.385% and 0.485%. This is because the BIt ratio of the sampler initial compression phase is governed by the thickness andzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIH while the extension phase, especially in the vicinity of cutting shoe, is controlled by the precise geometry of the cutting shoe. This finding, however, contrasts with those reported by Baligh (1985) and Baligh et al (1987) where peak axial strains in compression and extension at the centreline have been found to be equal for Simple samplers (with or without inside clearance). (iii) The magnitude of axial strain varies across the diameter of the sampler. Soil elements located near the inside edge of the sampler are strained much more severely than those near the centreline of the sampler. For example, the peak axial strain in extension for a soil element located at 0.9Rt from the centreline of the sampler is about 3.4 times that for a soil element located at O.IRt from the centreline. Baligh's closed-form solution, however, only estimated the strains on the centreline of the sampler. The typical strain paths shown in Fig. 6.2 for the flat-ended sampler illustrate that only for the initial compressive phase ahead of the sampler do strains have a peak. There are no peaks for the extension phases near the cutting edge of the sampler and the soil elements do not undergo a second compressive strain phase inside the sampler tube. It can also be seen that peak axial strains in compression are considerably higher than the maximum axial strain in extension for all the strain paths shown. All these aforementioned findings contrast with those reported by Baligh (1985) for flat-ended samplers. Baligh (1985) analysed the performance of 209 a Simple sampler with a round-end wall and compared it to a flat-ended wall. From the comparison, he states that the analyses "indicate no significant effect of sampler geometry on the strain history at the centreline". Baligh (1985), however, did not show the strain paths of the flat-ended sampler. From Fig. 6.2, it can also be seen that maximum axial strains in extension are equal for all the strain paths. This is because the sampler has no inside clearance. The peak axial strain in compression below the sampler, however, varies across the diameter of the sampler; soil elements in the vicinity of the cutting edge are strained much more than those near the centreline of the sampler. In case of a flat-ended sampler, the strain paths are (BIt), basically controlled by the thickness, or rather the diameter to thickness ratiozyxwvutsrqponmlkjihgfedcbaZY of the sampler. Because of the profound influence of BIt ratio over the actual geometry of the cutting shoe, peak compressive strains are significantly higher than the maximum strains in extension. Baligh et al (1987) predicted the strain history of an element of soil located at the centreline of Simple samplers having aspect ratios BItzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPO = 20, 40 and 50; and inside clearance ratios, ICR = 1.86%, 0.98%, and 0.79%, respectively. The walls of the Simple samplers had rounded ends and a slight reduction in the internal sampler diameter above the tip with a minimum internal diameter located at small distance (O.IB) behind the tip. Fig. 6.3 shows the strain histories obtained by Baligh et al (1987). The maximum peak axial strains reached by soil elements at the centreline are approximately equal to 2.14%, 1.00% and 0.78% for BIt = 20, 40 and 50 respectively. An appropriate comparison of these results with those obtained from the present work can not be made since the samplers studied have cutting shoe designs completely different from the Simple samplers. Nevertheless, an approximate comparison can be made by comparing the strains at the centreline of the samplers which have approximately the same BIt and ICR as those reported by Baligh et al (1987). For a sampler with BIt = 19.16 and ICR = 1.98%, the peak axial strains in compression and extension at the centreline were found to be 0.432% and 1.656% respectively. This sampler has a tapered cutting edge; inside and outside cutting edge taper angles were respectively 0.11~ and 9.9°. For a Simple sampler with a roundend wall, BIt = 20 and IeR = 1.86%, the maximum strains in both compression and extension were 2.14%. Although the BIt ratio and ICR are approximately same for the two samplers, a large discrepancy in the magnitude of the strains results due to differences in cutting shoe designs of the samplers. This shows that cutting shoe 210 designs have a significant effect on the predicted strain paths. Examples of differences in strain histories at various locations within the sampler tube due to differences in cutting shoe geometries for samplers of similarzyxwvutsrqponmlkjihgfedcbaZYXWV Bit ratio will be illustrated in section 6.5 6.3 COMPARISON OF NGI, SGI AND UIOO SAMPLERS Figs. 6.4 to 6.9 show comparisons of the strain paths at six different locations within the sampler tube due to undrained penetration of the samplers. From Figs. 6.4 to 6.9,zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA it can be seen that for all the strain paths, the peak axial strain in compression is highest for the UIDO (type II) sampler and least for the NOI sampler. This is also illustrated in Fig. 3.22(a). Although the thickness of 'the sampler tube of the both UIDO samplers are the same, in the case of the UIDO (type II) sampler peak axial strains in compression are higher than those of the UIDO (type I ) sampler. This can be attributed mainly to the effect of the outside cutting edge taper angles. The outside cutting edge taper angle of the UIDO (type II) sampler is higher than that of the UIDO (type I ) sampler. Also the thickness of sampler tube at the cutting edge is higher for the UIDO (type II) sampler than for the UIDO (type I) sampler. Figs. 6.4 to 6.7 show that for the strain paths at distances of up to O.7~ from the centreline of the samplers, the peak axial strain in extension is highest for UIDO (type II) sampler and least for NOI sampler. At 0.8~ from the centreline, however, peak axial strain in extension for the NOI and SOl samplers are approximately equal [see Fig. 3.22(b)], and at 0.9~ from the centreline and at the inside edge of the sampler tube (see Figs. 6.8 and 6.9), peak strains in extension are higher for the NOI sampler (ICR = 0.93%) than for the SOl sampler (lCR = 0.4%). This is because in the vicinity of the cutting shoe the design details, especially inside clearance ratio, control the magnitudes of peak strains in extension. The higher the inside clearance ratio, the higher is the magnitude of peak axial strain in extension. It can also be seen from Fig. 6.9 and 3.22(b) that at and near the cutting edge, soil elements for the UIDO (type I) sampler (IeR = 1.44%) suffer higher peak strains in extension than those adjacent to a UIDO(type II) sampler (lCR = 1.1%). At the inside edge of the sampler tube, the peak axial strain in extension is highest for the Ul00 (type I) sampler and least for the SGI sampler; while at the centreline, the peak axial strain in extension is highest for the Ul00 (type Il) sampler and least for the NO! sampler. 211 For all the samplers the minimum peak axial strains (at the centreline of the sampler) in compression and extension were determined by extrapolating the curves shown in Fig. 3.22. The minimum and maximum peak axial strains (at the inside edges of the samplers) in compression and extension for all the samplers have been listed below for comparison. Peak axial strains (%) Sampler Extension Compression type minimum maximum minimum maximum NOI 0.309 0.554 0.260 1.912 SOl 0.573 0.897 0.510 0.869 Ul00 (Type I) 0.651 1.224 0.781 2.929 U100 (Type II) 1.114 2.307 1.427 2.340 6.4 COMPARISON OF CUTTING SHOE DESIGNS FROM THE PARAMETRIC STUDY A parametric study of samplers of different area ratios, inside clearance ratios, and inside and outside cutting edge taper angles has been presented in chapter 3. The strain paths of soil elements due to penetration of the samplers are shown and their nature discussed. In this section, comparisons of the strains paths of the samplers having varying area ratios, inside clearance ratios, and cutting edge taper angles are presented. It should be noted that the sampler with an area ratio 29.64%, inside clearance ratio 0.99% and, inside and outside cutting edge taper angles of 0.7160 and 9.90 respectively has been included for comparison in all the following sections. 212 6.4.1 SAMPLERS WITH DIFFERENT AREA RATIOS Comparisons of strain paths at four locations within the sampler tube are shown in Figs. 6.10-6.13. For all the samplers inside clearance ratio, inside cutting edge taper angle and outside cutting edge taper angle were kept fixed and their values were respectively 0.99%, 0.716° and 9.9°. It can be seen that the peak axial strains in compression below the sampler and peak axial strains in extension inside the sampler tube depend on the area ratio of the samplers. In Figs. 6.14(a) and 6.14(b), the peak axial strains in compression and extension have been plotted respectively as a function of area ratio. It can be seen from Fig. 6.14(a) that peak axial compressive strain increases with increasing area ratio of the samplers. Fig. 6.14(b), however, shows that an increase in area ratio has a minor effect on the peak axial strain in extension. Peak axial strain in extension increase only slightly with increasing area ratio. BIt ratio (due to changes in area ratio) on the The effect of changes inzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA predicted axial strains was also evaluated. Figs. 6.15(a) and 6.15(b) show the variation of the peak axial strain in compression and extension respectively due to changes in BIt ratio of the samplers. It can be seen from Fig. 6.15(a) that peak axial compressive strain reduces with increasing BIt ratio of the samplers. No significant effect of BIt ratio on the peak axial strain in extension has been observed as can be seen from Fig. 6.15(b). It can, therefore, be concluded that an increase in area ratio (or a decrease in BIt ratio) by increasing the thickness of the sampler has a marked influence on the initial compression phase ahead of the sampler while changes in area ratio and BIt ratio have little effect on the extension phase in the vicinity of the cutting shoe. The minimum peak axial strains (at centreline) and maximum peak axial strains (at inside edge) in compression and extension are shown in the following table: 213 Peak axial strains (%) Area BIt Compression ratiozyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA ratio (%) Extensionzyxwvutsrqponmlkjih Minimum Maximum Minimum Maximum 10.14 53.0 0.243 0.447 0.538 1.913 29.64 17.7 0.694 1.177 0.598 2.131 50.73 11.3 0.933 1.504 0.677 2.237 100.46 7.0 1.180 1.835 0.738 2.362 6.4.2 SAMPLERS WITH DIFFERENT INSIDE CLEARANCE RATIOS The area ratio, inside cutting edge taper angle and outside cutting edge taper for these samplers were unchanged and their values were 29.64%, 0.716° and 9.~ respectively. Figs. 6.16 to 6.19 show the comparison of strain paths at various locations within the sampler tube. It can be seen that for all the strain paths peak axial strain in extension in the vicinity of the cutting shoe inside the sampler increases markedly with increasing inside clearance ratio. This is shown systematically in Fig. 6.20, where the peak axial strains in extension have been plotted as a function of inside clearance ratio of the samplers. Fig. 6.21 presents a comparison of the peak compressive strains ahead of the sampler. Fig. 6.21 illustrates that peak axial strain in compression decreases only slightly with increasing inside clearance ratio. This decrease is due to the decrease in thickness of the samplers or rather due to an slight increase in the BIt ratio of the samplers. It is also evident from Figs. 6.20 and 6.21 that an increase in inside clearance ratio influences the peak extension strains much more significantly than the peak compressive strain ahead of the sampler. The minimum (at the centreline of the sampler) and the maximum (at the inside edge of the sampler) peak axial strain in compression and extension are listed in the following table for comparison. 214 Peak axial strains (%) InsidezyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA BIt Compression clearance Extension ratio ratiozyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA (%) Minimum Maximum Minimum Maximum 0.495 17.0 0.906 1.483 0.333 1.352 0.990 17.7 0.694 1.177 0.598 2.131 1.980 19.2 0.432 0.750 1.656 3.447 3.960 23.0 0.210 .0.370 3.239 5.157 Another marked feature which can be seen from Figs. 6.16 to 6.19 is that in a sampler with a very high inside clearance ratio (3.96%), soil elements suffer considerable strain in compression (about 2.6%) during the second compression phase inside sampler tube. This is, however, not observed for the other samplers. 6.4.3 SAMPLERS WITH DIFFERENT INSIDE CUTTING EDGE TAPER ANGLES Comparisons of strain paths at four locations within the sampler tube are shown in Figs. 6.22-6.25. For all the samplers area ratio, inside clearance ratio and outside cutting edge taper angle were kept fixed and their respective values were 29.64%, 0.99% and 9.9°. Figs. 6.22-6.25 demonstrate that inside cutting edge taper angle has no influence on the peak axial strains in compression ahead of the sampler but it has a profound effect on the peak axial strain in extension. the strain paths shown peak axial strains in compression It can be seen that for all do not vary with inside cutting edge taper angle because both the thickness (or BIt ratio) and outside cutting edge taper angle of these samplers are identical. have been plotted against the respective shown in Fig. 6.26. The peak axial strains in extension inside cutting edge taper angle and are Fig. 6.26 shows that peak axial strain in extension decreases 215 with increasing inside cutting edge taper angle for strain paths up to 0.9Ri from the centreline of sampler. Since the inside clearance ratio of the samplers are the same, the peak axial strain in extension is controlled by the thickness of the sampler tube in the vicinity of the cutting shoe. A higher inside cutting edge taper angle reduces the thickness of the sampler tube near the cutting shoe provided the outside cutting edge taper angle is unchanged. Consequently, peak axial strain in extension decreases the sampler, however, with increasing cutting edge taper angle. At the inside edge ofzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA peak extension strain is governed by the amount of clearance along the length of the cutting shoe. This clearance is measured as the distance from the centreline of the sampler to the inside edge along the length of the cutting shoe. A higher inside cutting edge angle provided higher clearance and consequently at the inside edge, the peak axial strain in extension increases with the increase in inside cutting edge taper angle. This relationship can be seen in Fig. 6.26. The minimum (at the centreline of sampler) and maximum (at the inside edge of sampler) peak axial strains in compression and extension are shown in the following table: Peak axial strains (%) Inside cutting Compression Extension edge taper angle (degree) Maximum Minimum Minimum Maximum 0.358 0.729 1.209 0.937 1.978 0.716 0.694 1.177 0.598 2.131 1.432 0.729 1.182 0.356 2.213 216 6.4.4 SAMPLERS WITH DIFFERENT OUTSIDE CUTTING EDGE TAPER ANGLES For these samplers, the area ratio, inside clearance ratio and inside cutting edge taper angle are unchanged and their values are 29.64%, 0.99% and 0.716° respectively. Comparisons of strain paths at various locations are presented in Figs. 6.27-6.30. Figs. 6.27- 6.30 show that both the peak axial strains in compression and extension depend on the outside edge angle of the cutting shoe. Fig. 6.31(a) shows the variation of peak axial strain in compression with the increase in outside cutting edge taper angle. It can be seen that peak axial strain in compression increases considerably with increasing outside cutting edge taper angle. Although the thickness andzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA BIt ratio of these samplers are the same, the outside cutting edge taper angle increases the thickness of the sampler tube near the cutting edge. As a result, the peak axial strain in compression increases with increasing outside cutting edge taper angle. Fig. 6.31(b) shows curves of peak axial strain in extension as a function of the respective outside cutting edge taper angles. It can be seen from Fig. 6.31(b) that peak axial strain in extension increases only slightly with the increase of outside cutting edge taper angle. Outside cutting edge taper angle, therefore, has little effect on the peak extension strains in the vicinity of the cutting shoe inside the sampler tube. The minimum (at the centreline) and maximum (at the inside edge) peak axial strains in compression and extension for the samplers are quoted below. Peak axial strains (%) Outside cutting Extension Compression edge taper angle (degree) Maximum Minimum Minimum Maximumzyxwvutsrqponml 5.000 0.330 0.513 0.581 1.937 9.900 0.694 1.177 0.598 2.131 19.290 1.185 2.112 0.686 2.352 217 6.4.5 SUMMARY In the previous sections (sections 6.4.1 to 6.4.4), the calculated strain paths and peak axial strains in compression and extension due to undrained penetration of samplers of different area ratios, inside clearance ratios, and inside and outside cutting edge taper angles have been presented and compared. It is apparent that the degree of disturbance, which is likely to depend upon peak axial strains in compression and extension, is a function of all the design features of a sampler, e.g., area ratio, inside clearance ratio, inside cutting edge taper angle and outside cutting edge taper angle. A well designed sampler should have an appropriate combination of all these design parameters in order to reduce the amount of disturbance during sampling. In order that the maximum peak axial strains in both compression and extension do not exceed more than 1%, the following values of the design parameters should be adopted for a sampler. (a) The sampler should have a low area ratio, preferably not more than 10%. The sampler should, however, be strong enough to penetrate without buckling. (b) The sampler should have a low inside clearance ratio (not more than 0.5%). La Rochelle et al (1981) have Larger values allow excessive straining of the soil.zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPO reported that the quality of a sample obtained using the Laval sampler (which has no inside clearance) is similar to that of a block sample. Cc)The sampler should have a moderate inside cutting edge taper angle of 1 to 1.5°. Providing inside clearance in the vicinity of the cutting shoe by increasing the inside cutting edge taper angle has been found to reduce the amount of disturbance. (d) The sampler should have a small outside cutting edge taper angle, preferably not more than 5°. Larger outside cutting edge taper angles have been found to cause severe disturbance to soil. 6.5 COMPARISON OF FLAT-ENDED SAMPLERS AND SAMPLERS OF IDENTICALzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA BIt RATIO Figs. 6.32-6.35 show the comparison of strain paths at four locations within the sampler tube. It can be seen from Figs. 6.32-6.35 that both the peak axial strain in compression ahead of the sampler and the maximum axial strain in extension inside the sampler are dependent on thezyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA BIt ratio of the sampler. It can also be seen that the strains for the sampler of thickness 5.9 mm are less than those for the sampler 218 of thickness 4.9 mm. This, therefore, indicates that the strains in case of flat-ended samplers are not controlled by the thickness of the samplers. The variation of peak axial strain in compression with thezyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA BIt ratio of the samplers is presented in Fig. 6.36. Fig. 6.36 illustrates that the peak axial strain in compression decreases with increasing BIt ratio of the samplers. The dependency of maximum axial strains in extension (which are approximately the same at all locations) on BIt ratio of the samplers is also shown in Fig. 6.36. It can be seen that maximum axial strain in extension also decreases with the increase in BIt ratio of the samplers. Comparing the curves in Fig. 6.36 with those of Fig. 6.15(a), it is evident that at any BIt ratio, the peak axial strains in compression (at all locations within the sampler) for the flatended samplerszyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA are considerably higher than those of samplers of different area ratios with a fixed inside clearance ratio and inside and outside cutting edge taper angles. This certainly demonstrates the effect of cutting shoe geometry of a sampler on soil disturbance. The minimum and maximum peak axial strains in compression and the maximum axial strains in extension for all the samplers are listed in the following table for comparison. Sampler BIt No. ratio t Peak axial Maximum axial strains in strain in compressionzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONM extension (%) Minimum (%) Maximum I 45.6 1.25 1.125 3.381 0.493 II 23.0 2.50 2.345 6.416 1.004 III 12.2 4.90 4.750 10.790 1.958 IV 19.9 5.90 2.583 6.780 1.284 Baligh (1985), from his analyses of a Simple sampler and a flat-ended sampler, concluded that the sample straining (disturbance) depends only on the BIt ratio of the sampler and that sampler geometry has no significant effect on the strain history on the centreline. The two geometries also were reported to provide equivalent shear 219 distortion at the centreline [see Figs. 2.29(a) and 2.29(b»). The effect of the cutting shoe geometry on soil distortions was found to be visible only in the vicinity of the sampler walls. From the present study, however,zyxwvutsrqponmlkjihgfedcbaZYXWVUTS it has been found that the precise geometry of the cutting shoe has a profound influence on disturbance, as described by the level of axial strain, not only in the vicinity of the sampler edge but also near the centreline of the sampler. The strain histories of the NOI 54 mm dia. sampler and those of a flat-ended sampler of identicalzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA BIt ratio and thickness have been compared in Fig. 6.37. Figs. 6.37(a) and 6.37(b) respectively show the strain histories near the centreline (0.1~ from the centreline) and near the inside edge (O.lRj from the inside edge) of the samplers. Despite the same BIt ratio and thickness of the samplers, it can be seen from Figs. 6.37(a) and 6.37(b) that at both these locations the strain history associated with flat-ended sampler is completely different from that associated with NOI sampler. The flat-ended sampler causes severe straining in compression ahead of the sampler. Similar results have been found when an SOl 50 mm sampler has been compared with a sampler of identical BIt ratio and thickness as that of sal sampler (Fig. 6.38) and also when UIOO samplers and a sampler of same BIt ratio and thickness as those of UIOO samplers have been compared among themselves (Fig. 6.39). Figs. 6.37-6.39, therefore, imply that BIt ratio or thickness is not the dominant factor that controls the straining (disturbance). It is rather the cutting shoe design which plays a significant role in determining the disturbance during sampling. The effect of cutting shoe design on sample straining is visible both near the centreline and near the inside edge of the sampler. The BIt ratio controls the level of strains at the centreline or at other locations only when the samplers have identical cutting shoe designs. as has already been seen from the comparison of flat-ended samplers of different BIt values (see Figs. 6.32-6.36). 6.6 ONE-DIMENSIONAL CONSOLIDATION AND PERMEABILITY PROPERTIES OF NORMALLY CONSOLIDATED LONDON CLAY The consolidation and permeability characteristics of reconstituted normally consolidated soft London Clay were investigated by carrying out incremental loading oedometer tests. A soft London Clay sample. as mentioned in. chapter 4, was prepared in the laboratory by Ks-consolidation from slurry to a maximum vertical effective stress of 100 kPa. The compressibility and expansibility characteristics of 220 reconstituted soft London Clay undergoing incremental loading in an oedometer are presented in Figs. 6.40 and 6.41. loading and unloading consolidation stages In Fig. 6.40, void ratios (e) at the end of each have been plotted against log. (vertical effective pressure) while Fig. 6.41 shows the plotting of coefficient of volume compressibility, m, and coefficient vertical effective consolidation of volume increase,zyxwvutsrqponmlkjihgfedcbaZY Ill, as a function of log. pressure. Figs. 6.40 and 6.41 show the features similar to that expected for a normally consolidated clay. The values of compression index, Cc and swelling index, Cl' estimated from the loading and unloading curves in Fig. 6.40, are respectively 0.55 and 0.15. For loading up to 200 kPa, m, increases from 0.47 m2/MN to 0.85 m2/MN; m, then decreases up to 0.21 m2/MN at a stress range of 400 to 800 kPa. During unloading from 800 kPa, m, increases from 0.056 m2/MN to 0.514 m2/MN. Hopper (1988) also carried out standard incremental loading oedometer tests on reconstituted overconsolidated London Clay [LL = 70, PI = 49,( Q"y)mu = 100 kPa, OCR = 10] for three different stages of loading. The average Cc values for the stress ranges 100 to 200 kPa and 200 to 1000 kPa were For loading up to 100 kPa, m, increased from 0.33 0.57 and 0.42 respectively. m2/MN to 0.8 m2/MN; m, then decreased up to 0.06 m2/MN at a stress range of 800 At a stress range of 400 to 800 kPa, m, was 0.17 m2/MN. to 1600 kPa. Dial gauge readings for each loading stage were plotted as a function of both logarithm of elapsed time and square root of elapsed time. consolidation, Cy The coefficients have been calculated from these curves using Casagrande's fitting method and Taylor's method (B.S. 1377: 1975). called log-time (log t) and square-root-time of curve These methods are also ('{i) method respectively. Fig. 6.42 shows coefficient of consolidation, c, as a function of vertical effective stress. It can be seen that c, reduces quite rapidly on passing through the preconsolidation (lOO kPa). Beyond the preconsolidation range, change in c, is insignificant. reconstituted pressure pressure, i.e., in the normally consolidated Similar observations have been reported for low plasticity (PI=17) Magnus Clay (Jardine, 1985; Hight et al, 1987). From Fig. 6.42, it is also evident that cv-values obtained from the log t are less than those obtained from the \Sf method for all the stress ranges. considerable at preconsolidation stresses less than the preconsolidation pressure, however, the difference is small. The difference stress. Beyond is the The average value of c, in the normally consolidated range (100 kPa to 800 kPa) obtained from log t method is 0.27 ma/yr, while it is 0.3 ma/yr when calculated using 'if method. 221 Hopper (1988), however, reported an average value of 0.24 m2jyr (at a stress range of 50 to 2000 kPa), calculated usingzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA \If method, for London Clay. ) were also computed for different stress Coefficients of permeability,k (= c.m..'Yw levels of loading. In Figs. 6.43 and 6.44, the vertical permeability of the clay is shown in relation to changes in vertical effective consolidation stress and to void ratio at the end of each loading stage respectively. It can be seen from Fig. 6.43 mls. In the plots of void that permeabilities vary between 2 x 10-11 and 3 x 10-10zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJ ratio against log. (permeability), shown in Fig. 6.44, the relationships are approximately linear over the complete stress ranges. The average slope of the relationships, which is termed the permeability change index, ~ is 0.54 for this clay. Hight et al (1987) reported a value of 0.66 for reconstituted Magnus Clay. The approximately linear relationship between void ratio and log. (permeability) has been found to apply to other reconstituted clays (Lambe and Whitman, 1969; Hight et al, 1987) and has been shown to extend to high void ratios in very soft sedimented clays (Been and Sills, 1981). The permeability change index, ~ has been shown by Tavenas et al (1983) to be numerically equal to one-half of the initial void ratio, Co, which for this clay was 1.33. The ratio of compression index to permeability change index, i.e., CJC" for the soft London Clay studied is 1.02. Berry and Wilkinson (1969) reported that for many soils CjCrr. often lies within the limits of 0.5 and 2.0 while Mesri and Rokshar (1974) observed that the experimental values of Cj~ were found to vary between 0.5 and 5.0, more or less as a function of initial void ratio. 6.7 STRESS AND STRAIN PATH TEST RESULTS 6.7.1 Ko-CONSOLIDATION In these tests, as mentioned earlier, the specimens were consolidated anisotropically under Ko-conditions, from their estimated "in-situ" stresses. Specimens were consolidated up to 1.8 to 1.9 times the maximum past effective vertical stress (100 kPa) in order to eliminate the effects of sampling (Ladd and Foott, 1974; Gens, 1982; Hight et al, 1985). The estimated Ko-value is approximately 0.64. Although the true Ko-condition (i.e., ~ = e, = 0) was not met precisely, the lateral strains during 222 Ko-zyxwvutsrqponm consolidation were significantly smaller than the axial strain, as can be seen from Figs. 6.45 and 6.46. Fig. 6.45 shows that the maximum radial strain during Koconsolidation is approximately 0.3%. The average ratio of the radial strain to axial strain at the end of consolidation was less than 0.04. Typical plots of local volumetric strain versus local axial strain are shown in Fig. 6.47. It can be seen that e.) is not great. the deviation of the experimental Ko-lines from the true Ko-line (e, =zyxwvutsrqponmlkjihgfedcbaZYX Despite loading the specimens at a rate of only 0.7 kPa/hr, during Ko-consolidation some excess pore pressures were generated at the mid-height of the specimens. Typical curves showing the variation of excess pore pressure at mid-height with effective vertical consolidation stress are presented in Fig. 6.48. It can be seen that the excess pore pressure at mid-height at the end of consolidation is between 8 and 15 kPa, which is less than 10% of the effective vertical consolidation stress (at the end of consolidation). For London Clay (LL = 62, PI = 38), Jardine (1985) found an excess pore pressure (at mid-height of 38 mm dia. x 78 high specimens) of 8 kPa for loading rates between 0.5 kPa/hr and 0.9 kPa/hr. The specimens tested by Jardine (1985) were allowed to drain from top and bottom; no side drains were provided. At the end of consolidation, two specimens were sheared to undrained failure; one in compression and the other in extension. These two tests were carried out to evaluate the reference "undisturbed" behaviour in compression and extension. In the other tests, tube penetration disturbances were imposed on Ko-consolidated specimens. The application of tube penetration disturbances were followed by undrained shearing in compression up to failure. The object of performing these tests was to investigate the effects of tube penetration disturbances on the subsequent stress-strain behaviour of the soil. 6.7.2 INVESTIGATIONS OF VARIOUS APPROACHES TO CORRECT TEST RESULTS The deviator stress is computed by dividing the piston load by an effective area. Because the specimen often deforms substantially during both consolidation and shear, it is necessary to calculate a corrected area based on the initial area, the measured axial and volumetric defonnations, and an observed deformation pattern. 223 Besides,zyxwvutsrqponm restraints are also imposed on the specimen by the rubber membrane enclosing it and by the filter paper drainage strips, and a correction on the measured stresses has to be made. Limited researches have been carried out to assess the possible effects of area correction and the effects of the rubber membrane and filter paper drains on the triaxial shear characteristics of soils (Henkel and Gilbert, 1952; Bishop and Henkel, 1962; Olson and Kiefer, 1963; Germaine and Ladd, 1988; Leroueil et al, 1988; La Rochelle et aI, 1988). Henkel and Gilbert (1952) report that in the undrained triaxial compression test the rubber membrane enclosing the specimen gives rise to an apparent increase in strength which is proportional to the stiffness of the membrane. Bishop and Henkel (1962) also reported similar observations. Olson and Kiefer (1963) investigated the effect of lateral filter paper drains on undrained triaxial shear properties of sodium kaolinite (LL = 50, PI = 19). The test results demonstrated that the drains may either increase or decrease the shearing parameters, depending on the relative strengths of the filter paper and the soil. Recent investigations by Leroueil et al (1988), however, showed no effect of the filter drains on shear strength. Mitachi et al (1988) investigated the influence of filter strip shape on consolidated undrained triaxial extension test results. Based on their test results, a filter strip with a spiral slit was recommended for use during in consolidated undrained extension tests, because of fast consolidation and low tensile strength. Germaine and Ladd (1988) reported the influence of area correction on computed shear stress for constant volume triaxial compression tests. They showed that the reduction in shear stress (also equal to per cent increase in area) due to the area correction depended on the deformation mode during shearing, namely, cylindrical, parabolic or bulging. The cylindrical assumption ( the specimen is assumed to deform as a right cylinder) overpredicts the strength while bulging correction (the strains are assumed to be concentrated on the central portion of the specimen) underpredicts the strength. The most appropriate choice of correction should be based on the observed geometry of the specimen at the end of the test. A comprehensive discussion of membrane and area correction may be found in La Rochelle et al (1988). They reported that both membrane and area correction are influenced by the mode of failure (bulging, shearing on a single plane, or splitting) and the strain at failure. La Rochelle et al (1988) introduced an appropriate procedure for correcting principal stresses for both types of failure.zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA 224 In all the tests carried out in the present research, an attempt has been made to assess the effects of area and membrane corrections on the stress-strain parameters during undrained shearing for both "undisturbed" and "disturbed" specimens. effect of lateral filter paper drains was not taken into account. The The following three approaches were considered. (i) Firstly, the deviator stresses were corrected for changes in cross-sectional the soil specimens only. During undrained shearing, it was assumed area of that the specimens deform as a right cylinder with zero volumetric strain. The corrected area, Ac during shear was calculated using the following equation:zyxwvutsrqponmlkjihgfedcbaZYXW ....(6.1) ~ = N(1-e) where, Ao = cross-sectional e = axial strain during shear area of the specimen after consolidation The undrained shear characteristics obtained on the basis of area correction alone are listed in Table 6.1 (ii) Secondly, the deviator stresses were corrected for the changes in cross-sectional area of the soil specimen and also for the contribution of the rubber membrane. Corrected areas were calculated using Equation 6.1. The correction due to membrane constraint was based on the assumption that the rubber membrane and the test specimen deform as a unit. No buckling of the membrane is likely to occur and the membrane acts as reinforcing compression shell around the specimen. The specimen was assumed to deform as a right cylinder and the rubber membrane correction at any strain, e during shear was computed using the following equation: Om =zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA [1tDoMe(1-e)]/Ao ....(6.2) where, om = decrease in deviator stress to be applied to allow for membrane stiffness Do = diameter of specimen after consolidation M = compression modulus of membrane Ao = cross-sectional area of the specimen after consolidation When a membrane is placed around a specimen, it applies a lateral confming pressure on the specimen. This lateral pressure is a function of the initial tangent modulus (1% modulus) of the membrane and initial diameters of the membrane and of the specimen. For a latex membrane with an initial modulus of 0.6 N/mm, the initial confining pressure, which has to be added to the cell pressure, was 0.61 kPa on a 225 102 mm diameter specimen. Since all the tests were carried out with a moderately high cell pressure (382 kPa), the increase in initial lateral pressure is 0.16% which is very small indeed and truly negligible. Also no buckling of membranes at high cell pressure was observed and therefore the hoop tension in the membrane which results in an increase in minor principal stress was not considered. (iii) Finally, for comparative purposes no correction for changes in cross-sectional area and due to membrane restraint were applied for the deviator stresses. Table 6.2 presents a comparison of the undrained stress-strain parameters for the three approaches considered. The initial tangent modulus,zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONML El and undrained shear modulus, G, are not shown in Table 6.2 because these refer to very small strains and were the same for all cases considered. It was found that for the tests where the specimens were sheared in compression (i.e., test 1 and tests 3 to 8), the undrained shear strength, Cu was approximately 4.5% to 7% higher when deviator stresses were not corrected at all, than when deviator stresses were corrected both for area and membrane effects. sheared in extension), however, c.. In test 2 (specimen corrected for changes in area and for the contribution of membrane is about 10% higher than that without any correction applied for deviator stresses. The effective angle of internal friction, <1»' for compression failure was typically 3% to 5% higher when no corrections were applied than those corrected for changes in area and membrane effects. For the test in which the specimen failed in extension, however, <1»' corrected for changes in area and membrane constraint was as much as 22% higher than that computed without any correction. Axial strains at peak strength, e, were also much higher when deviator stresses were not corrected for the compression failure tests. In the extension test, e, was the same with and without the corrections applied for deviator stresses. The secant modulus at half the maximum deviator stress, Eso increased due to area correction in compression tests. For example, in tests 1 and 8, Eso increased by about 12% and 7% respectively. In the other tests, e.g., tests 3 to 7, the increase in Eso was insignificant. approximately In the extension test (test 2), however, Eso decreased 10% due to corrections for changes in cross-sectional area. Skempton's pore pressure parameter A at peak strength, A, was also modified when deviator stresses were corrected for both area and rubber membrane effects as can 226 be seen from Table 6.2. No significant difference was observed when the stress-strain parameters. listed in Table 6.2. computed applying only the area correction are compared with those corrected using both area and membrane correction. This is because the decrease in deviator stress to be applied to allow for membrane stiffness is insignificant at small failure strains. Deviator stresses were also calculated using the actual cross-sectional areas at approximately the mid-height of the specimens. In computing the actual cross-sectional areas. the diameter of the specimens were determined from the lateral caliper readings. Fig. 6.49 shows typical comparisons of deviator stresses corrected using different approaches for two tests. In Fig. 6.49. uncorrected deviator stresses are also plotted as a function of overall strain. From Fig. 6.49, it can be seen that the deviator stresses corrected for changes in cross-sectional area On the assumption of plastic failure (i.e., specimen deforms as a right cylinder) agree well with those corrected using actual cross-sectional area of the specimen during shearing. From Fig. 6.49, comparing the deviator stresses corrected for changes in area only with those corrected for both area and membrane. it is also evident that the effect of the rubber membrane on deviator stress is insignificant. Based on these assessments, it was concluded that deviator stresses could be corrected for changes in cross-sectional area only. without introducing any significant error in the computed stress-strain parameters. Consequently, all the test results presented in the following sections have been corrected in this way. Restraints imposed by the rubber membrane and lateral filter paper drains have not been considered. 6.7.3 OBSERVED BEHAVIOUR IN COMPRESSION AND EXTENSION FOR "UNDISTURBED" SPECIMENS 6.7.3.1 STRESS PATHS In Fig. 6.50, the observed stress paths in compression and extension are plotted in terms of Cambridge effective stress parameters (p', q'). It can be seen from Fig. 6.50 that the mean effective stress, p' decreases during both loading (i.e., q' increases) and unloading (i.e., q' decreases), i.e., shearing in compression and extension respectively. The failure conditions are represented by straight lines passing through the origin. Assuming failure occurs On the critical state line (CSL), then the equations of the 227 critical state lines shown in Fig. 6.50 can be given by the following equations: = M.,p' q'zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA ....(6.3) and q' = -MeP' ....(6.4) where, M., and M, are the critical state parameters in compression and extension respectively. The values of M.,and M, are 0.757 and 0.94, indicating that the critical state lines are not symmetrical about the p' axis. The effective angle of internal friction (or critical state friction angle), which can be conveniently calculated from the M values, are 19.6° and 33.9° in compression and extension respectively. Since 4>'. is significantly higher than the compression, 4>'c' it appears that the relative 4>' effective angle of internal friction in extension, effective angle of internal friction in values in compression and extension are influenced by' stress anisotropy in the soil. These results contrast with those reported by other research workers (e.g., Jardine, 1985; Gens; 1982; Hight et al, 1987), who found 4>'c approximately equal to Ko-normally consolidated reconstituted samples of clay. however, found 4>'. higher than 4>'c 4>'. for Some research workers, for Ko-normally consolidated samples of kaolin (e.g., Parry and Nadarajah, 1974; Ho, 1985; Atkinson et al, 1987). Using a database derived from 100 different clays, Mayne and HoltzzyxwvutsrqponmlkjihgfedcbaZYXW (198S) found that 4>'. was typically 20% tozyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA SO% greater than <1>' e- A comparison of the values of 4>'e and <1>' e reported by various researchers for different clays consolidated under Ko- condition is presented in Table 6.3. 6.7.3.2 STRESS-STRAIN BEHAVIOUR Values of deviator stress (q) are plotted against axial strain (measured both locally and externally) for the compression and extension tests in Figs. 6.S1(a) and 6.Sl(b) respectively. The following are the main observations: (i) In triaxial compression, the peak undrained strength is mobilised at a small axial strain (e, = I.S%). (ii) The strength mobilised at larger strains in triaxial compression is slightly lower (1%) than that mobilised at peak. The clay, therefore, does not show any significant undrained brittleness of strain-softening behaviour when sheared in compression. (iii) In triaxial extension, the clay also appears to be non-brittle. Both peak and ultimate strength are mobilised at large axial strain.zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA 228 (iv) The stress-strain relationships in both compression and extension are non-linear. For this soft London Clay, the triaxial compression and extension strengths are found approximately equal. The undrained triaxial strength is 43.1 kPa. The initial tangent modulus, El (determined from the slopes of the initial part of the stress-strain curves) and the undrained shear modulus, G, (determined from the initial part of stress-local shear strain relationships) were, however, considerably higher during unloading to triaxial extension than during loading in triaxial compression. E; and G, for triaxial compression are 9815 kPa and 3999 kPa respectively while the respective values in triaxial extension are 14024 kPa and 5029 kPa. The secant stiffness at half the maximum deviator stress, Esa, however, was found to be higher in compression than that in extension. Esa in compression and extension are respectively 5998 kPa and 5049 kPa. The values of El' Esa and G, are listed in Table 6.1. Parry and Nadarajah (1974) and Gens (1982) also found higher stiffness in extension than in compression for Ko-normally consolidated specimens of kaolin (PIzyxwvutsrqponmlkjihgfedcbaZYXWVUTSR = 32) and low plasticity Lower Cromer 1111 (PI = 13) respectively. Koutsoftas (1981), however, observed higher stiffness in compression than in extension for a marine clay (PI = 18 ± 5). The small stress-strain behaviour of the specimens in compression and extension can be visualised by plotting stress-strain data on a semi logarithmic scale as suggested by Jardine (1985). Deviator stress versus axial external strain (on a logarithmic scale) plots for the compression and extension tests are shown in Fig. 6.52. For the extension test stress-strain data are plotted only to an axial strain of 10%. The shape of the curves shown in Fig. 6.52 are similar to those shown by Hight et al (1987) for a reconstituted low plasticity clay. The secant stiffness <Eu) determined over different strain levels are plotted against strains (on a logarithmic scale) for both the compression and extension tests and are shown in Figs. 6.53 and 6.54 respectively. It can be seen from these plots that secant stiffness decreases rapidly with increasing strain levels both in triaxial compression and extension. Similar behaviour was also reported by Hight et al (1987) for normally consolidated reconstituted clay and also for slightly and moderately overconsolidated low plasticity (OCR = 1.4 to 4) Lower Cromer Till, Magnus Clay and London Clay. The small stress-strain (EJO.ol.1P'o and L l= characteristics (EJo.l.1<Euk.ol1.] were also evaluated in terms as suggested by Jardine (1985). of indices The values of (EJO.ol..!P'O and L were calculated for the extension and compression tests and these 229 values are 108 and 0.476 respectively extension are 248 and 0.609. in compression. The respective values in It is, therefore, evident that the size of the small strain zone (or rather initial stiffness) is considerably larger in extension than that in ThezyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFED L values for the compression and extension test, however, indicate compression. that the stress-strain relationships are more non-linear in compression than in Jardine (1985) reported the values of <Euk.Ol'Jp'oand L for Ko-normally extension. consolidated reconstituted Magnus Clay (Pl = 17) and London Clay (PI = 38) sheared in compression and extension. Magnus Clay and London The mean effective stresses prior to shear, p'o, for the Clay were approximately 266 kPa and 291 kPa Values of (EJO.ol./P'O and L for these two clays are listed in the respectively. following table for comparison. L Clay type Shearing mode Magnus Compression 831 0.185 Magnus Extension 1190 0.439 London Compression 482 0.319 London Extension 623 0.497 It can be seen from the above table that for both clays stiffness index, <Eu)O.olJp'Ois considerably higher in extension than that in compression. Also for both the clays the degree of non-linearity is higher in compression than that in extension. Similar type of results have been obtained for the Ko-normally consolidated soft London Clay (PI = 45, p'o.139kPa)from the present investigation. 6.7.3.3 PORE PRESSURE RESPONSE DURING SHEARING Porewater pressure responses were observed throughout the whole shearing stage of each test. In Figs. 6.55(a) and 6.55(b), the changes in pore pressure developed due to change in deviator stress only are plotted against external axial strain for the compression and extension test respectively. 230 It can be seen from Fig. 6.55(a) that, during undrained shearing in compression, the pore pressure increases rapidly with the increase in deviator stress. The pore pressure parameter at peak deviator stress, Ap is 1.25. The pore pressure continues to increase after the peak deviator stress is reached and therefore results in higher values of pore pressure parameter A at failure,zyxwvutsrqpo At. The value of At is 1.96. For the extension test, however, it can be seen from Fig. 6.55(b) that the pore pressure change is very small as compared to that in the compression test. At very small strains (up to 0.5%), the pore pressure decreases, after which it increases. Pore pressure changes become positive after a strain of about 3.5% is reached. At peak strength the pore pressure change is slightly positive and the value of ~ is only -0.003. It is, therefore, concluded that for this normally consolidated soft London Clay, the pore pressure response due to change in deviator stress is much more significant when the specimen is sheared in compression than that when the specimen is sheared in extension. The evidence of yielding during shearing in compression was also investigated. The excess pore pressures which develop during undrained tests are governed by the volumetric strain characteristics of the soil. Thus, a yield condition should be associated with a sharp increase in porewater pressure. Some confirmation of this expected behaviour can be seen in results for the undrained triaxial compression tests on lightly overconsolidated laboratory prepared kaolin (Parry and Nadarajah, 1974). The results have been reported by Parry and Wroth (1981). For the compression test, the pore pressure changes, Au are plotted against change 6.56. A sharp increase in pore pressure can be seen in deviator stresses, Aq in Fig.zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA atzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA Aq = 16 kPa (approximately), indicating a distinct yield point. The point of marked change in pore pressure behaviour, i.e., yield point is located by the intersection of a simple straight line extrapolation of the portions of the curves with different slopes. This is similar to methods used by Mitchell (1970), Parry and Nadarajah (1974), and Parry and Wroth (1981). In a simple undrained triaxial compression test with constant cell pressure, the following relationship should hold for isotropic elastic soil: ....(6.5) Au = 1/3 (Aq) This relationship is plotted in Fig. 6.56. It can be seen that the pore pressure response is much higher than the isotropic elastic value. Parry and Wroth (1981), 231 however, reported that for slightly overconsolidated kaolin the actual response early in the test corresponds approximately to elastic behaviour. The approximate location of the yield point on the stress path for the test performed on soft London Clay is shown in Fig. 6.57. 6.7.4 STRAIN PATH TESTS MODELLING TUBE PENETRATION DISTURBANCES In these tests, as mentioned earlier, different degrees of tube penetration disturbances, as predicted by the Strain Path Method, were imposed on Ko-normally consolidated reconstituted soft London Clay specimens. The application of tube penetration disturbances were followed by undrained shearing in compression up to failure. The different levels of disturbance imposed on the specimens are shown in Table 5.1. In three typical tests (tests 4, 7 and 8), the maximum axial strain during the first compression phase and the maximum axial strain imposed during the extension phase were kept approximately the same.zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGF In these tests the minimum axial strains suffered by the specimens during the second compression phase were all equal to zero. In the other tests, however, maximum axial strain suffered by the specimens during the first compression phase and the extension phase were different in magnitude. 6.7.4.1 STRESS PATHS The normalised effective stress paths in p'-q' space for all the tests carried out are shown in Figs. 6.58 to 6.63. mean effective The stress paths have been normalised by the initial stress at the end of Ko-consolidation, i.e., p'; The normalised effective stress paths of the "undisturbed" specimens (i.e., tests 1 and 2) are also shown in Figs. 6.58-6.63 as dashed lines. The stress paths in Figs. 6.58-6.63 show the following features: (i) During the first compression phase, mean effective stress decreases with the increase in deviator stress. (ii) During the extension phase, deviator stress decreases fairly rapidly while the mean effective stress decreases only slightly. The effective stress paths did not touch the yield line in extension, although in tests 5 and 6, the effective stress paths almost hit the yield line in extension. 232 (iii) During the second compression phase, the mean effective stress remains almost constant for the less "disturbed" specimens (tests 7 and 8) as seen from the approximately vertical effective stress paths. For the more "disturbed" specimens (tests 3, 4, 5 and 6), however, the mean effective stress decreases with the increase in deviator stress. (iv) The relative position of the stress point on the stress path at the end of the application of tube penetration disturbance (i.e., prior to shear) depends on the degree of disturbance applied. For symmetric disturbance, i.e., equal maximum axial strain in compression and extension during the first compression and extension phases respectively (tests 4, 7 and 8), this point lies on or slightly above the Ko-line. Baligh et al (1987) also found similar results from a test on reconstituted Ko-normally consolidated Boston Blue Clay. (v) The shape of the stress paths during undrained shearing (i.e., after disturbance) are typically those of lightly overconsolidated soils. With the exception of test 3, the stress paths are substantially vertical until they meet the failure or critical state line. In test 3, mean effective stress increases with the increase in deviator stress, indicating more overconsolidated behaviour. The normalised effective stress path of the "undisturbed" specimen which was sheared in compression up to failure (i.e., test 1) is shown in Fig. 6.64. Comparing the part of the stress path (Figs. 6.58-6.63) during undrained shearing for the "disturbed" be specimens with that for the "undisturbed" specimen shown in Fig. 6.64, it canzyxwvutsrqponmlkjih seen that tube penetration disturbances produced appreciably different effective stress paths to failure. In the "undisturbed" specimen, during undrained shearing in compression, mean effective stress decreased with the increase in deviator stress. In the case of "disturbed" specimens, however, the effective stress paths during undrained shearing in compression are approximately vertical, indicating constant mean effective stress under increasing deviator stress. Because of tube penetration disturbance, the specimens therefore produced effective stress paths similar to those of lightly overconsolidated clays. Assuming failure occurred on the critical state line, the critical state parameter, ~ was calculated for all the tests using Equation 6.3. The effective friction angles were then computed from the values of M; The values of effective friction angles, cj»' arc shown in Table 6.1. It can be seen from Table 6.1 that, compared with the 233 "undisturbed" specimen (test 1), the effective angle of internal friction increased only (1 to 6%) due to the application of tube penetration disturbances. slightlyzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA The most significant effect due to the application of tube penetration disturbances, however, is the reduction in mean effective stress. This effect can be observed from effective stress paths shown in Figs. 6.58-6.63 by comparing the mean effective stresses before and after the application of tube penetration disturbances, In Table 6.1, the mean effective stresses of the specimens before and after the application of disturbances have been listed. The per cent reduction in mean effective stresses for all the specimens are shown in Table 6.4. It can be seen from Table 6.4 that the reduction in mean effective stress depends on the total strain applied during different phases. Total axial strain in this context means the algebraic summation of the axial strains suffered by a specimen during the first compression phase, extension phase and second compression phase. It has been found that the relationship between the total strain imposed and the reduction in mean effective stress is approximately (Fig. 6.65). PI = linear For Ko-normally consolidated reconstituted Boston Blue Clay (LL = 42, 20), Baligh et al (1987) found an appreciable reduction (about 59%) in mean effective stress due to the application of tube penetration disturbances strain = 4%). (total axial The applied tube penetration disturbances was equal to that predicted by the Strain Path Method for soil elements along the centreline of an S-sampler with BIt of 40 and inside clearance ratio of 1%. For approximately an aspect ratio,zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA identical magnitude of tube penetration disturbances when imposed on Ko-normally consolidated reconstituted soft London Clay (LL = 69, PI = 45), it has been found that the decrease in mean effective stress is only 26% (test 4 in Table 6.4). These results evidently imply that the severity of reduction in mean effective stresses due to the application of tube penetration disturbances is much more acute in less plastic clays than in more plastic clays. disturbances are approximately centreline of sal In tests 7 and 8 where the applied tube penetration same as those predicted for soil elements along the 50 mm dia. piston sampler (A.R. = 44%. ICR = 0.4%, BIt = 12.2) and NOI 54 mm dia. piston sampler (A.R. respectively, the corresponding = 11.4%, ICR = 0.93%, BIt = 45.6) reductions in mean effective stresses are 17% and 10%. Baligh et al (1987), however, did not investigate the effect of the application of varying degrees of tube penetration disturbances for the Boston Blue Clay. For ideal sampling disturbance on Ko-normally consolidated Boston Blue Clay, Baligh 234 et al (1987) reported a decrease of 57.5% in mean effective stress. Reduction of mean effective stresses due to "perfect" sampling of Ko-normally consolidated clays was also reported by several investigators. For Weald Clay (LL = 46, PI = 24), Skempton and Sowa (1963) found 20 to 22% decrease in mean effective stress. Ladd and Lambe (1963) reported a decrease in mean effective stresses of 5 to 22% for Kawasaki Clays (LL = 48-106%, PI = 16-46%) while Ladd and Varallyay (1965) found a reduction of only 8% for remoulded Boston Blue Clay"(LL = 33, PI = 15). (LL = 32, PI = 17) from the North Sea, Hight For low plasticity reconstituted clayzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIH et al (1985) reported a reduction of as much as 28% in mean effective stress due to "perfect" sampling. 6.7.4.2 STRESS-STRAIN BEHAVIOUR DURING THE APPLICATION OF STRAIN PATHS Figs. 6.66 to 6.71 show the normalised deviator stress versus strain during the application of undrained strain paths simulating tube penetration disturbances. Both external strain and local axial strain imposed during the strain paths are shown on all figures. The following key features can be seen in Figs. 6.66 to 6.71. (i) External axial strains are typically less than the local axial strains during the first compression phase. Similar behaviour was also found during shearing the "undisturbed" specimen in compression [see Fig. 6.51(a)]. (ii) With the exception of test 8, the external strains are higher than strains measured locally at the end of the extension phase. (iii) The unequal magnitudes of local axial strain and local shear strain indicate that despite loading and unloading under undrained conditions, small but unrecoverable volumetric strains occur in the central section of the specimen. In tests where the external axial strain did not exceed more than 0.5% during the first compression phase, local axial strains were higher than shear strains, indicating increase in volume or dilation. In tests 3 and 4 where the axial strains were 1% to 2% during the initial compression phase, however, local shear strains were greater than local axial strains. This demonstrates that volume contraction occurred in these tests. During the reloading or second compression phase, local shear strains were always higher than the local axial strains, indicating dilation during this phase. (iv) The stiffness during loading (initial compression phase), unloading (extension 235 phase), and reloading (second compression phase) are markedly different in each test. The relative magnitude of secant stiffnesses at various strain levels during the application of strain paths are listed in Table 6.5 for all the tests. It can be seen from Table 6.5 that the secant stiffness is a minimum during the initial compression phase and maximum during the initial part of the extension phase. Stiffness reduces during the latter part of the extension phase and then increases again while reloading during the second compression phase.zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHG In test 3, however, significant creep occurred at the end of the initial compression phase which modified appreciably the stress- strain and stiffness characteristics during the subsequent extension and recompression phases. The typical variation of secant stiffness with level of applied strain during different strain paths is shown schematically in Figs. 6.72(a) and 6.72(b) for tests 4 and 7 respectively. Normalised deviator stresses were also plotted as a function of local radial strain during the application of strain paths. Two typical plots (for tests 4 and 7) are presented in Figs. 6.73(a) and 6.73(b) respectively. In Figs. 6.74(a) and 6.74(b), local axial strains have been plotted as a function of local radial strains for tests 4 and 7 respectively. It can be seen from Fig. 6.74 that local axial strain versus local radial strain relations are approximately undrained strain paths. linear during all the different phases of Undrained Poisson's Ratios were computed from the slopes of these straight line relationships. In Table 6.6, undrained Poisson's the tests during different phases of strain paths are shown. Ratio for all It can be seen that undrained Poisson's Ratio are typically higher during the extension phase than those during the initial or second compression phases. Typical normalised pore pressure change curves during the application of strain paths are presented in Figs. 6.75(a) and 6.75(b) respectively for tests 4 and 7. demonstrates (i) During Fig. 6.75 the following: the initial compression phase pore pressure increases rapidly due to increase in deviator stress. (ii) Porewater pressure reduces fairly quickly during the initial stages of extension phase and it changes little during the latter stage of the extension phase. (iii) During the second compression phase the porewater pressure again increases rapidly. 236 Skempton's pore pressure parameter A was calculated for the different phases of undrained strain paths. These values are listed in Table 6.7 for all the tests. It can be seen that Skempton's A-values during the initial compression phases are 1.7 to 3 times greater than those during the second compression phase. Much of these differences can be attributed to the non-linearity of the relationship between pore. pressure change and deviator stress change, but they are also caused, in part, by the non-reversibility of the pore pressure changes caused by the release of deviator stress during the extension phase. It can be seen from Table 6.7 that the A-values during the extension phase.i.e., during unloading, are comparable to Au values for "perfect" = 88, sampling of Ko-normally consolidated undisturbed San Fransisco Bay mud (LLzyxwvutsrqponmlkj PI = 45, Au = = .12 to .24). 0.16 to .24) and remoulded Boston Blue Clay (LL = 33, PI = 15, Au Rather more interestingly, A-values during recompression were similar to those to be expected for an elastic material (i.e., A = 1/3). The soil should behave elastically inside the yield surface. The average A-values during the initial compression phase and the second compression phase are 1.07 and 0.41 respectively. During the extension phases, the A-values are much less; the average value being only 0.17. 6.7.4.3 STRESS-STRAIN AND PORE PRESSURE CHARACTERISTICS AFTER TUBE PENETRATION DISTURBANCES Typical deviator stress versus external axial strain plots for tests 3, 4 and 7 are shown in Fig. 6.76. The important features to note are the following: (i) The peak undrained compressive strength is mobilised at axial strains considerably larger than those for the "undisturbed" specimen. (ii) As with the "undisturbed" specimen, the strength mobilised at larger strains is slightly lower than that mobilised at peak. Therefore, no significant strain-softening occurred in the "disturbed" specimens. (iii) The stress-strain relationships are non-linear. The undrained compressive strength, c, of the specimens are shown in Table 6.1. It can be seen that with the exception of test 7, in all other tests the undrained shear strengths are slightly reduced (1.6 to 6.7%). In test 7, a small increase in undrained shear strength (about 1.4%) was obtained. 237 Baligh et al (1987) reported a 21% reduction in undrained strength ratio (c/a'yJ due to tube penetration disturbances in a Boston Blue Clay. Lacasse and Berre (1988), however, reported that the peak triaxial compression shear strength is about the same for the disturbed and = 2.5) undisturbed specimens of normally consolidated and overconsolidated (OCRzyxwvutsrqponmlkjih plastic Drammen Clay (PI = 27). The disturbed specimen was strained to a magnitude identical to that predicted by Strain Path Method at the centreline of a (BIt = 40, ICR "" 1%). Lacasse and Berre (198"8) reported that the Simple samplerzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA shear resistance at high strains was much higher for normally consolidated disturbed specimen. Skempton and Sowa (1963) found that the undrained strength of the "perfect" samples of low sensitive (S, = 2) Weald Clay were only 1 to 3% less than that of "ground" samples. Other investigators, however, reported moderate reduction in undrained shear strength due to "perfect" sampling. . For example, Noorany and Seed (1965) observed a 6% reduction of the strength for San Francisco Bay mud (SI = 8 to 10); Ladd and Varallyay (1965) found a 10% decrease in strength for normally consolidated Boston Blue Clay; Kirkpatrick and Khan (1984) reported approximately 34% and 47% reductions in undrained strength for unaged "perfect" samples of illite and kaolin respectively. Lacasse and Berre (1988) reported approximately 11% decrease in undrained shear resistance for "perfect" samples of overconsolidated (OCR = 2.5) plastic Drammen Clay (Pl = 27). The axial strains (measured externally) at peak deviator stress, a; for the "disturbed" specimens are listed in Table 6.1. It can be seen that the axial strain at peak strength are increased by about 1.3 times to 3.1 times that for the "undisturbed" specimen (test 1). It is also apparent that the axial strain at peak strength increases with the increase in the level of tube penetration disturbances. Baligh et al (1987) found a quite significant increase in axial strain at peak strength due to tube penetration disturbance for the Boston Blue Clay. Significant increase in failure strain due to "perfect" sampling was also reported by a number of investigators, e.g., Skempton and Sowa (1963), Ladd and Varallyay (1965), and Kirkpatrick and Khan (1984). The initial tangent modulus, El' undrained shear modulus, Gil and secant modulus at half the maximum deviator stress, E50 are shown in Table 6.1 for all the tests carried out. Compared with the "undisturbed" specimen (test I), the "disturbed" specimens suffered considerable reduction in Et, Gil and Esoo 238 The degree of reduction in different moduli depends on the relative values of tube penetration Table 6.1 shows that even for the specimen which is subjected disturbances. to least tube penetration disturbances (approximately identical to that predicted at the centreline of a NOI 54 mm dia. piston sampler), El' G, and Eso are reduced by approximately 49%, 60% and 27% respectively. For an application of tube penetration predicted by the Strain Path Method for soil elements disturbances along the centreline of a (BIt = 40, ICR "" 1%), Baligh et al (1987) found that the undrained Simple samplerzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA modulus ratio (i.e., EsJa'.J However, for approximately decreased as much as 95% for the Boston Blue Clay. identical tube penetration disturbance applied to soft London Clay, the undrained modulus ratio decreased by approximately 65%. Lacasse and Berre (1988) also reponed normally and overconsolidated significant specimens reduction in initial moduli for both of plastic Drammen application of an equivalent tube penetration disturbance. reponed Atkinson Clay due to the Ladd and Varallyay (1965) a slight decrease in E50 due to "perfect" sampling of Boston Blue Clay. and Kubba (1881), however, found considerably lower (50%) secant stiffness, Eso for anisotropically consolidated "perfect" specimens of kaolin clay. The small strain behaviour of the specimens due to the application of tube penetration disturbances was also investigated. Typical variation of deviator stress and secant stiffness with axial strain (plotted on a logarithmic scale) are shown in Figs. 6.77 and 6.78 respectively. Compared with the "undisturbed" stiffness characteristics are markedly different. specimen (see Fig. 6.53), the At all strain levels, the stiffnesses are significantly smaller than those for the "undisturbed" specimen. The stiffness index, (EU)ODI"/P'O and the linearity parameter, L as proposed by Jardine (1985) were also determined for the "disturbed" specimens. These values are listed in Table 6.8. The stiffness index and the linearity parameter of the "undisturbed" specimen (testzyxwvutsrqponmlkjihgfe 1) are also shown in Table 6.8 for comparison. (EJo.oI"/p'o are disturbance. appreciably because of the applied tube penetration The non-linearity in the stress-strain behaviour was reduced considerably only in test 6. observed. reduced It can be seen that the stiffness indices, In the other tests no significant change in non-linearity As a result of "perfect" and block sampling of reconstituted has been samples of North Sea Clay, Hight et al (1985) also observed considerable reduction in the degree of non-linearity in the stress-strain behaviour. The stiffness index, o;.>OD1Jp'O, however, increased significantly (65%) for normally consolidated "perfect" sample and decreased considerably (about 27 to 37%) for overconsolidated 239 (OCR = 2) samples. In Table 6.8, secant stiffnesses at strain levels of 0.01%, 0.05% and 0.1% are also shown. An attempt was made to correlate secant stiffnesses at various strain levels with the initial mean effective stress prior to shear, p'o. The secant stiffnesses corresponding to various strain levels are plotted as a function of mean effective stress prior to shear (which is the same as the mean effective stress after disturbance in case of tests 3 to 8) in Fig. 6.79. Normally one would expect that stiffness would be higher for a specimen when it was sheared from a larger mean effective stress and vice versa. However, this was not found to be the case while examining the stiffness characteristics of the "disturbed" specimens. The stiffnesses were found to be dependent on the initial mean effective stress prior to shear only when the specimens were subjected to "symmetric" tube penetration disturbances (i.e., the maximum axial strain in compression and extension during the initial compression phase and extension phase are equal). This is shown in Fig. 6.79 where the curves are drawn considering only the points from tests 4, 7 and 8 (where applied tube penetration disturbances are approximately "symmetric") and test 1 (i.e., "undisturbed" specimen sheared in compression). It can be seen that secant stiffnesses at various strain levels increase with the increase in mean effective stress prior to shear. It can be seen from Table 6.8 and Table 6.1 that although the mean effective stress prior to shear in test 5 is larger than test 3 and 4, the stiffness parameters, i.e., Gil are all considerably lower secant stiffnesses at various strain levels, El' E50t andzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJ than the corresponding values for tests 3 and 4. This inevitably shows that the initial stiffness characteristics are not controlled only by the preshear effective stress but also other factors, especially the relative level of axial strain imposed on the specimen during the extension phase. In test 5, the specimen suffered considerable axial strain (maximum of -1.44%) during the extension phase of the strain path as compared with only 0.5% axial strain imposed during the initial compression phase of the strain path. It therefore seems that when the maximum axial strain imposed during the extension phase is excessive, the stiffness characteristics are basically controlled by its relative magnitude rather than the preshear mean effective stresses. Further investigation should, however, be carried out to confmn this behaviour. Typical pore pressure change characteristics due to change in deviator stress during shearing are presented in Fig. 6.80 for the tests 4, 7 and 8. In Fig. 6.80, the pore pressure response of the "undisturbed" specimen (test 1) is also shown. It can be 240 seen that pore pressure change reduced appreciably due to the application of tube penetration disturbances. The variation of Skempton's pore pressure parameter A during undrained shearing is presented in Fig. 6.81 for tests 1, 4, 7 and 8. Skempton's A-values at peak strength, i.e., ~ for all the tests are listed in Table 6.1. It can be seen that Ap-values are significantly smaller for the "disturbed" specimens than for the "undisturbed" specimen. The degree of reduction depends on the relative magnitudes of the applied tube penetration disturbances. For tests 4, 7 and 8, for example, ~ decreased by about 78%, 56% and 41% respectively. Significant reduction in pore pressure parameter A due to "perfect" sampling has been reported by several investigators (e.g., Skempton and Sowa, 1963; Seed et al, 1964; Noorany and Seed, 1965; Ladd and Varallyay, 1965). The possible change in the yielding behaviour during undrained loading because of applied disturbances was also studied.zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLK In Fig. 6.82, the change in porewater pressures are plotted against change in deviator stresses for the tests 1,4,7 and 8. It can be seen that for tests 7 and 8 (where the disturbance is small to moderate), there is still evidence of yielding as indicated by the sharp increase in pore pressure. The location of the yield points, however, modified markedly due to applied disturbance. In test 4 (where the applied tube penetration disturbances is higher than those in tests 7 and 8), no sharp change in pore pressure response indicating possible yield condition was observed. Similar observations were also found in tests 3, 5 and 6 where the applied disturbances are also larger compared with tests 7 and 8. It is also evident from Fig. 6.82 that disturbances modified significantly the pore pressure response during the early stages of shearing. It can be seen that the actual response early in the test for the "disturbed" specimens correspond approximately to isotropic elastic behaviour. 6.8 CONCLUDING REMARKS The numerical analyses conducted on NOI, SOl and Ul00 samplers and also samplers analysed in the parametric study indicate that soil disturbance, as described by the level of peak axial strains in compression and extension reduces towards the centre of the sample. Soil elements suffer least disturbance at the sample centreline and maximum disturbance at the inside edge of the sampler. It was also observed that the relative increase in disturbance in the inner half (i.e., from the centreline of 241 sampler to 50% of R, from centreline of sampler) is much less than that in the outer R, from centreline of sampler to the inside edge of sampler). half (i.e., from 50% ofzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA For example, in case of the NOI 54 mm dia. sampler, the peak axial strain in extension at 50% of R, from centreline is about 34% higher than that at the centreline; while the peak axial strain in extension at the inside edge of sampler tube is about 4.47 times greater than that at 50% of R, from centreline. For the UIDO (type I) sampler the increase in the peak axial strain in extension within the inner half and outer half of the sampler are respectively 24% and 201%. These observations suggest that soil located in the outer half of a tube sample should be avoided in the preparation of representative specimens for laboratory testing. Although from the parametric study of cutting shoe designs some limiting values of the design parameters were reported"so that the peak axial strains in compression and extension do not exceed 1%, it was found experimentally that for an equivalent amount of disturbance imposed on unaged Ko-normally consolidated specimen of London Clay the reduction in mean effective stress is 26%. Although the strength parameters, Cu and «1>' remained unaffected, the initial stiffnesses and pore pressure changes were considerably reduced. This, perhaps, indicates that it is extremely difficult to design a tube sampler which could sample undisturbed samples of unaged clays. For sampling in aged clay deposits, however, the reported values of the design parameters may be appropriate. Strain path tests conducted on Ko-normally consolidated unaged London Clay show that tube penetration disturbances have significant effects on the undrained shear behaviour of the clay. The most pronounced effects are the reduction in mean effective stress, initial stiffness parameters and pore pressure changes. The undrained shear strengths, however, were reduced only slightly. As mentioned earlier, numerical analyses showed that soil disturbance is least along the centreline of a sampler. However, for an application of tube penetration disturbances approximately equivalent to those predicted at the centreline of NGI 54 mm dia. sampler (AR = 11.4%, leR = 0.93%,zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA Bit = 45.6) and SOl sampler (AR = 44%, ICR = 0.4%, Bit = 12.2) respectively, the corresponding reduction in mean effective stress were 10% and 17%. and El, E50, Gu, secant stiffnesses at various strain levels, (EU)ODl1Jp'O reduced considerably. A, were all These results indicate that even with these good quality samplers it is virtually impossible to carry out good quality sampling in normally 242 consolidated unaged clays. Experimental evidence has already suggested that the problem is more severe in less plastic sensitive soft clays. For an application of tube penetration disturbances equal to that predicted at the centreline of a Simple samplerzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA (BItzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA = 40, ICR "" 1%), Baligh et al (1987) reported a significant reduction (about 59%) of mean effective stress in Ko-normally consolidated unaged low plastic Boston Blue Clay (PI = 20 ± 2.5). For an equivalent amount of disturbance, the reduction in mean effective stress for a specimen of more plastic London Clay (PI = 45) is 26%. The reduction in strength ratio (cja'vJ and modulus ratio (E~a'vJ for Boston Blue Clay were respectively 21% and 95% compared with 6% and 65% respectively for London Clay. These results also indicate that the effects of tube penetration disturbances depend not only on the design of the sampler but also to a great extent on the plasticity of the soil. Although the effect of reconsolidating the "disturbed" specimens in order to recover the "undisturbed" behaviour was not investigated, the effects of tube penetration disturbances on this unaged soft London Clay evidently suggest the need to reduce sampling disturbance by reconsolidating before undrained shearing. Samples should be reconsolidated using either the Bjerrum (Bjerrum, 1973) or SHANSEP (Ladd and Foott, 1974) procedure in order to eliminate the effects of sampling disturbance. Baligh et al (1987) reported that unaged specimens of Boston Blue Clay, consolidated to 1.5 to 2.0 times the maximum past effective vertical stress after the application of ideal sampling disturbance (i.e., tube penetration and "perfect" sampling disturbances), exhibited virtually the same normalised behaviour of the "undisturbed" behaviour. This finding is consistent with that on which the SHANSEP approach is based. For aged samples, Burland (1990), however, reported a lower undrained strength ratio than for undisturbed specimen. This was attributed to the effect of destructuration when reconsolidating according to SHANSEP procedure. For reconstituted normally consolidated low plasticity clays, Hight et al (1985) found that the effects of "perfect" sampling were only fully removed when reconsolidation was continued to a vertical effective stress greater than 1.75 times the maximum past effective vertical stress. 243 Table 6.1 Undrained shear characteristics of soft London Clay (deviator stresses corrected using area correction only) Mean effective Test stresses Undrained shear behaviour No. E, p'ozyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFED ~' CuzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHG E$O Gu Ap f" a'vc p', (kPa) (kPa) 1 184.4 138.9 19.6 43.1 2 183.4 138.8 33.9 43.1 -15.4 3 192.7 147.6 92.9 20.8 40.2 4 185.6 139.2 103.6 19.8 5 189.5 144.5 109.7 6 188.3 143.1 7 190.4 8 183.0 (kPa) (deg) (kPa) (%) (kPa) 1.5 9815 (kPa) (kPa) 5998 3999 1.25 14024 5049 5029 -.003 4.4 2431 2252 713 0.185 40.6 3.4 2241 2101 780 0.27 20.1 41.6 4.7 1453 1292 456 0.66 116.7 20.4 41.7 3.4 3542 3317 1267 0.47 144.S 119.6 20.2 43.7 3.3 370S 3271 1046 0.55 138.0 123.6 20.0 42.4 1.9 4968 4350 1579 0.74 a'yC = effective vertical stress at the end of Ko-consolidation p', = initial mean effective stress at the end of Ko-consolidation p'o = mean effective stress prior to shear after tube penetration disturbance ~' = effective angle of internal friction c, = undrained shear strength e.. = external axial strain at peak ~ = strength initial tangent modulus E$O= secant modulus at half the maximum deviator stress G, = ~ = undrained shear modulus Skempton's pore pressure parameter A at peak strength 244 Table 6.2 Undrained shear characteristics of soft London Clay (deviator stresses corrected using different approaches) Test No. ESI) CorrectionzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPON C Af fzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA applied y 1 N.C. A.C.O. A. &M.C. 20.2 19.6 19.5 45 43.1 42.9 4.2 1.5 1.5 5360 5998 1.5 1.25 1.27 2 N.C. A.C.O. A. &M.C. 27.7 33.9 32.7 37.4 43.1 41.4 15.4 15.4 15.4 5604 5049 -.003 -.003 -.003 3 N.C. A.C.O. A. &M.C. 21.3 20.8 20.4 42.6 40.2 39.8 5.6 4.4 4.4 2245 2252 .091 .188 .185 4 N.C. A.C.O. A. &M.C. 20.7 19.8 19.6 42.1 40.6 40.2 4.3 3.4 3.4 2100 2101 .26 .27 .28 5 N.C. A.C.O. A. & M.C. 21.0 20.1 20.0 43.6 41.6 41.2 4.7 4.7 3.8 1260 1292 .531 .66 .678 6 N.C. A.C.O. A. &M.C. 21.0 20.4 20.2 43.3 41.7 41.4 3.7 3.4 3.4 3302 3317 .455 .47 .48 7 N.C. A.C.O. A. & M.C. 20.8 20.2 20.1 45.5 43.7 43.4 4.3 3.3 3.3 3157 3271 .583 .55 .56 8 N.C. A.C.O. A. & M.C. 20.6 20.0 20.0 44.1 42.4 42.4 4.5 1.9 1.9 4055 4350 .945 .74 .757 Note: N.C. A.C.O. A. & M.C. = No correction = Area correction only = Area and membrane correction 245 Table 6.3 Effective friction angle in compression and extension for different Ko-normally consolidated clays Soil type Plasticity index Reference Effective friction cjl' (degree) angle,zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA Ko-value Compression Marine silty clay Extension 29.2 31.7 Koutsoftas, 0.5-0.53 18zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA ±5 1981 Spestone kaolin 32 0.64 20.8 28.0 Parry and Nadarajah, 1974 Lower Cromer Till 13 0.50 30.0 30.0 Gens, 1982 Speswhite kaolin 30 0.67 22.8 36.7 Ho, 1985 Speswhite kaolin 30 0.63 24.8 27.5 Ho, 1985 London Clay 38 0.60 22.5 22.5 Jardine, 1985 Magnus Clay 17 0.50 30.0 30.0 Hight et al, 1985 Spestone kaolin 30 0.66 22.0 29.0 Atkinson et al, 1987 London Clay 45 0.64 19.6 33.9 This study 246 Table 6.4 Comparison of reduction in mean effective stress due to tube penetration disturbances Test No. Applied tube penetration disturbances Reduction in mean effective stresszyxwvutsrqponml Ec (%) E. Emili (%) er (%) (%)zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA (%) 3 2.0 -1.0 0 6.0 37 4 1.0 -1.15 0 4.3 26 5 0.5 -1.44 0 3.88 24 6 0.5 -1.15 0.5 2.8 18 7 0.5 -0.6 0 2.2 17 8 0.25 -0.27 0 1.04 10 Note: Test 3 suffered creep strain between Ec = 1.0% and 2.0% 247 Table 6.5 Comparison of secant stiffnesses during the application of tube penetration disturbances Test No. Secant modulus (lcPa) Strain paths 0% to 1% 1% to 0% 0% to -1.15% -1.15% to 0% 1377 8021 1982 6223 7 0% to 0.5% 0.5% to 0% 0% to -0.5% -0.5% to 0% 2845 12712 3610 9513 8 0% to 0.25% 0.25% to 0% 0% to -0.25% -0.25% to 0% 3588 16480 6544 14898zyxwvutsrqponmlkjihgfedcbaZYXWVUT o to 0.5% 0.5% to 0% 0% to -1.3% -1.3% to 0% 2714 11968 2929 6414 4 5 o 6 to 0.5% 0.5% to 0% 0% to -1.13% -1.13 to 0.5% 2962 12385 3530 10586 3 0% to 1% 1% to 2%2% to 0% 0% to -1% -1% to 0% 2379 1159 4732 4539 6820 - Creep occurred during unloading 248 Table 6.6 Undrained Poisson's ratio during different phases of strain paths Undrained Poisson's ratio TestzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA Initial compression Extension 2nd compression No. phase phase phase 3 0.54 0.54 0.42 4 0.61 0.61 0.58 5 0.62 0.62 0.62 6 0.49 0.61 0.51 7 0.41 0.53 0.45 8 0.36 0.45 0.36 249 Table 6.7 Skempton's pore pressure parameter A during different phases of undrained strain paths Test No. 3 4 5 6 7 8 Strain paths Skempton's A-value 0% to 1% 0.88 1% to 2%· -0.36 2% to -1% 0.18 -1% to 0% 0.52 0% to 1% 1.28 1% to -1.15% 0.17 -1.15% to 0% 0.42 0% to 0.5% 1.14 0.5% to -1.44% 0.17 -1.44% to 0% 0.45 0% to 0.5% 0.95 0.5% to -1.15% 0.15 -1.15% to 0.5% 0.39 0% to 0.5% 1.05 0.5% to -0.6% 0.15 -0.6% to 0% 0.34 0% to 0.25% 1.14 0.25% to -0.27% 0.21 -0.27% to 0% 0.36 • Creep occurred during unloading 250 Table 6.8 Comparison of secant stiffnesses and linearity parameter (L) of the "disturbed" specimens Test p'O 1;. (kPa)zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQ L No.zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA (kPa) at 0.01 % 1 138.9 15011 9228 7143 108 0.476 3 92.9 4857+ 2638 2152 52 0.443 4 103.6 4853+ 1970 1819 47 0.375 5 109.7 3007+ 1601 1467 27 0.488 6 116.7 3395+ 3012 2673 29 0.787 7 119.6 6683+ 3337 3002 56 0.449 8 123.6 8225+ 4812 4406 67 0.536 + at 0.05% extrapolated values 251 at 0.1% 2.5 2.0zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA ,...., CD <, N 1.5 EXTENSION "'-J Z CJ ,_, s. P. ................ 5. P• --S.P. -------S. P. S.P. C.S.R. C.L. .5 CJ _J I- z w ~ w _J w 10% 50% 70% 90% OF OF OF OF C. S. R. C. 5. R. C. S. R. C. S. R. FROM C. L.zyxwvutsrqponmlkjih FROM C. L. FROM C. L. FROM C. L. 1.0 I- < u AT AT AT AT • STRAIN PATH a CUTTING SHOE RADIUS a CENTRELINE OF SAMPLER 0.0 -.5 _J < -1. 0 u ,_, COMPRESSION ICl::: w -1. 5 > -2.0 -2.5 -.75 -.50 -.25 -1.00zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA 0.00 .25 AXIAL STRAIN .50 .75 1. 00 GI,) Fig. 6.1 Typical strain paths due to undrained penetration of a piston sampler 2.5 2.0 ,...., m <, N 1.5 EXTENSION "'-J Z CJ ,_, LEGEND I SAME AS ABOVE 1.0 I- -e u .5 CJ _J IZ W :::i: W _J w 0.0 -.5 _J < u -1. 0 ,_, COMPRESSION ICl::: w > -1. 5 -2.0 -2.5 -4.5 -3.0 -1.5 ].5 0.0 AXIAL STRAIN 3.0 4.5 (7.) Fig. 6.2 Typical strain paths due to undrained penetration of a flat-ended sampler 252 --8--r;- 2.0zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA 81 1.5 .. 1.0 c; 0.5 I III ..... s u 0 ..J C II -8.--,zyxwvutsrqponmlkjih 0 E ..!! UJ 0 u -... ~ -0.5 1 -1.0 -1.5 zyxwvutsrqponml 'CR' 8,- 8. )(100 8. I,. - 2.0 -2 o -I Vertical Fig. 6.3 Minimum 010""'" 2. Strain, Ell (0'.) Straining history at centreline of Simple samplers (after Baligh et aI, 1987) 253 2.5 2.0zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA NGI SAMPLER. ................ SGI SAMPLER ,.... 1.5 --- UlDD <TYPE SAMPLER zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA 1) CD <, N ------- .. UlDO (TYPE 11> SAMPLERzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPO EXTENSIONzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA "-J z: 1.0 ...;i .. ~, --<../ CJ ...... f< u .; .5 ,~,~" ------- .~.""'( .... '--------<-- -~ ~- .")) ..-- ..---, CJ ...J ' <, f- 0.0 W zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIH -.5 f-zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA z: :::E w ...J w ......... . .: \ " y' "", i:~'---~. ...J < -1.0 f- u ...... ; , ; I COMPRESSION f- a:: w -1. 5 > -2.0 -2.5 -3 o -1 -2 2 3 AXIAL STRAINzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA Cl.) Fig. 6.4 Strain paths of NOI, SOl and UlOO samplers at 10% of R from centreline 2.5 2.0 ,.... CD <, N t- 1.5 "-J z: CJ ...... f-e u 1.0 z: SAME AS ABOVE ... ~i .'; -: --------:f~( '--.... .5 f- ",-.... -----~--........ '" 0.0 W :::E w ...J w I _-:""j' CJ ...J f- LEGEND EXTENSION -.5 I- 0 -, ~-:.------ ----, J ..= ' .:/ 7,' .> ...J -e -1. 0 w ...... f- a:: w -1.5 > 0 ,," "", r COMPRESS fON -2.0 -2.5 I -3 -2 -1 o AXIAL STRAIN 2 3 Cl.) Fig. 6.5 Strain paths of NOI, SOl and UIOO samplers at 30% of R; from centreline' 254 2.5 2.0zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA --NGI SAMPLER ,-.. CD <, N '-" :z C) ................ SG I SAMPLER 1.5zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA --UIOO (TYPE 1) SAMPLER EXTENSION -------- UlOO (TYPEzyxwvutsrqponmlkjihgfedcbaZ II) SAMPLER 1.0 I- <: u .5 C) __J I- :z w ~ w __J w 0.0 -.5 __J -e -1.0 u .... COMPRESSION l- e::: lJ.J > -1.5 -2.0 -2.5 -3 o -1 -2 AXIAL Fig. 6.6 2 3 STRAINzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONM (X) Strain paths of NOI, SOl and UIOO samplers at 50% of R from centreline 2.5 2.0 ,-.. CD <, N 1.S EXTENSION '-' :z LEGEND : SAME AS ABOVE 1.0 .... C) I- <: u .5 Cl __J I- 0.0 w __J w -.5 z w ~ -...!!'~-----...~zyxwvutsrqponmlkjihgfedcbaZYXWV -----" zyxwvutsrqponmlkj ..../ zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQP ",..,.,,,,,, _ '" ..- , ~,.,.; ".,. I " ,.,."" ...-; __J ..,," ,, <: u -1. 0 ...... e::: I l- COMPRESSION w -1. 5 > -2.0 -2.5 -3 -2 o -1 AXIAL Fig. 6.7 2 3 STRAINzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA (X) Strain paths of NOI, SOl and UIOO samplers at 70% of 255 R from centreline 2.5 NGI SAMPLER 2.0zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA ,..... ID <, N "-J Z 0 ...... I< u --- UIOO (TYPE 1) SAMPLER 1.5zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA -------UIOO <TYPE ID SAMPLER EXTENSIONzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA 1.0 ........ .5 -~ ,zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA .' --------------------..:<"~'I ,--zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA .... _- ~.. ___ ---.. ...... 0 _J IZ W ................ SGI SAMPLER :', "'- 0.0 ~~4"~~~:h------.---~zyxwvutsrqponmlkjihgfedcbaZYXWVU .: "J ~ W _J ....~~' ... .,. .... ' ... .:/. ... , ... ' -.5 lJ.I ------- -' 0'; " '/,/ ,, _J < u -1. 0 I' I ...... I COMPRESSION I- ~ w -1.5 > - -2.0 -2.5 o -1 -2 -3 AXIAL STRAIN 2 3 (7.) Fig. 6.8 Strain paths of NOI, SOl and UIOO samplers at 90% of R, from centreline 2.5 2.0 ,..... ID <, N 1.5 Z 0 LEGEND EXTENSION "-J 1.0 .,' ...... .5 I- 0 .. ' ,-------------------- ..----.~.:::---- t _ _......: .. --- _J IZ W SAME AS ABOVE "" I- < u I 0.0 ------ ~ ..-----.::. " ___ - ~ w _J -.5 I- lJ.I _J < -1. 0 u ..... I- COMPRESSION ~ lJ.I > -1. 5 -2.0 -2.5 -3 -2 -1 o AXIAL STRAIN 2 3 (7.) Fig. 6.9 Strain paths of NOl, SOl and UlOO samplers at inside edge of sampler tube 256 2.5 --- 2.0 A. R. - ,.... CD <, 1.5 EXTENSION .!j z 0 ....... I< u 1.0 ,: I 29. 64? --- A.R." 50.737. -------- A. R." 100.467. - 53.0 .. 17.7 ..11.3 ..7.0 )' :j : ,k~···"" .5 zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA .,fi::····zyxwvutsrqponmlkjihgfedcbaZYXW 0 .....J IZ W I Bit Bit Bit Bit 10.147. ................ A. R.· _::.. '..zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA 0.0 ~ w .....J w -.5 .....J < -1.0 u ....... COMPRESSION I0::: w > -1. 5 -2.0 -2.5 L- __~~ -2.5 -2.0 ___J -1.5 ~ -1.0 -L -.5 AXIAL ~ ~ 0.0 .5 STRAIN GO ~ ~ ).0 1.5 ~ __~ 2.0 2.5 Fig. 6.10 Comparison of strain paths at 10% of R, from centreline for samplers of different area ratios 2.5 2.0zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA ,.... CD <, .!j z 1.5zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA LEGEND : SAME AS ABOVE EXTENSIONzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDC 1.0 0 ....... I- < u .5 0 .....J IZ W 0.0 ~ w .....J w -.5 .....J < -1.0 u ....... COMPRESSION I0::: w -1. 5 > -2.0 -2.5 -2.5 -2.0 -1.5 -1. 0 -.5 AXIAL 0.0 STRAlN .5 1.0 1.5 2.0 (;0 Fig. 6.11 Comparison of strain paths at 50% of R, from centreline for samplers of different area ratios 257 2.5 ~----------------------------T-----------------------------' 2.5zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA 81t ..53.0 ---A. R." 1O. 147. 2.0zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA ................ A. R. a 29.zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQP 64Y. 81t .. 17.7 I I 81t .. 11. 3 --A.R."zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPON 50.73Y. ! I Bit .. 7.0 -------A. R." 100.467. EXTENSION ! I 1.0 z zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLK 1.5 ! zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA II ~ ! }zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA j CJ I I- < U CJ .5 : I-:'!- ..... / _j IZ W ::::;: W _j W _j L5 ...... -1.0 I0::: ~ -1. 5 -2.0 -2.5 L-__~~ -2.5 -2.0 ___J ~ -1.5 -1.0 ~ -L ~ -.5 0.0 .5 AXIAL STRAIN Fig. 6.12 ~ ~ 1.0 1.5 ~ __~ 2.0 2.5 (7.) Comparison of strain paths at 90% of R; from centreline for samplers of different area ratios 2.5 2.0 ,.... ID <, N 1.5 EXTENSION ...., z 1.0 CJ ...... I< u I I I I I I I I I I I I I I I I .5 CJ _j IZ W 0.0 W -.5 ... - I I I LEGEND I I I w SAME AS ABOVE I I .....~ .. .. '" .... ~--'>..- ..-, ::::;: _j : /'~---- ..-:;;:;;~'~:---- - -" .. _j < -1. 0 ...... u r I0::: COMPRESSION I w -1. 5 > -2.0 -2.5 -2.5 -2.0 -1.5 -1. 0 -.5 0.0 AXIAL STRAIN Fig. 6.13 .5 1.0 1.5 2.0 (7.) Comparison of strain paths at inside edge of sampler tube for samplers of different area ratios 258 2.5 2.5~--------------------------------------------------' zyxwvuts AT C.L. OF SAMPLERzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA ................ 3 0 r. OF C. S. R. FROM C. L.zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA ATzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA -------- h T SOY. OF C. S. R. FROM C. L. 2.0 ---AT 70Y. OF C.S.R. FROM C.L. zzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA --~ CJ ...... lJ) lJ) L1J ~ a, ::E CJ W _,- AT 90Y. OF C. S. R. ---- hI INSIDE EDGE OF ShMPLER .s->: _-------- a-- 1.5 ",.-' Z I- l.0 lJ) _J , < ...... x -c ::s::: -c w _, ",.-" ~ ._. _._. _,_.€I ----_-El _. »->: _-- _----..0 r: /' ~' /..a-----~-------~.~.~~~.:.: " .. ' _El----:.:..:." .. ",.- z ...... < ~ _---zyxwvutsrqponmlkjihgfedcbaZY FROM C. L. / ' /'/ .> ,/ _-- _------ ....... 101 -- .[)" . ."" '~ T ,--~,.",." ~ .>: ~~ .>: / . /. ' Iil'"- ••••• '.0..... ///,(" /,,~ .. '/,:' ',~~~'~ .." . .5 ~ n, , '0 , 00 ,:'.0 !-!-o' 0.00l_--------2~0---------4~0---------6~0--------~8~O~------71~OO~------~120 AREA RATIO OF SAMPLER (1.) (c) 2.5~---------------------------------------------, .0------_. __-.0-----_ . -------------------------~ zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA ",,- 2.0 z ... .../ ...... 13"""- CJ ...... lJ) z W I- ~ 1. 5 z a---o__ '__ '_o __ .__ .--._o-.£l _a-'_' a--' z ...... < ~ tii _._. _.- l. 0 _J < ...... x < ~ w .5 n, LEGEND SAME AS ABOVE 0.OOL---------2~0----------4~0---------6~0--------~8~0---------1~O-0--------1~20 AREA RATIO OF SAMPLER (1.) (b) Fig. 6.14 Variation of peak axial strains with area ratio of samplers: (a) in compression (b) in extension 259 2.S ~----------------------------------------------------------.zyxwvutsrqponmlk z 2.0 OF SAMPLERzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIH ---- AT C. L. ................ ATzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFE 3 0 r. OF C. S. R. FROM C. L. -------- AT SOY. OF C. S. R. FROM C. L. o --ATzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA 7070 OF C. S. R. FROM C. L. , enzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA W -.AT 90Y. OF C. S. R. FROM C. L. '\ tn ~ u, ::;: o u l.S z z ...... -c ~ l- ~'\ ,'"'''' GI ,\., m-,',"\. . " 0,'".' \s, 1.0 ------ , '-, <, ,"'"' .......... b.', 'n', "" .... .......... ~ --.----- '--~' ,,--...... ........, -c ...... X -c -e ~-.....zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA ...._ ~<, ~--... .~ ...... ........_ ... ~-.-~ .__ ·__ ·__-~:-~·-....._ ......·..·::::::::::.:.7.:::-:::::-· _J :::s:: <, "'8 .... .......... "''s~' ......... ............ tn AT INSIDE EDGE OF SAMPLER .............. .......... ......... ::.---.----- --- __ .__ ~<, .. ·..·..::::.';.7.--_ -__ •S .__ ~'-..... .......~.';'.~.:::--...... ...... -~- ui n, --- 0.0 ~--------~--------~--------~--------._--------~------~ o ]0 20 30 8/t __ . ..... :-.~:-:::::.,..;.'" ":-- 40 •__ '~I!) ':---. SO 60zyxwvuts RATIO OF SAMPLER Ca) 2.S ~----------------------------------------------------------; G......, "'s.._ LEGEND .......... _- '""6- I SAME AS ABOVE ... __ -------_ -----_ zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA ------------------e z o ......zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCB 2.0 in Z W l- X W l.S Z z ...... < ~ l- tn 13-........ ~'- o ----a._._._._._._._._._ '_'-'--£I 1.0 _J < ...... x < :::s:: < w •S u.. 0.0 ~------~~------~--------~--------~--------~------~ o 10 20 30 8/t 40 SO RATIO OF SAMPLER Cb) Fig. 6.15 Variation of peak axial strains with Bit ratio of samplers: (a) in compression (b) in extension 260 60 2.5 r---------------------------.---------------------------,zyxwvutsrqponmlkjihgfedcb I 2.0 I 1.5 I """,,/zyxwvutsrqponmlkjihgfedcbaZYXWVUTSR EXTENSION z C) ,,,,j"" 1.0 ,---- izyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONML .> zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA .r 1/ I- < W -",-'-" /./:! .5 O• 0 ...... I ---------~--~~ r- _.J ••~~~~~~~ '_____ -- ..- " .... , .:zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA : '. j,' :' I _J -1.0 I0:: ~ \.. __._••~ ~~------_+--------_+--------_1 •..•..• -.5 W l5 ...... :'" '._ ~ W ..... (. "" C) _.J IZ W / I --................ 1. C.R.- 0.4957. 8/t J. c. R. a O. 9901. BIt _._ 1.C.R." 1.980r. -------- I.C.R.- -1.5 3.9601. .. .. 17.0 I. IzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA a 17. 7 - , BIt.. 19.2 BIt - 23.0 COMPRESS ION -2.0 -2.5_~6--------~-4----------~2---------0L-------~2--------~4--------~6 AXIAL STRAIN Fig. 6.16 (7.) Comparison of strain paths at 10% of R, from centreline for samplers of different inside clearance ratios 2.5 LEGEND : SAME AS A80VE 2.0 ,.... CD <, N 1.S EXTENSION '-' z 1.0 C) ...... I- < u .5 Cl _.J I- Z -- _----- - .__ ..:.: 1'--.\."'\ 0.0 W ~ w _.J w -, , -.5 ) ! \:i//zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCB '/i" _.J < -1. 0 w ...... I0:: w -1. 5 > COMPRESSION -2.0 -2.5 -4 -6 -2 o AXIAL STRAIN Fig. 6.17 2 4 (i0 Comparison of strain paths at 50% of R, from centreline for samplers of different inside clearance ratios 261 6 2.5 ,.... CD <, N ~ z CJ ...... I < I " zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA f""" ....zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPON : EXTENSION " " 1.0 " ,;. ........,J.. V· """", .5 """, "" _J w _J w " .,.- CJ fZ W ::E , 1.5zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA f- u I f I I I I I I I 2.0zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA ".,._, '.... 0.0 ...--...: ...--....--........ ....... . C. - ...---.-~...~::::.--....zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPON ...... ~.. -~ » t " -.5 .. ) ~:..... .: • '• .' 0.4951. 17.0 Bit J. C. R. =zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA ................ 1. C. R. = 0.9907. 8/t = 17.7 _J < u -1. I f 0 f- a::: w -1. 5 > _.- J. C. R. = 1.9807. 8/t -------- I. C. R. = 8/t 3.9601. = 19.2 = 23.0 COMPRESSION -2.0 -2.5 o -2 -4 -6 AXIAL Fig. 6.18 STRAlN 4 2 6 (7.) Comparison of strain paths at 90% of R; from centreline for samplers of different inside clearance ratios 2.5 LEGEND 2.0 ,.... CD <, N ~ z I SAME AS A80VE 1.5 EXTENSION 1.0 CJ ...... f- < u .5 CJ _J fZ W ::E w --' lJJ 0.0 -.5 --' < u -1. 0 ...... COMPRESSION f- a::: w -1. 5 > -2.0 -2.5 -4 -6 o -2 AXIAL Fig. 6.19 STRAIN 2 4 (7.) Comparison of strain paths at inside edge of sampler tube for samplers of different inside clearance ratios 262 6 6 --- AT C. L. OF SAMPLER ................ ATzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA 30r. OF C. S. R. FROM C. L. -------- AT sore OF C.S.R. FROM C.L. 5 z C) 4 ......zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA (J') Z ._ LU X LU Z ...... z ...... -< 3 ~ ._ (J') --' -e ...... x -e :::s= -< 2 LU CL 1 o o 1 2 4zyxwvutsrqponmlk 3 INSIDE CLEARANCE RATIO OF SAMPLER (r.) Fig. 6.20 Peak axial strain in extension versus inside clearance ratio of sampler 263 6 5 ,.....zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLK - N ~ lJJ _J CL :::E: < U) u, CJ AT C. L. OF SAMPLER 4 Cl < :c -e ................ AT 30% OF C. S. R. FROM C. L. -------- AT 50% OF C. S. R. FROM C. L. --- AT 70% OF C. S. R. FROM C. L. _.- AT 90% OF C. S. R. FROM C. L. ---------- AT INSIDE EDGE OF SAMPLER lJJ z Cl U) U) lJJ ~ 3 CL :::E: CJ u - z z < ~ tU) 2 _J < e-e < ~ < lJJ CL 1 o o 1 2 3 4zyxwvutsrqponml INSIDE CLEARANCE RATIO OF SAMPLERzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA er.) Fig. 6.21 Peak axial strain in compression (ahead of sampler) versus inside clearance ratio of sampler 264 2.5 ~--------------------------r---------------------------' 2.0 1.5 EXTENSION ----- I. C.E. TAPER ANGLE· 0.358 DEG. ................ 1. C. E. TAPER ANGLE • O. 7 J 6 DEG. _._ I. C.E. TAPER ANGLE a 1. 432 OEG. 1.0zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA zzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA Cl I.C.E.a INSIDE CUTTING EDGE I- ;:J .5 Cl _j ~ 0.0 r---~~---1-----+---=~~~~~~:r----~----;-----+---~ w ::E ~ w -.5 _j ;:J -1.0 ...... COMPRESSIONzyxwvutsrqponml I- a:: ~ -1. 5 -2. 0 -2.5 L-__~ -2. 5 ~ -1. 5 -2. 0 ~ -1. 0 ~ -.5 ~ __~~ __~ O.0 .5 1. 0 AXIAL STRAIN Fig. 6.22 2.5 ~~~~ 1. 5 __~ 2. 0 2. 5 (7.) Comparison of strain paths at 10% of R, from centreline for samplers of different inside cutting edge taper angles ~---------------------------r--------------------------~ 2.0 1.5 z 1.0 ...... I< .5 SAME AS ABOVE LEGEND EXTENSlDN CJ U Cl _j ~ 0.0 r---~r---_,----~~~~~~=f~~:t~---r----;-----+---~ w ::E ~ w -.5 _j ;:J -1.0 ...... COMPRESSJON I- a:: ~ -1. 5 -2.0 -2.5 L-2.5 L- -2.0 L- -1.5 ~ -1.0 ~ ~ ~ ~ ~ -.5 0.0 .5 1.0 1.5 AXIAL STRAIN Fig. 6.23 ~ __~ 2.0 (7.) Comparison of strain paths at 50% of R, from centreline for samplers of different inside cutting edge taper angles 265 2.5 2.5 ~--------------------------~----------------------------, I! 2.0zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA --- I! 1.5 a O. 358 OEG. ................ 1. C. E. TAPER ANGLE • O. 716 DEG. I! EXTENSION 1. C. E. TAPER ANGLE _._ 1. C. E. TAPER ANGLE· 1. 432 DEG. II 1.0 _J tJ -1.0 .....zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLK COMPRESSION l- e::: ~ -1. 5 -2.0 -2.5 L-__~ -2.5 -2.0 ~ -1.5 ~ -1.0 ~ AXIAL Fig. 6.24 2.5 CD <, N '-J 0.0 .5 ~ ~ 1.0 1.5 ~ __ ~ 2.0 2.5 STRAINzyxwvutsrqponmlkjihgfedcbaZYXWVUT (i.) Comparison of strain paths at 90% of R; from centreline for samplers of different inside cutting edge taper angles ~--------------------------r--------------------------. II 2.0 ,.... ~ ____J -.5 Ii Ii II. : 1.5 LEGEND : SAME AS ABOVE zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIH EXTENSION z 1.0 -e .5 Cl ..... IU Cl _J IZ W ~ W _J -.5 W _J tJ ..... -1.0 COMPRESSION l- e::: ~ -1. 5 -2.0 -2.5 L- __~~ -2.5 -2.0 __~ -1.5 ~ -1.0 ~ -.5 AXIAL Fig. 6.25 ~ 0.0 STRAIN -L .5 ~ ~ 1.0 1.5 ~ __~ 2.0 (i.) Comparison of strain paths at inside edge of sampler tube for sampler of different inside cutting edge taper angles 266 2.5 2.5 --------------------------------------------~zyxwvutsrqponml ,....a------ --------- _----0zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA 13""--/-- 2 AT C.L. OF SAMPLERzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDC ................ --------------------.----z zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA ...... ---aJ 1.5 C) lI') AT 30% OF C. S. R. FROM C. L. AT AT AT AT 50% OF 70% OF 90% OF INSIDE C.S.R. FROM C.L. C.S.R. FROM C.L. C.S.R. FROM C.L. EDGE OF SAMPLER l- X W Z ...... z ...... < et::: l- ll') ~ ...... x 1 < :x: < w a, .5 o o .5 1 1.5 INSIDE CUTTING EDGE TAPER ANGLE (DEG.) Fig. 6.26 Variation of peak axial strain in extension with inside cutting edge taper angle of sampler 267 2 2.5 2.0 ----@ <, - EXTENSION -.- 1.0 J 0 u .5 0 O. C. E. TAPER ANGLE = 19. 29 DEG• IzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA O.C.E.= .: J ../ I- <: ANGLE • 5. 00 DEG. ................ O. C. E. TAPER ANGLE .. 9. 90 DEG. 1.5zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFE N ....., z o. C. E. TAPER OUTSIDE CUTTING EDGE _j 0.0 IZ W x w _j w -.5 _j -c -1.0 u ...... COMPRESSION I- 0:: w > -1. 5 -2.0 -2.5 ~ __ ~~ __ ~ -2.5 -2.0 -1.5 ~zyxwvutsrqponmlkjihgfedc -L ~ ~ __~~ __~~ __~ -.5 0.0 .5 1.0 1.5 2.0 2.5 ~ -1.0 AX I AL STRA 1N Fig. 6.27 (7.) Comparison of strain paths at 10% of R; from centreline for samplers of different outside cutting edge taper angles 2.5 ~----------------------------~----------------------------, 2.0 1.5 LEGEND EXTENSION z o ...... -c u o .5 I- 0.0 ( _j z x ~ w /,/. / ~----~----~----~--~~~~+-~==+-----+-----+-----T---~ ...... ) .... / ............. -.5 .' ............ ......./ .... .>. zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA ...J LS ...... /. -1. 0 ... I- COMPRESSION (/ 0:: ~ SAME AS ABOVE 1.0 I- w I -1. 5 -2.0 -2.5 ~--~~--~----~----~----~----~----~----~----~--~ -2. 5 -2. 0 -1. 5 -1. 0 -.5 O. 0 AX I AL STRA I N Fig. 6.28 .5 1.0 1.5 2.0 (7.) Comparison of strain paths at 50% of R; from centreline for samplers of different outside cutting edge taper angles 268 2.5 2.5 2.0zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA ...... CD 1.5 <, EXTENSION N ....., z --- O. C. E. TAPER ANGLE" S.OO OEG• ................ O. C. E. TAPER ANGLE" 9.90 _.- O. C. E. TAPER ANGLE" 19.29 OEG• DEG.zyxwvutsrqp 1.0 Cl >-1 I- <: u .5 Cl _J IZ W :::0: w _J w 0.0 -.5 _J <: u -1. 0 >-1 COMPRESSION I0:: w -1. 5 > -2.0zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA -2.5 zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONM L~ ~ LL~ ~ ~ __ ~~--~~--~ -2.5 -2.0 -1.5 -1.0 -.5 0.0 .5 1.0 1.5 2.0 2.5 AXIAL Fig. 6.29 STRAIN (7.) Comparison of strain paths at 90% of R, from centreline for samplers of different outside cutting edge taper angles 2.5 2.0 ...... CD 1.5 <, z LEGEND EXTENSION .....,N SAME AS ABOVE 1.0 Cl >-1 I- -c u .5 Cl _J IZ W :::0: w _J w 0.0 ................ ::- · -__ · .7 ;::__... __ .__ . .... ;;>/ ......... -.5 _J <: u -1. 0 /. ..... ... I0:: :. ' ... COMPRESSION / w -1. 5 > -2.0 -2.5 -2.5 -2.0 -1. 5 -1. 0 -.5 AXIAL Fig. 6.30 0.0 STRAIN .5 1.0 1.5 2.0 (;0 Comparison of strain paths at inside edge of sampler tube for samplers of different outside cutting edge taper angles 269 2.5 2.5 ~--------------------------------------------------------__.zyxwvutsrqponmlkjih --- AT C.L. OF SAMPLER 30Y. OF C.S.R. FROM C.L. ATzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA --------AT SOY.OF C.S.R. FROM C.L. 2.0 zzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA --AT 70Y. OF C.S.R. FROM C.L. Cl lJJ lJJ W _.- AT 90r.OF C.S.R. FROM C.L. 0:: Cl... ::::E Cl 1.5 U Z z ...... -e 0:: IlJJ 1.0 _J <: ...... x <: ::os:: <: .5 w Cl... O.OOL-----------5~----------lLO-----------l~5-----------2~0----------~25 OUTSIDE CUTTING EDGE TAPER ANGLE (DEG.) (c) 2.5 ~--------------------------------------------------------__. LEGEND : SAME AS ABOVE 2. 0 z _.B---_-_a---- ... --- --- --.0 _-------zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA .. Cl ...... lJJ Z W lX W 1.5 Z ...... a--'_' z ...... -'_ __ __ __ __ . a--'--' . . -e 0:: tn 1. 0 _J -a------- --0------- ... _--0 ---El <: ...... x -e ~ w '_.--EI :~.~.~.~.~.~~~.~.~.~.~.~~~.:.~.~.~~~.~.~.~.~.~~~~.~.~.~.~ ....~ o B .5 Cl... 0.00~----------5~----------1~O-----------1~5-----------2~O----------~25 OUTSIDE CUTTING EDGE TAPER ANGLE (DEG.) Cb) Fig. 6.31 Peak axial strain versus outside cutting edge taper angle of sampler: (a) in compression (b) in extension 270 2.5 2.0zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA BIt .. 45. 6 --- --Bit • 23.0 1.5zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA ,.... CD <, N EXTENSION '-' l.0 z -------- BIt· _._ BIt .. 12.2 19.9 = 1.25 mm t • 2.50 mm t t .. 5.90 mm t .. 4.90 mmzyxwvutsrqponm CJ _, I- -c .5 u CJ _J 0.0 I- z W ::::0: w -.5 _J w _J -c u _, + l,0 I0::: w > COMPRESSION -1. 5 -2.0 -2.5 -12 -9 -6 o -3 AXIAL Fig. 6.32 6 3 STRAIN 9 12 (7.) Comparison of strain paths at 10% of R; from centreline for various flat-ended samplers 2.5 2.0 r"o OJ <, II :,:1 1.5 z , :1 EXTENSION !j ,I ::,I 1.0 0 I- -e .5 _J _J w SAME AS ABOVE \,J~\ CJ IZ W ::::0: W I I :' ....... u LEGEND ;~~, 0.0 ) )- -.~ ~~) // -.5 _,/' r-: _J -< u -1.0 _, /'zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQ 11'/' V I0::: w -1.5 > COMPRESSION -2.0 -2.5 -12 -9 -6 -3 AXIAL Fig. 6.33 o STRAIN 3 6 9 (7.) Comparison of strain paths at 50% of R; from centreline for various flat-ended samplers 271 12 2.5 --2.0zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA Bit 45. 6 a t a 1.25 mm I ! --- Bit ..23.0 t .. 2.50 mm ·I I, ,... 1.5 zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGF t = 5.90 mm Bit .. 19.9 EXTENSION ·I I, -------_._ Bit = 12.2 = 4.90 mm 1.0 z ·I I, .5 I I u CD <, N "-' t Cl I- <: Cl ...J IZ W :::0: w ...J w ~:~ 0.0 -.5 ...J -e -1.0 u ....., l- n:: w -1. 5 > -2.0 -2.5 -12 -9 -6 o -3 6 3 AX I AL STRA I N 9 12 (X) Fig. 6.34 Comparison of strain paths at 90% of R, from centreline for various flat-ended samplers 2.5 I I, I I, 2.0 ,..... CD 1.5 "N z LEGEND SAME AS ABOVE I I I I I I I _lzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIH ~ ~;~__ ._._-:r EXTENSION '-' I I 1.0 Cl ...... I I- -e u .5 Cl ...J IZ W :::0: w _J w 0.0 /:: .~.~>.--..--- -.5 '/' _J <: u ...... -1.0 V' l- n:: w -1. 5 > COMPRESS JON -2.0zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA -2.5 -)2 -9 -6 -3 AXIAL o STRAIN 3 6 9 GO Fig. 6.35 Comparison of strain paths at inside edge of sampler tube for various flat-ended samplers 272 12 Clzyxwvutsrqponmlkjihgfedcb Lf1 ~ w ....J c, ~ -c U) u, CJ ..J _j ....J U U U W ::E CJ ::E CJ ::E CJ ::E CJ u, lJ.. a:::: u. a:::: U. u I-- < a:::: a:::: n, ::E Ul u. u. u, Cl Cl U. N Cl N Cl t.n N Cl ('T) I-- I-- -e -e r-I-- U) UJ Cl z -c I u Cl ....J I ....J ....J Z ....... <: I-- I I -e <: I I I I I / / / / I / .,.,/" [3"" ../" " I .3 .- /" / r:(""" ./' /' ;" (3; t: et' (T) ...... Cl 00 I I,' / Cl N t ...., <, OJ / / ~ zyxwvutsrqponmlkj 1=:.8 '§.g til ~ Cd ~ .~~ i~ / ~§..... I=: en o I=: 'p 4.) * 'Ia ~ / ~ ::E ...... ::E Cl U ....... w X I-<: x ::E w ::E ('t') \t:i .bh .... ~ U) z N .-.zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA Cl .... >.5 \0 ..... ::J 273 8·~ til CJ / w GO NrV~lS lVIXV ~r;:: f;/ CJ Z Cl (D I ...... l< 0:= U) Cl N ~ I* ::E: Cl a:::: ::c: a.. 5 til en .... :f .... U) CL o .5 > U) CL Cl _J ..... u, i :f Cl ...... < UJ w 8'~ />.<.... z 1=:'0 0:= -e },~.: ..... ~1S I I I />.~.~ ... ' / ,,;. ",' r1 / ~8' 'Oen I I I I ,'f j I ,:! ,. I :: / I m '§ I I ,,'f I I /,:f,:,: / I " / I I-- I .- /" / I U) -e I I-- w w Cl ....... N Cl Cl I I ....... Cl I-- -e CJ L!) u, Cl i I U) U) ! ,: / II! // ,':III t til I Cd I ....~ I ~ Cl I -.:t S :::I ~ I .§~ zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA I / ,.t:t ; I S ~ I <: et::: . ~ . ~ · et:::·· . . u· w· u w U) ~ UJzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONML I=: ....J ~ U) Cl ....J · · · _j 2.5 ~-------------------------r------------------------~zyxw I I 2.0 ! NGJ SAMPLER BItzyxwvutsrqponmlkjihgf -45.6 I : -------- FLAT-ENDED SAMPLER BIt· 45.6 ,, ,,zyxwvutsrqponmlkjihgfedcbaZYXW 1.5 I EXTENSION z zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA 1 .. 0 : o : I :: 2S .5 _j \, '( " - ~~~-~~~"'1-_ ~----~----~----~----+-~~_~--~~~~--~_t_-----+-----T----~ -~ ~~~ ~ z 0.0 LU ~ LU _j LU " " -.5zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGF c-:" ;:s _j " ~zyxwvutsrqponmlkjihgfedcbaZYX "", " /zyxwvutsrqponmlkjihgfed -1.0 COMPRESSION I I I f!' -2.0 -2.5L_--~----~--__j~--~~~~--~~--~--~~5~~2~0~~2 -2.5 -2.0 -1.5 -1.0 -.5 0.0 AX 1 AL STRA 1 N .5 ],0 1. 5 .. (X) (0) 2.5 2.0 ,..., CD <, N 1.5 EXTENSION '-' z LEGEND : SAME AS ABOVE 1.0 0 ~ < LJ CJ .5 _j ~ z w ~ w _J w -------- ~- 0.0 -.5 _j < u -]. 0 ..... ~ Cl:: w -1. 5 > -2.0 -2.5 ~--~~--~----~----~----~----~----~----~----~--~ -2.5 -2.0 -1.5 -1.0 -.5 AXIAL 0.0 STRAIN .5 1.0 1.5 2.0 2.5 (I.) (b) Fig. 6.37 Strain histories of NGI sampler and a flat-ended sampler of identical BIt ratio: (a) at 10% of R, from centreline (b) at 90% of R, from centreline 274 2.5 2.0zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBAII ,.... BIt - 12.2 -------- FLAT-ENDED SAMPLER BIt .. 12.2 1.5zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA , zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHG EXTENSION I CD <, N -.J SGI SAMPLER I I I I I I I :z: I I 1.0 Cl ,_., I- -< u \._", .5 Cl _J I- :z: w ~ w _J w -( _ zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLK "'-zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA ... ...zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA ---------- ;;;. 0.0 -.5 " " " " ~t1;"" , -c -1. 0 ,, u COMPRESSION I I0:: > ) , " _J w ---------, ..... " I v, -1. 5 I -2.0 -2.5 -8 -6 -4 o -2 2 AXIAL STRAIN 8 6 4 (7.) (c) 2.5 , , ,, 2.0 I ,..., 1.5 zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA LEGEND EXTENSION .!j 1.0 :z: CD <, , l ~. Cl ,_., I- -< u .5 " ~~--- -_ _J :z: w ~ w _J w 0.0 --- -e -------- ------- -,-------,-------------,. -. 5 ,-."., "_",, _J u SAME AS A80VE , Cl I- I -1. 0 " " I , ,, EXTENSION I0:: I COMPRESSION I w , > -1. 5 -2.0 -2.5 -B -6 -4 -2 o AXIAL STRAIN 2 4 6 (7.) Cb) Fig. 6.38 Strain histories of SOl sampler and a flat-ended sampler of identical Bit ratio: (a) at 10% of Ri from centreline (b) at 90% of R, from centreline 275 B 2.5 --- 2.0zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA UIOD (TYPE UIOO (TYPE I I I I I I I I I I I ,.... J) 11) SAMPLER SAMPLER 1.5zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA -------- FLAT-ENDED SAMPLER <, NzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA EXTENSION 1.0 Z zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQP C) zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA :0 ...... co -...J -: I I- < u .5 C) _j I- Z ~"" 0.0 W ......... ..... _ .. ------ .. _ :::E: W _j w -.5 _j -e -1.0 Bit RATIO OF EACH SAMPLER IS 19.9 u ...... I0:: w -1.5 > -2.0 -2.5 -5 -4 -3 -2 -) AXIAL o 2 STRAIN 3 4 5 OD (a) 2.5 2.0 ,.... co LEGEND : SAME AS ABOVE 1.5 <, N EXTENSION -...J Z Bit RATIb OF EACH SAMPLER 1S 19.9 1.0 C) ...... I< u .5 C) _j I- Z 0° :::E: w _j w --,..------- .. --~ ............................. 0.0 W ...... -.5 I- _---------------- ~.::~~.~.::~.--- --- ......,, _j ~;' -e -1. 0 Iu ...... : , :1 1 I0:: COMPRESSION w -1. 5 > -2.0 -2.5 -5 -4 -3 -2 -) AXIAL o STRAIN 2 3 4 5 (r.) (b) Fig. 6.39 Strain histories of UIOO samplers and a flat-ended sampler of identicalzyxwvutsrqponmlkj Bit ratio: (a) at 10% of R, from centreline (b) at 90% of R, from centreline 276 ,....zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA OJ m 0 +J (f) m c .,.., 1.2 "0 0 0 .....zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA s: o 0 OJ 40 "0 1.0 c OJ OJzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA s: +J +J 0 C) ._ >-1 < et: 0.8 CJ >-1 C) >zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA 0.6~~--'_~~~~------~----'_~--~~~~~ 100 30 1000 VERTICAL EFFECTIVE STRESS (kPc) Fig 6.40 Void ratio - log cry relationship for normally consolidated London Gay 277 0.9 ~ , 0.8 l- O. 7 I- " I' III l0.6zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA '"' :z -... 0.5 I-, .......... E ......zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA ::E "' •.. E > 0.4 E - ,...... ~ 0.3 \1/ I- ~ 0.2 r~ O. ] I-zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA l_ 0.0 I I I I I I I ]0 I I I lOO I I 'II I I IJ ]000 VERTICAL EFFECTIVE STRESS (kPc) Fig. 6.41 Compressibility/expansibility - log o', relationship for normally consolidated London Clay 278 2.00r--------------------------------------~ z o ..... 1. 75 1------------, 1. SO I- 1. 25 F------~------- --- - - ---- 'Jt Method 109' t Method I- < ..... Cl _J o Ul z o 1. 00 -- U 4. o IZ UJ ..... U ..... 4. 0.75 - 4. UJ I u y o I 0.50 - I L __ + __ ~~~~+--~~--~--"'--'+--"_--~--~ ... 0.25I- p O.OO~----~I---~I-L-I~II-~II~I~I------L-I~I~~I~I-~II-ILUIzyxwvutsrqponmlkj ID 100 1000zyxwvuts VERTICAL EFFECTIVE STRESS (kPc) Fig. 6.42 c, - logzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA o', relationship for normally consolidated London Gay 279 3.0 'It M'Ithod zyxwvutsrqponmlkjihgfedcbaZYXWVU ----]og.t M'Ithod 2.5zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA tF 1, ,.... Ul <, E 2.0 I-zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA 0...... TO r------~------ .. X .. .l:: >~ _, 1.5 l- .....zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA _J co < w ~ a=: w " I I I V I L_.,___ I 1.0 r- CL ... --. - 0.5 t- --"'--1 _l 0.0 _l I I I I 100 I I I 10 ---'+--- I I I _l _l .irr 1000 VERTICAL EFFECTIVE STRESS (kPo) Fig. 6.43 Permeability - log cry relationship 280 for normally consolidated London Clay 1.4zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFE .---. 0zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA O'lt Method ,...zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA OJ 01 0 log.t Method +' (J) 01 .....c 1.2 "'IJ 0 0 .....zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA s: u 0 OJ 40 "'IJ c 1.0 OJ OJzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA s: +' +' ....,0 Cl ...... I- 0.8 < c::: Cl ...... Cl > PERMEABILITY Fig. 6.44 (m/s) Void ratio - log. permeability relationship for normally consolidated London Clay 281 225 ~------------------------------------------------~zyxwvutsrqponmlkjih 200 TEST 1 -------- TEST 7 ................ TEST B .-: _, .' "zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQP zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJ ....zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJ V) V) ,/ W 0:: fV) 175 ,,,,' ;,' _J , < U f0:: W > " ! / - -- > ...... .....,.-/. fU W u, W ...: ,;'zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA . W u, f ... ,.<iI'" 150 .~zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIH ----- . . .... 125 --»>: -------.:.-----.-.-.-.-.-.. / ~~-----=~--~ ,.' .' " tII' 100 ~~~~ O.00 _L .05 ~ .10 ~ .15 .20 .25 .30 .35 LOCAL RADIAL STRAINzyxwvutsrqponmlkjihgfedcbaZYXWVUTSR (r.) Fig. 6.45 Typical vertical effective stress versus local radial strain plots 225 r-------------------------------------------------------. LEGEND SAME AS ABOVE 200 ,-'zyxwvutsrqponmlkjihgfedcbaZYXW If) ill'" V) W 0:: fIf) -- - -:--- ---- ::::: 175 _J < u ...... > > ...... f- .'.' .' U W W . , 150 W u, u, // ........... . f0:: W ,,' ",,,,, 125 /.~~~~ ....... .' "" " .' . ,'.... ;.~ ..... ,~.t- 100 U~~· 023 L- L- L- L- L--- ~ --: 4 5 6 7 LOCAL AXIAL STRAIN (7.) Fig. 6.46 Typical vertical effective stress versus local axial strain plots 282 7.-------------------------------------~--~ /' ,-:~. ...,.' /. ,.' N ............ TEST ] 6 ,.$~" ,..'.''.' N ------ ,(~.~ .. TEST B '.' 5 N :z:zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA ..... < ,~.~ ... ,..,_. ,,.'.' Cl::: lll) ,'.' .' ,,.' .' 4 ,~.~.~ .... ,..,.. W ..... ,~, .',.' Cl::: IW :::E: :::J True Ko-line ,..'.' ,.' ,~.~.~.~ ... ,_. ,.. ,.' /~.~ .... N 3 _J CJ > _J < w 2 ,,.'.'zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA I~.~~··· ,..,.' ,,.'.'zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA CJ _J I~'" ,.''.'zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA I.!o··zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA N !,' °0~----L-----~2----~3----~4~----5~----~6----~7 LOCAL AXIAL STRAIN (7.) Fig. 6.47 Typical local volumetric strain versus local axial strain plots 20 ~--------------------------------------------------------__, 15 w Cl::: :::J ---- TEST 1 --- TEST 3 _.- TEST 5 ....•........... TEST 6 -------- TEST 7 lI) lI) w Cl::: o, 10 W 0::: CJ a, lI) lI) w w x w 5 EFFECTIVE VERTICAL STRESS (kPo) Fig. 6.48 Typical excess pore pressure versus vertical effective stress plots 283 ]00 ~----------------------------------------------------------. 90zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA ....................................................................... ..... ........zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONM ~-. ---------------------------- --- -"'-zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA ... (f) (f) 80zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA LW a::: f(f) a::: CJ f- -c ,_, ................ No correction 70 oppl ied at all > LW ---- Area corrected us ing convent iana 1 method CJ ------- Area corrected using -------- Correction applied actual diameter of the specimen for both area and rubber 60zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA 50 ~ o ~ _L ~~ ~ ~ EXTERNAL AXIAL STRAIN ~ 5 4 3 2 membrane ~ 6 7 (7.) (a) 100 ~----------------------------------------------------------~ 90 (f) (f) 80 LW a::: f(f) a::: CJ f- -e ,_, 70 LEGEND SAME AS ABOVE > LW CJ 60 50~------.-------~------~------~------~--------~-__~ 7 o 3 2 5 4 6 EXTERNAL AXIAL STRAIN (7.) (b) Fig. 6.49 Comparison of deviator stresses corrected using different approaches: (a) Test 1 (b) Test 7 284 120zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA 80 ,zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA eT -40 -80 p'" (kPa) Fig. 6.50 Stress paths for compression and extension tests in p' - q' space 285 90 ~----------------------------------------------------~zyxwvutsrqponmlkjihgfedcbaZYXWV ................. 85zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA .................. ..•. zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA .........zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA ..'.. ' 80 tf) tf) W 0:: Itf) --- / EXTERNALLY LOCALLY MEASURED MEASURED 75 0:: Cl I- < > w 70 Cl 65 600L---------L---------2L---------L3---------L4---------7S---------:6 AXIAL STRAIN (7.) (0) 80 --- 60 EXTERNALLY MEASURED ••...........••. LOCALLY MEASURED 40zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA r-.. 0 o, 20 .xzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA o..J tf) tf) UJ 0 -, 0:: Itf) 0:: ............ -20 Cl I- ................... -c ....... > UJ -40 Cl -60 -80 -100 0 2 4 10 8 6 AXIAL STRAIN ]2 (7.) Cb) Fig. 6.51 Deviator stress versus axial strain plots: (a) Compression test (b) Extension test 286 14 16 enzyxwvutsrqponmlkjihgfedcbaZY .... 0 ~ .~ ::sen r1 " N 'oJ Z ....... < 0::: 'a .~ -~ e~ ~ en ~ c:: ~ 0 en Q) .... I-zyxwvutsrqponmlk X -=~ < en en ....0 (J') _. < ....... cl .c:: ~~ en ~ Q) _.< t ac::zyxwvutsrqponml ;> Z 0::: UJ l- x w g~ en ~ I-< e- ~ 0 '> 5 o~ Co) Q) l- U') LLI II-U') LLI M Z l- .... 0 -Z _ U') U') 0 0 cl LLlU') 0:: Z Cl..LLI XI- ox U LLI N 0 0 0 0 CD 0 1.0 0 oot ~ oh u: zyxwvutsrqponmlkjihgfedcbaZYXWVUT .0 0 II) 0 0 N 0 N I 0 oot I 0 1.0 I ocl CD I (Dd~)zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA SS3~lS ~OlvrA30 287 ..... .~ In ,..., N ...... :z ...... cl 5 < ~ .... e:: (J) bile:: l- ..... .... tUzyxwvutsrqponmlkjihgfedcbaZY ~ tU .... ell ~ ...J .... ~ 0 • eIl 09 < ...... X -e ..... >< zyxwvutsrqpo V ~ ~ ...J tt:l~ < :z ~ lLJ I- x lLJ I g~ 'J:2 In :a ell ....§ ~'J:2 ~~ CIle ~ II') I.d .... bh if: Cl Cl Cl Cl ID (Y) Cl Cl Cl Cl N 0 Cl Cl ID N Cl Cl Cl N .... od Cl Clzyxwvutsrqponmlkjihgfedc ..,. (Cd~)zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA SS3N~~rlS INVJ3S .... .~ ell tU ,..., N ...... z ...... < 0:: 'a y .~ s tU~ ~ l- . ..... Cl (J) ...J .e ~8" -e ...... .....0 -e ~ ~ X ...J < Z 0:: lLJ I- X lLJ I ell Cl Cl Cl .... (0 Cl Cl Cl Cl Cl Cl Cl ..,. Cl .... N Cl CD (Cd~) SS3N~~rlS lNVJ3S 288 ell .... e:: ~.~ ~i I.d -- U ~; '..t= ('f') II') Cl ell bh if: 40 35zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA 30zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA '"' a CLzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA x ....... UJ L!) z -e :z: 25 w UJ a:: 20 ::::J lJ) lJ) UJ a:: 15 CL UJ a:: CJ CL 10 5 0 0 3 2 4 EXTERNAL AXIAL STRAIN 5 6 (7.) (a) 6 4 '"' a 2 CL x ....... UJ L!) z -c :z: 0 w w a:: -2 ::::J lJ) lJ) w a:: -4 CL l..&J a:: 0 CL -6 -8 -10 0 2 4 6 8 10 EXTERNAL AXIAL STRAIN 12 14 (7.) Cb) Fig. 6.55 Pore pressure change versus external axial strain plots: (a) Compression test (b) Extension test 289 16 40~------------------------------------1zyxwvutsrqponmlkjihgfed ,.... 30zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA o n, ..l: ...., I..i.J l!l Z < :J: U Yi"ld point 20 UJ a:: ::::J Ul Ul I..i.J ~_~ _~_- Isotro\PiC"las::~~:_ a:: n, I..i.J a:: ~ 10 ~~ __ .- __ O~~~~--~--~--~--~--~--~--~ o _ ~.,... .-~ ;";" 10 20 30 40 DEVIATOR STRESS CHANGE (kPc) Fig. 6.56 Au - Aq relationship for undrained compression test indicating probable yield point l50~---------------------------------------------------------------------------------' 125 100 c ~ ...., 75 '0- so 2S 25 50 75 100 125 150 p' (kPc) Fig. 6.57 Approximate location of yield point on the stress path in undrained compression test 290 1.2~--------,-------------------------------~zyxwvutsrqponmlkjihg Solid symbol represents presheor condition 1.0 .8 .6 '0:- ,~ .4zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA .2 O.D~---4----~~--;-----r---_'~--~~/r--r----~ zyxwvutsr ,zyxwvutsrqponmlkjihgfe I I , £,~ -.2 -LIS,., ,/ _.4L-. 4 -L ~ -.2 D. D _L ~ •2 •4 ,, , ~-----~~------L-----~ •6 1. 0 •8 1.2 Fig. 6.58 Normalised stress paths in pi -zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQ q' space for test 4 1.2.-------------------r--------------------------------, Solid symbol represents presheor condition 1. D ·8 ·6 'ci.,~ .4 •2 O.D~--~~---4~---+----~-----r-----~£~/7~~~~.6~%~----, I , I -.2 I ,, , , I -.4L-------L--------~----~---------~-----~-----~~-----~~~ -.4 -.2 D.O. 2 .4 .6 .8 1.0 1.2 Fig. 6.59 Normalised stress paths in p' - q' space for test 7 291 1.2~--------~----------------------------'zyxwvutsrqpo Solid symbol represents presheor condition 1.0 .8 .6 I I I I I '0:- 'd: .4 ,,zyxwvutsrqponmlkjihgfe I I ,zyxwvutsrqponmlkjihgfedcba .2 I t). -0.27% I O.O~--~~--~~---4-----4-----+-----+-,r'--t----; ,,, I ,,, I -.2 ,, I -.4L---~----~--~----~--~--~~--~~~ _.4 -.2 O.0 .2 .4 .6 1. 0 .8 1. 2 Fig. 6.60 Normalised stress paths in pi -zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONML q' space for test 8zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPO 1.2----------r---------------------------~ Solid symbol represents presheor condition 1.0 .8 .6 '0:- 'er .4 .2 O.O~--_4----~~--~----,_----~T_~r-T,--+---~ I , ,, I -.2 .,-1" I £ ,, I _.4L-.4 -L -.2 L- O.0 ~ .2 ~ ~~~ .4 .6 .8 ~ 1. 0 ~ 1.2 p~p; Fig. 6.61 Normalised stress paths in p' - q' space for test 3 292 1.2~--------,---------------------------1zyxwvutsrqponmlkjihgfedc 1.0 Solid symbol represents presheor condition .8 .6 '0:- ,-;;::..4 .2 -.2 _.4L_--~--~zyxwvutsrqponmlkjihgfedcbaZYXWVUT L- __ ~--~~~~--~~~ _.4 -.2 0.0 .2 .4 .6 .8 1.0 1.2 Fig. 6.62 Normalised stress paths in p' - q' space for test 5 1.2--------~----------------------------'zyxwvutsrqponmlkjihgfedcbaZYXWV 1.0 Solid symbol represents preshear condition .8 .6 'r:e '?r .4 .2 I -.2zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQP ,, ".8 1.0 1.2 -.4L---~--~--~----~--~~~----~--~ -.4 -.2 0.0 .2 .4 .6 I Fig. 6.63 Normalised stress paths in pi - q' space for test 6 293 1.2r---------,------------------------------,zyxwvutsrqponmlkjihg 1.0 Solid symbol represents presheor condition .8 .6 'cCzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA 'd=- .4 .2 -.2zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA -.4L-__ zyxwvutsrqponmlkjihgfedcbaZYXWVUTS -L zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQ L- __-L ~ __ ~ ~ __ ~~~ -. 4 -. 2 O. 0 •2 •4 •6 •8 1. 0 1. 2 Fig. 6.64 Normalised stress paths in pi - q' space for test 1 50 ~----------------------------------------------------__, ,..., ~ >J Ul Ul W 0:: 40 GI l- tn W > ...... IU 30 W U. U. W Z < w :::£ 20 Z ...... Z 0 ...... I- u ::J 0 W 0:: ID TOTAL STRAIN APPLIED eX) Fig. 6.65 Mean effective stress reduction - total strain applied relationship 294 ~-------------------------r------------------------~ .75zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA EXTERNAL .50 AX1AL ................ LOCAL AX J AL -------- LOCAL SHEAR '0:.-zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA 'if .25 -.25 L___ ~ -1.5 __ -L __ ~ -1.0 ~ __ ~ -.5 __ -L __ ~ STRAIN Fig. 6.66 ~ 0.0 .5 __ ~ __ ~~~ __ ~ 1.0 1.5 00 Normalised deviator stress versus strains during the application of strain paths in test 4 .75 LEGEND SAME AS ABOVE -.25 ~--~~~~~~--~--~--~--~--~--~--~--~ -1.00 -.75 -.50 -.25 0.00 .25 __ ~ __ ~ __ ~~ .50 .75 1.00 STRAINzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCB Cl.) Fig. 6.67 Nonnalised deviator stress versus strains during the application of strain paths in test 7 295 .750 ~----------------------------~------------------------------,zyxwvutsrqponmlkjihg .625 --- EXTERNAL AXIAL ................ LOCAL AX] AL .500 -------- , .. \ ," LOCAL SHEAR := : I .. I , : .'." I " ..... ,, .'.'.' , '0:.-zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA ,,' . , .' 'if· 375 zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA ~,,'.~. ." :.:::.:.::;:. . .250 .125 0.000zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONM L~ ~ __ ~~ ~ ~ __ ~ __----~---~ -.500 -.375 -.250 -. 125 0.000 STRAIN • 125 .250 .375 • SOD (;0 Fig. 6.68 Normalised deviator stress versus strains during the application of strain paths in test 8 .75 r--------------------,----------------------------------, SAME AS ABOVE LEGEND -------------., """"'1':..~::-, -,:,, : I : I : I .50 : i: : , I :, :, : : I I /~./ :I :, :, .', .', '0:.- ,- er .25 ...{/ /., ~-::.":.~..... ,,# • ..~" ...,~~...... ..;.0;" ,.<.~....' " •••• .a.:';v ,J",,,Jy.-.i> ,'.... .... ,.0 , ,<# 0° ;"'" ""," 0.00 ~--~--~--~'.~:-+---+--~--~---1~~--~~,~~~~--r_--~--r_--~__1 ,./---' ~~ ................ ",' rr' -.25 L-~~~ -1.5 __ _. __ ~ __ ~ __ ~ __ ~ __ -1.0 -.5 0.0 L_ __ ~~~~ .5 STRAIN __ 1.0 ~ __ 1.5 ~ __ ~ OD Fig. 6.69 Normalised deviator stress versus strains during the application of strain paths in test 3 296 __ 2.0 ~~ 2.5 .75 ~--------------------------r---------------------------'zyxwvutsrqp ---EXTERNAL AXIAL ................ LOCAL AX J AL .50 -------- LOCAL SHEAR .25zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA -.25 zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA L-__~ L- __ ~ L- __ ~ ~ __ ~ ~_~ __ ~_~_~ -1.5 -1.0 -.5 0.0 .5 1.0 1.5 STRAIN Fig. 6.70 .75 (7.) Normalised deviator stress versus strains during the application of strain paths in test 5 ~--------------r-------------------' LEGEND SAME AS ABOVE -.25 ~-~----~--~---~--~---~--~---._--~----~--_.--~ -1.5 -1.0 -.5 0.0 STRAIN Fig. 6.71 .5 1.0 (7.) Normalised deviator stress versus strains during the application of strain paths in test 6 297 1.5 10000 ~---------------------------r---------------------------' 8000zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA ,......zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA a Il... _r ...... 6000 Ul ::l ....J ::> Cl CJ ~nloading q~ ~ I- z 4000 < LJ Ls.J Ul ~ ~ ~oadmg 2000 o L-1.5 ~ -1.0 ~ -.5 ~------~------~~----~ 0.0 .5 1.0 l.S EXTERNAL AXIAL STRAINzyxwvutsrqponmlkjihgfedcbaZYXWVUTSR (X) (c) 15000 ~--------~-----------------r----------------------------. 12000 ,...... a Il... _r ...... 9000 Ul ::l ...J ::> ~nWadIDg Cl Cl ::::£ I- z 6000 < LJ Ls.J VI \LoadIn g 3000 o L- -. 75 ~~ -. 50 ~ ~ ~ -. 25 O.00 •25 EXTERNAL AXIAL STRAIN ~------~ .50 .75 (X) Cb) Fig. 6.72 Schematic diagram showing the variation of secant stiffnesses with strains during the application of strain paths: (a) Test 4 (b) Test 7 298 .750 .625 .500 .375 '0:.- ,~. 250 . 125 0.000 -. 125 -.250 -.75 -.50 -.25 .50 .25 0.00 .75 1.00zyxwvutsrqpo (in LOCAL RADIAL STRAINzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIH (0) .750 .625 .500 .375 '0:.- '~ .250 .125 0.000 -. 125 -.250 -.500 -.375 -.250 -. 125 0.000 • 125 .250 .375 LOCAL RADIAL STRAIN 00 (b) Fig. 6.73 Typical normalised deviator stress versus local radial strain plots during the application of strain paths: (a) Test 4 (b) Test 7 299 .500 1.5r--------------------r------------------~zyxwvutsrqponmlkjihgf 1.0zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA ,.... N "'-J • 5 z: ..... < 0:: IU) ~O.O~-+---r--+---~-+--~r-+---~-+---r--+--, < ..... >< < ~ L3 Cl _J -.5 -1.0 -1.5~~--~--~--~~--~--~--L-~--~--~~ -1.5 -1.0 -.5 0.0 .5 1.0 1.5 LOCAL RADIAL STRAINzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFED (Yo) (c) .50 ,.... N "'-J .25 z: ..... < 0:: IU) ~ 0.00 < ..... >< < ~ < u Cl -.25 _J -.50 LOCAL RAD1AL STRA1NzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQ (X) Cb) Fig. 6.74 Typical local axial strain versus local radial strain plots during the application of strain paths: (a) Test 4 (b) Test 7 300 .25.----------------------------,----------------------------, ----- EXTERNALzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA ............. ................ LOCAL . 20 - ................ · ]5 -, '0::" :J zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA <I · 10 ..•. • 05 .................................................... AXIAL STRAlN GO (a) .25,----------------------------,----------------------------, LEGEND I SAME AS ABOVE .20 ............ • 15 ........... ,... · 10 .zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA l····· zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA ........ • 05 /( ,) ........................................ ..... 0.00~----~------~----~~-----L------~----~-------L----~OO -1.00 -.75 -.50 -.25 0.00 .25 .50 .75 AXIAL Fig. 6.75 1.zyxwvut STRAINzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJ (I.) Cb) Typical normalised pore pressure change versus strains plots during the application of strain paths: (a) Test 4 (b) Test 7 301 90 r---------------------------------------------------------,zyxwvutsrqponmlk --- --- ---~~----------------------------------------- zyxwvutsrqponmlkjihgfedcbaZYXWV zyxwvutsrqponmlkjihgfedcbaZYXWVUTS zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJ , ...... ' ,.1 '" • • ::;.:c!!! :.!!~ !.» 80zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA '..................... , /'/ ...... o CL oX 70 I I '-' ", ", -~- en TEST 3 W 0::: tn 60 . "..".."."" ..zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIH tn l- -.....,.,==-. TEST 4 ;/// -------- 0::: TEST 7 Cl I- < ...... > w 50 Cl 40 30 ~-----~------~------~------~------~--------~----~ 3 2 o 5 4 EXTERNAL AXIAL STRAIN 6 7 (7.) Fig. 6.76 Typical deviator stress versus external axial strain plots during undrained shearing of "disturbed" specimens 90~----------------------------------------~ o TEST 4 • TEST 7 80 U) U) w A TEST 8 70 a::: I- U) a::: Cl I- -e 60 ..... > W Cl 50 40 0.01 o. 1 EXTERNAL AXIAL STRAINzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCB (Yo) Fig. 6.77 Typical deviator stress versus log. external axial strain plots during undrained shearing of "disturbed" specimens" 302 9000zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA EzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA o TEST 4 • TEST 7 A TEST 8 E EXTRAPOLATED 7500 ,...... 0 VALUES a... ~ '-' BODO U) U) LU Z u, u, ..... IU) 4500 I- z < U LU U) 3000 lSOOL- __ O. OJ ~~-L_J-L-L~~ -L __ L-J_~LL~ o. 1 1 EXTERNAL AXIAL STRAINzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA CX) Fig. 6.78 Typical secant stiffness versus log. external axial strain plots during undrained shearing of "disturbed" specimens 16000 ~------------------------------------------------------, o ,.., POINTS FOR SECANT STIFFNESS AT 0.0]% STRAIN • POINTS FOR SECANT STIFFNESS AT 0.05% STRAIN ~ POINTS FOR SECANT STIFFNESS AT O. lX STRAIN E EXTRAPOLATED VALUES 12000 c o, ...r .._, III III LLJ Z u, L&.. .... BODO I- U) tZ < E 0 u LLJ Ul 4000 • ~ o ~ __ ~ 90 ~ 100 ~ ~ ~~ 110 120 __ _' -L ~ 130 MEAN EFFECTIVE STRESS PRIOR TO SHEAR (kPa) Fig. 6.79 Secant stiffness - mean effective stress relationships 303 __~ 140 40 ~------------------------------------------------------, ----- TEST 35 TEST 4 -------TEST7 --- 30 TEST 8 ----- wzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA 25 ~ .. z /- L:) ..: :c u ui ,/-/- 20 oc :::J lJ1 lJ1 W oc ,~ / 15 0.. ~' --- ,------------- ----_ ..-,.---------zyxwvutsrqponmlkjihgfedcbaZ ~,~' W Cl:::: Cl CL ,- ,- ... > 10 zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA ,(:,.""'-" 5 .' . .zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA /: l,'.~.···········zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA loo' ~.' °O~------~------_L2------~3~------~4--------5~------~6------~7 EXTERNAL AXIAL STRAIN Fig. 6.80 (7.) Typical comparisons of pore pressure changes during undrained shearing of "disturbed" and "undisturbed" specimens 2.00 LEGEND 1. 75 SAME AS ABOVE 1. 50 w :::J _J -c 1. 25 ~----- :> I ..: lJ1 z ... ...- 1. DO /- ,/- Cl I0.. ~ w »> ,/ .75 :::cC lJ1 ./ ----------~/zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA --------- zyxwvutsrqponmlkjih _-_.zyxwvutsrqponmlkjihgfe ",/ ~ .50 ... \Ia I ",""" ~- .,..,,,,;' ,/ _------ --- --~.,..-- ------; ------- . .................................................................................................................. . 25 O.OOoL_------~-------2~------~3~------~4--------5~------~6~----~7 EXTERNAL AXIAL STRAINzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPO (X) Fig. 6.81 Typical comparisons of the variation of Skempton's A-values during undrained shearing of "disturbed" and "undisturbed" specimens 304 40zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGF TestIzyxwvutsrqponmlkjihgfedcbaZYXWVU ----Test 4 -------Test 7 ---Test 8 30 • Yield paint UJ ~ z \ -c I U UJ 0:: I I 20 ::J (J) (J) Isotropic elastic behaviour UJ ~ a.... UJ ~ CJ a.... 10 10 20 30 DEVIATOR STRESS CHANGE (kPo) Fig. 6.82 Typical comparisons of yielding behaviour during undrained shearing of "disturbed" and "undisturbed" specimens 305 40 CHAPTER 7 CONCLUSIONS AND RECOMMENDATIONS 7.1 CONCLUSIONS The research was carried out in three major phases, namely, (a) design and development of a Hall effect gauge for measuring large axial strains locally on triaxial specimens; (b) development of an approximate numerical technique to predict the strain paths of soil elements due to undrained penetration of a sampler; and; (c) performance of stress and strain path tests in order to model tube penetration disturbances on specimens of Ko-normally consolidated soft London Clay. In order to simulate tube penetration disturbances in the triaxial apparatus, a computer control program was also developed. The main findings and conclusions drawn from the various aspects of the research may be summarised as follows: - A Hall effect strain device has been developed for the measurement of large local axial strains on triaxial specimens. In this particular strain measuring device, the concept of using magnetically soft materials, known as pole pieces, at the end faces of the magnet, in order to obtain a high range of linearity, has been utilised. This new device is capable of measuring axial strains upzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQP to 10% over a 70 mm gauge length on the middle third of a 102 mm dia. x 203 mm high triaxial specimens. The gauge is, therefore, suitable for measuring strains on both clay and sand specimens. The gauge can resolve to approximately, 1 urn, which is equivalent to an axial strain of less than 0.002% on a 70 mm gauge length. The gauge is temperature and voltage stabilised, small and simple in design. Results from stress and strain path tests on Ko-normally consolidated specimens of London Clay, however, showed no significant difference between the measured local and external strains during loading in compression. During unloading, however, the specimens generally showed slightly stiffer response with respect to strains monitored by local axial gauges than that measured externally by LSCDT. As a result, determination of stress-strain parameters involving axial strains, e.g., Elt Bu, Eso and secant stiffnesses at various strain levels, were calculated using external axial strains. 306 Similar configuration of Hall effect sensor-magnet-pole piece system used in the axial gauge was also employed in the lateral strain caliper to measure radial strains locally at approximately the mid-height of the specimens. The lateral strain caliper can monitor strains better than 0.001 %. Experience gained with these Hall effect axial and lateral strain measuring devices has shown that they are simple and easy to use and sufficiently accurate. - An approximate numerical technique has been developed to predict the strain paths of soil elements due to undrained penetration of samplers having different cutting shoe designs. All the samplers have been modelled as piston samplers. In this numerical technique the soil has been treated as an incompressible. inviscid fluid flowing around the sampler under conditions of axial symmetry. The finite element technique has been adopted to model the complete flow' domain around the samplers. The application of the [mite element method has made simple the modelling of different cutting shoe designs of samplers. The principal findings from the numerical analyses are as follows: (a) Due to undrained penetration of typical samplers (e.g.• NOI. SOl and U100 samplers). the soil elements on the sample centreline are subjected to three distinct phases of triaxial shearing. namely. (i) an initial compression phase ahead of the sampler where axial strain increases from zero to a maximum value; (ii) an extension phase near the cutting edge of the samplers where the axial strain reverse from compression to extension and attain a maximum value in extension; and; (iii) a .second compressive phase inside the sampler tube where axial strain decreases and attain a constant value. These results agree with those reported by Baligh (1985) for Simple samplers (b) For the flat-ended samplers. however. only for the initial compression phase below the samplers do strains have' a peak. There are no peaks for the extension phases in the vicinity of the cutting edge of the samplers and the soil elements do not undergo a second compression phase inside the sampler tube. These results. however. contrast with those reported by Baligh (1985) for flat-ended samplers. (c) Soil disturbance. as described by the level of peak axial strains, varies across the diameter of the sampler tube. Soil elements located near the sample centreline suffer much less disturbance than those near and at the inside edge of the sampler tube. It was also observed that the relative increase in soil disturbance within the inner core (i.e.• up to 50% ofzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA R, from centreline of the sampler) of the sample was 307 not significant as compared with that within the outer core (i.e., 50% ofzyxwvutsrqponmlkjihgfedc R, from centreline of sampler to the inside edge of the sampler tube). Practically, this implies that the soil located in the outer half of the tube sample should be avoided in the preparation of representative specimens for laboratory testing. Baligh et al (1987) also reported that soil disturbance, as described by the level of shear distortions, due to a Simple sampler (BIt = 40, ICR = 1%) decreases towards the centre of the sample; soil disturbances in the outer half of the sample involve significant nonuniformities. (d) The peak axial strain in compression and extension are not equal. The relative values have been found to depend on thickness of the sampler tube and design features of the cutting shoe. At or near the centreline of the sampler, the peak axial strain in compression during the initial compressive phase is basically controlled by the thickness of the sampler tube. The peak extensive strains in the vicinity of the cutting shoe, however, are governed by the precise geometry of the cutting shoe, especially the inside clearance ratio. For the NOI sampler, it was found that at and near the centreline of the samplers, the peak compressive strains are usually higher than the peak extensive strains. This indicates that at these locations, the effect of thickness predominates over the effect of actual cutting shoe design. At or near the cutting edge, however, the peak extension strains are much higher than the peak compression strains, thereby indicating the profound effect of cutting shoe geometry over the thickness of the samplers. In case of the SOl sampler, for all the strain paths, the peak axial strain in compression is greater than that in extension. This shows that the effect of thickness on strain paths is more pronounced than the effect of cutting shoe geometry. For the Ul00 samplers, however, peak extensive strains at all locations within the sampler tube are higher than the peak compressive strains. (5.875 mm and 5.9 mm) of these This demonstrates that despite high thicknesseszyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDC samplers, the influence of the cutting shoe designs on strain history is more significant than the thickness of the samplers. Comparing the NOI, SOl and UlOO samplers, it was found that, at the inside edge of the sampler tube, the peak axial strain in extension was highest for the UlOO (type I) sampler and least for the SOl sampler; at the centreline, it was highest for the UlOO (type II) sampler and least for the NOI sampler. At both the centreline and inside edge of the sampler tube, the peak compressive strains are, however, highest for the UlOO (type II) sampler and least for the NOI sampler. At the centreline of the Simple samplers, Baligh et al (1987) reported identical peak axial strains in compression and extension which 308 obviously contrasts with the findings for all the samplers investigated in this research. - A detailed investigation of the effects of various design parameters of the sampler, e.g., area ratio, inside clearance ratio, and inside and outside cutting edge taper angles, showed that: (i) Increasing area ratio by increasing the thickness of the sampler tube causes significant increase in the peak compressive strains. The peak axial strains in extension, however, are increased only slightly. For a sampler with an area ratio equal to 100.46%, the peak axial strains in compression and extension are respectively 410% to 485% and 23% to 37% higher than those for the sampler of area ratio 10.14%. (ii) Increasing the inside clearance ratio by increasing the inside diameter of the sampler tube causes a significant increase in peak axial extension strains and a small reduction in peak axial compression strains (during the first compression phase below the sampler). At the centreline, the peak axial strain in extension for the sampler = 3.96% was 9.7 times higher than for a sampler with ICR = 0.495%. with ICRzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA The peak axial compressive strain during the initial compression phase was, however, 77% less for the sampler with ICR = 4.96% than for the sampler with ICR = 0.495%. (iii) A change in the inside cutting edge taper angle had no apparent influence on the initial compression phase of strain paths as indicated by identical peak axial compressive strains. Changes in the inside cutting edge taper angle, however, affected the extension phase in the vicinity of the cutting edge. Increasing inside cutting edge taper angle reduces the peak axial strain in extension. By increasing inside cutting edge taper angle from 0.358° to 1.432° (i.e., a four-fold increase), the peak axial strain in extension was reduced by 62%. (iv) Outside cutting edge taper angle has a profound influence on soil straining. Peak axial strains in both compression and extension increase with the increase in outside cutting edge taper angle. The effect of outside cutting edge taper angle, however, has been found to be more significant on the peak axial compressive strain. When the outside cutting edge taper angle increased from S° to 19.2go, the peak axial compressive strain was found to increase by 360% to 410% (depending on the position of the soil within the sample). The peak axial strain in extension increased 309 by only 18% to 21 %. From the parametric study of the cutting shoe designs, it has been concluded that in order to restrict the degree of disturbance (peak axial strains in compression and extension) to less than 1%, a sampler should have the following values of the design parameters: (i) Area ratio not more than 10% (ii) Inside clearance ratio not more than 0.5% (iii) Inside cutting edge taper angle 1 to 1.5° (iv) Outside cutting edge taper angle not more than 5° It is, however, important to note that for an identical disturbance applied to unaged Ko-normally consolidated London Clay, significant reduction in mean effective stress,zyxwvutsrqponmlkjihgfedcb p'O, initial stiffnesses and pore pressure change was found. "undisturbed" specimen, p'O' El' E5Q, 0 .., {E..)ODl.lP/O and ~ Compared with an were reduced by approximately 26%, 77%, 65%, 80%. 78% and 56% respectively. although undrained strength was reduced by only about 6%. This essentially implies that it is extremely difficult to design a sampler which will sample undisturbed samples from normally consolidated unaged soil deposits. For good quality sampling in soft aged clays, the above design parameters of a tube sampler may be appropriate. - Four flat-ended samplers of varying thickness and BIt ratio were also investigated. It was found that both the peak axial strain in compression and maximum axial strain in extension were dependent on the BIt ratio of the samplers rather than the thickness of the samplers. An increase in BIt ratio reduced the level of straining in the soil. For a flat-ended sampler with BIt = 12.2, the peak axial strain in compression at the = 45.6. The centreline was about 4.2 times higher than that for the sampler with BItzyxwvutsrqponmlkjihgfedcb maximum axial strain in extension was also considerably higher (about 4 times) for the sampler of BIt = 12.2 than that for the sampler with BIt = 45.6. Comparing samplers of different designs but with identical thickness and BIt ratios (e.g., NOI, SOl and Ul00 samplers, and flat-ended samplers), it was concluded that the design of cutting shoe influences considerably the level of disturbance and that it affects the strain histories of soil elements not only at or near the inside edge of the sampler tube but also near the centreline of the sampler. 310 These conclusions, however, contrast with those reported by Baligh (1985). Baligh (1985), from his analyses on the Simple sampler and flat-ended sampler, concluded that sample disturbance (measured by the level of shear distortions) depends only on thezyxwvutsrqponmlkjihgfedcbaZYXWVUTSR BIt ratio of the sampler and that sampler geometry has no significant effect on the strain history on the centreline. The effect of cutting shoe geometry on soil distortions was only visible in the vicinity of the sampler walls. - Stress and Strain path tests were carried out using a computer controlled stress path system incorporating the Hall effect local strain measuring devices and local pore pressure transducer. The microcomputer- controlled system made it possible to apply precisely a combination of stresses with very small steps at specified rates. This resulted in stress paths followed as closely as possible. - The "undisturbed" behaviour of the Ko-normally consolidated London Clay was found to be markedly different in compression and extension. The effective angle of internal friction, cj>' was significantly higher in extension than in compression. The undrained strength, however was approximately the same. The clay also showed profound stiffness anisotropy. El' Eso. GzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJI u, secant stiffnesses at various strain levels, and the normalised stiffness parameter, <Eu)O.oI..,!P'O, were all considerably higher in extension than in compression. Pore pressure changes, however, were significantly smaller in extension than in compression. - Strain path tests (modelling tube penetration disturbances) on Ko-normally consolidated unaged London Clay show that the tube penetration disturbances have significant effects on the subsequent undrained stress paths, stress-strain, stiffness and pore pressure characteristics. Results of simulated tube penetration disturbances indicate the following effects in unaged normally consolidated soft London Clay: (i) The undrained stress paths were completely different for the "disturbed" specimens. The stress paths of the "disturbed" specimens were similar to those of a lightly overconsolidated specimen, giving approximately vertical effective stress paths during shearing in compression up to failure. after the application of tube penetration disturbances. (ii) Strength parameters, tu and cj>'. were not modified significantly as a result of the application of tube penetration disturbances. A small increase (1% to 6%) in cj>' and 311 a small reduction (up to 6.7%) in CuzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGF were observed. For plastic Drammen Clay (PI = 27), Lacasse and Berre (1988) also reported the same shear strength for the disturbed and undisturbed normally consolidated and overconsolidated specimens. The disturbed specimen was strained to a magnitude identical to that predicted by (BIt = 40, ICR ... 1%). Strain Path Method at the centreline of a Simple samplerzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLK For similar disturbances applied to less plastic Boston Blue Clay (PI = 20 ± 2.5), Baligh et al (1987), however, reported a 21% reduction in undrained strength ratio. (iii) A significant reduction in the mean effective stress occurs because of tube penetration disturbances. For the application of the predicted tube penetration disturbances at the centreline of NOI 54 mm dia. piston sampler (BIt = 45.6), the decrease in mean effective stress was about 10%; while the reduction in mean effective stress due to simulated tube penetration disturbances at the centreline of a thin walled Simple sampler with BIt = 40 and ICR ... 1% was approximately 26%. Baligh et al (1987), however, found an appreciable reduction (about 59%) in mean effective stress for Boston Blue Clay when subjected to disturbances predicted at the centreline of a Simple sampler. (iv) Axial strains at peak strength, ~ were considerably increased (26% to 313%). Baligh et al (1987) also reported considerable increase(about 27 times) in ~ due to tube penetration disturbances. (v) Initial tangent modulus, El, undrained shear modulus, G, and secant modulus at half the maximum deviator stress, Eso, were all reduced significantly depending upon the degree of disturbance applied. For example, the imposed tube penetration disturbances predicted at the centreline of the Simple sampler reduced the values of El, G, and Eso by 77%, 80% and 65% respectively relative to those for the "undisturbed" specimen. For an equivalent disturbance, Baligh et al (1987) found that the undrained modulus ratio (Eyja'yJ decreased as much as 95% for the Boston Blue Clay. Lacasse and Berre (1988) also found significant reduction in initial moduli for both normally consolidated and overconsolidated specimens of Drammen Clay sheared in compression after the application of disturbance. (vi) Significant reduction in secant stiffnesses at various strain levels occurred because of the applied tube penetration disturbances. The normalised stiffness index, (Eu)o'ot.lP'o was reduced, thereby indicating a reduction in the size of the small strain region. (vii) Considerable changes in pore pressure responses have been noted. Skempton's pore pressure parameter A at peak strength, i.e., ~ was 41% to 85% less for the 312 "disturbed" specimens than for the "undisturbed" specimen. For the "undisturbed" specimen a sharp increase in pore pressure during the early stages of loading was observed, indicating probable yielding behaviour. Yielding behaviour was noticed for less "disturbed" specimens (i.e., in tests 7 and 8). For more "disturbed" specimens, however, no such trend was observed. Because of disturbance, the pore pressure response early in the test corresponds approximately to elastic behaviour (Auzyxwvutsrqponmlkjih = 1/3 Aq). In the "undisturbed" specimen, however, the pore pressure during the early stages of the test was found to be much higher than that for isotropic elastic soil. From the aforementioned effects of tube penetration disturbances in unaged soft London Clay, it is evident that except for the strength parameters, Cu,4>', mean effective stresses, initial stiffnesses and pore pressure changes are reduced considerably because of tube sampling disturbances. This clearly indicates that it is virtually impossible to collect good quality undisturbed samples of unaged normally consolidated clays using thin-walled tubes. Although the effect of reconsolidating "disturbed" specimens in order to recover the "undisturbed" behaviour for these unaged specimens was not investigated, this suggeststhe need to minimise sampling disturbance (i.e., tube penetration and "perfect" sampling disturbances) by reconsolidating before undrained shearing using the Bjerrum or the SHANSEP procedures. Baligh et al (1987) reported that specimens of unaged Boston Blue Clay, consolidated to 1.5 and 2.0 times the maximum past effective vertical stress in accordance with the SHANSEP procedure, exhibited virtually the same normalised behaviour of the "undisturbed" specimens. For aged samples, Burland (1990), however, reported a lower undrained strength ratio than for undisturbed specimen. This was attributed to the effect of destructuration when reconsolidating according to SHANSEP procedure. It is also evident from the comparisons of the effects of tube penetration disturbances in unaged London Clay (PI = 45) and Boston Blue Clay (PI = 20 ± 2.5) that less plastic Boston Blue Clay suffers much more reduction in strength and stiffness compared with more plastic London Clay for an approximately equivalent degree of disturbances. This inevitably implies that the degree of disturbance depends not only on the design of a sampler but also the type of clay sampled. The severity. of the effects of tube penetration disturbances is much more acute in less plastic and sensitive clays than in more plastic and insensitive clays. 313 7.2 RECOMMENDATIONS FOR FURTHER STUDY Several aspects of the work presented in this thesis require further study. Some of the important areas of further research may be listed as follows: (1) The numerical technique which has been developed to predict the strain paths of soil elements due to undrained penetration of a sampler is based on relatively simple assumptions. The soil has been treated as incompressible, shearing resistance. inviscid fluid offering no The frictional drag at the soil-sampler interface was neglected. It would be interesting to incorporate the properties of real soils into the model by modifying the boundary conditions and then predicting the soil disturbance. This would clarify and compare the possible changes in predictive behaviour between a real soil and an imaginary soil. In this research, cutting shoe designs of NOI, SOl and Ul00 samplers have been (2)zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA investigated. Further study could be carried out to model other standard samplers used in practice, e.g., NOI 95 mm dia. sampler, Osterberg 127 mm dia. hydraulic piston sampler, 124 mm dia. research sampler developed at SOl. The results could then be compared with those obtained from the present investigation. (3) Further parametric study could be carried out to assess the effect of outside clearance, length of the sampler, L, external diameter of the sampler tube, B, and ratio on the predicted soil disturbance. LIB In this research, only the strain paths of soil elements entering the sampler tube have been studied. It may be important to know the behaviour of the soil around and outside the sampler. (4) The automated triaxial equipment used for carrying out all the tests is a stresscontrolled system. by allowing In the strain path tests, strains were imposed on the specimens them to follow pre-specified stress paths. Also during undrained shearing, the specimens were loaded at a constant rate of deviator stress change. More modifications extra features, to the system and software could be made in order to include such as to incorporate the provisions of strain-controlled testing, especially when applying the undrained tube penetration disturbances and undrained shearing in compression or extension. 314 (5) The scope of the testing programme has been limited to investigating only the tube penetration disturbance effects on reconstituted Ko-normally consolidated unaged soft London Clay. The fabric of natural soils may have a significant influence on the behaviour of soils and hence, further research is required on natural soils to identify any special features associated with fabric, composition, bonding and ageing. (6) To observe and identify the important effects of stress history on soil disturbance, it is perhaps desirable to extend the investigation to overconsolidated samples having a wide range of overconsolidation ratios. 315 REFERENCES ADACHI, K., TODO, H. and MIZUNO, H. (1981) "Quality of Samples of Soft Cohesive Soil", Proc., 10th Int. Conf. Soil Mech. and Found. Engng., Stockholm, Vol. 2, pp. 409-412. ADAMS, J.I. AND RADAKRISHNA, H.S. (1971) "Loss of Strength Due to Sampling in a Glacial Lake Deposit", Symposium on sampling of Soil and Rock, ASTM STP 483, pp. 109-120. ALONSO, E.E., ONATE, E. and CASANOVAS, J.S. (1981) "An Investigation into Sampling Disturbance", Proc., 10th Int. Conf. Soil Mech. and Found. Engng., Stockholm, Vol. 2, pp. 419-422. ALVA-HURTADO, J.E. and SELIG, E.T. (1981) "Survey of Laboratory Devices for Measuring Soil Volume Change", Geotechnical Testing Journal, ASTM, Vol. 4. No.zyxwvutsrqpon 1, pp. 11-18. APTED, J.P. (1977) "Effects of Weathering on Some Geotechnical Properties of London Clay", Ph.D. Thesis, Imperial College, University of London. ARMAN, A. and McMANIS, KL. (1976) "Effects of Storage and Extrusion on Sample Properties", Symposium on Soil Specimen Preparation for Laboratory Testing, ASTM STP 599, pp. 66-87. ATKINSON, J.H. (1973) "The Deformation of Undisturbed London Clay", Ph.D. Thesis, Imperial College, University of London. ATKINSON, J.H. and KUBBA, L.M. (1981) "Some Effects of Sample Disturbance on Soft Clay", Proc., 10th Int. Conf. Soil Mech. and Found. Engng., Stockholm Vol. 2, pp. 423-426. ATKINSON, J.H. (1985) "Simple and Inexpensive Pressure Control Equipment for Conventional and Stress Path Triaxial Testing of Soils", Geotechnique, Vol. 35, No. I, pp. 61-63. 316 ATKINSON, IH. and EVANS, J.S. (1985) Discussion on "The Measurement of Soil Stiffness in the Triaxial Apparatus", Geotechnique, Vol. 35, No.3, pp. 378-380. ATKINSON, J.H., EVANS, J.S. and HO, E.W.L. (1985) "Non-uniformity of Triaxial Samples Due to Consolidation With Radial Drainage", Geotechnique, Vol. 36, No. 4, pp. 611-614. ATKINSON, J.H., EVANS, J.S. and Scott, C.R. (1985) "Developments in Microcomputer Controlled Stress Path Testing Equipment for Measurement of Soil Parameters", Ground Engineering, Vol. 18, No.1, pp. 15-22. ATKINSON, J.H., RICHARDSON, D. and ROBINSON,zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONM P.I. (1987) "Compression and Extension ofzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA K, Normally Consolidated Kaolin Clay", Journal of the Geotech. Engng. Div., ASCE, Vol. 113, No. GTI2, pp. 1468-1481 BALDI, G., HIGHT, D.W. and mOMAS, G.E. (1988) "A Reevaluation of Conventional Triaxial Test Methods", Symposium on Advanced Triaxial Testing of Soil and Rock, ASTM STP 977, pp. 219-263. BALIGH, M.M. (1975) "Theory of Deep Site Static Cone Penetration Resistance", Research Report No. R75-76, Order No. 517, Dept. of Civil Engng., MIT, Cambridge, Massachusetts, 133 pages. BALIGH, M.M. (1985) "Strain Path Method", Journal of the Geotech. Engng. Div., ASCE, Vol. 111, No. GT9, pp. 1108-1136. BALIGH, M.M., AZZOUZ, A.S. and CHIN, C.T. (1987) "Disturbance Due to Ideal Tube Sampling Disturbance", Journal of the Geotech. Engng. Div., ASCE, Vol. 113, No. GTI, pp. 739-757. BALLA, A. (1960) "Stress Conditions in Triaxial Compression", Journal of the Soil Mech. and Found. Div., ASCE. Vol. 86, No. SM6, pp. 57-84. BARDEN. L. and McDERMOTT. R.J.W. (1965) "Use of Free Ends in Triaxial Testing of Clays", Journal of the Soil Mech. and Found. Div.• ASCE, Vol. 91. No. 317 SM6, pp. 1-23. BEEN, K. and SILLS, G.C. (1981) "Self Weight Consolidation of Soft Soils: An Experimental and Theoretical Study", Geotechnique, Vol. 31, No.4, pp. 519-535. BEGEMANN, H.K.S.Ph. (1977) "Sample Disturbance Influencing Shear Strength of Cohesive Soils", Specialty Session No.2 on Soil Sampling, 9th Int. Conf. Soil Mech. and Found. Engng., Tokyo, pp. 43-45. BERRE, T., SCHJETNE, K. and SOLLIE, S. (1969) "Sampling Disturbance of Soft Marine Clays", Proc., Specialty Session No.1 on Soil Sampling, 7th Int. Conf. Soil Mech. and Found. Engng., Mexico, pp. 21-24. BERRE, T. and BJERRUM,zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFED L. (1973) "Shear Strength of Normally Consolidated Clays", Proc., 8th lot. Conf. Soil Mech. and Found. Engng., Moscow, Vol. 1, pp. 3949 BERRY, P.L. and WILKINSON, W.B. (1969) " Radial Consolidation of Clay Soils", Geotechnique, Vol. 19, No.2, pp. 253-284. BISHOP, A.W., BLIGHT, G.E. and DONALD, I.B. (1960) Discussions and Closures to Session 2, Proc., Research Conf. on Shear Strength of Cohesive Soils, ASCE, University of Colorado, Boulder, pp. 1027-1042. BISHOP, A.W. and DONALD, lB. (1961) "The Experimental Study of Partly Saturated Soil in the Triaxial Apparatus", Proc., 5th Int. Conf. Soil Mech. and Found. Engng., Paris, Vol. 1, pp. 13-21. BISHOP, A.W. and HENKEL, D.1. (1962) "The Measurement of Soil Properties in the Triaxial Test", 2nd Edition, Edward Arnold, London. BISHOP, A.W. and GmSON, R.E. (1963) "The Influence of the Provisions for Boundary Drainage on Strength and Consolidation Characteristics of Soils Measured in the Triaxial Apparatus", Symposium on Laboratory Shear Testing of Soils, ASTM STP 361, pp. 435-451. 318 BISHOP, A.W. and WESLEY, L.D. (1975) "A Hydraulic Controlled Stress Path Testing", Geotechnique, Triaxial Apparatus Vol. 25, No.4, for pp. 657-670. BJERRUM,zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA L. (1973) "Problems of Soil Mechanics and Construction on Soft Clays and Structurally Unstable Soils (Collapsible, Expansive and Others)", State-of-the- Art Report, Session IV, Proc., 8th Int. Conf. Soil Mech. and Found. Engng., Moscow, Vol. 3, pp. 109-159. BLACK, D.K. and KENNETH, L.L. (1973) "Saturating Laboratory Samples by Back Pressure", Journal of the Soil Mech. and Found. Div., ASCE, Vol. 99, No. SM1, pp~ 75-92 BLIGHT, G.E. (1963) "The Effect of Nonuniform Measurements Pore Pressures of the Shear Strength of Soils", Symposium on Laboratory on Laboratory Shear Testing of Soils, ASTM STP 361, pp. 173-184. BLIGHT, G.E. (1965) "Shear Stress and Pore Pressure in Triaxial Testing", Journal of the Soil Mech. and Found. Div., ASCE, Vol. 91, No. SM6, pp. 25-40. BOZOZUK, M. (1971) "Effect of Sampling, Size, and Storage on Test Results for Marine Clay", Symposium on Sampling of Soil and Rock, ASTM STP 483, pp. 121131. BRAND, E.W. (1975) "Back Pressure Effects Characteristics of Soft Clay", Soils and Foundations, BROMHAM, S.B. (1971) "The Measurement on the Undrained Vol. IS, No.2, of Disturbance Strength pp. 1-16. in Samples of Soft Clay", Proc., Specialty Session on Quality in Soil Sampling, 4th Asian Conf., Int. Soc. Soil Mech. and Found. Engng., Bangkok, pp. 68-72. BROMS, B.B. (1980) "Soil Sampling in Europe: State-of-the-Art", Journal of the Geotech. Engng. Div., ASCE, Vol. 106, No. OTt, pp. 65-98. BROWN, S.F. and SNAITH, M.S. (1974) "The Measurement of Recoverable Irrecoverable Deformation in the Repeated Load Triaxial Test", Geotechnique, 319 and Vol. 24, No.2, pp. 255-259. BROWN, S.F., AUSTIN, G. and OVERY, R.F. (1980) "An Instrumented Triaxial Cell for Cyclic Loading of Clays", Geotechnical Testing Journal, ASTM, Vol. 3, No. 4, pp. 145-152. BS 1377 (1975) "Methods of Tests for Soils for Civil Engineering Purposes", British Standards Institution, London. BURLAND, lB. and SYMES, M. (1982) "A Simple Axial Displacement Gauge for Use in the Triaxial Apparatus", Geotechnique, Vol. 32, No.1, pp. 62-65. BURLAND, J.B. (1990) "On the Compressibility and Shear Strength of Natural Clays", Thirthieth Rankine Lecture, to be published in Geotechnique. BURMISTER, D.M. (1936) "A Method of Determining the Representative Character of Undisturbed Samples and Something of the Disturbance Caused by Sampling Operation", Proc., 1st Int. Conf. Soil Mech. and Found. Engng., Harvard, Vol. 3, pp. 26-28. CALHOON, M.L. (1956) "Effect of Sample Disturbance on the Strength of a Clay", Transactions, ASCE, Vol. 121, Paper No. 2827, pp. 925-954. CHIN, C.T. and BALIGH, M.M. (1983) "Deformations and Strains Due to OpenEnded Pile Installation in Saturated Clays", Research Report No. R83-17, Order No. 757, Dept. of Civil Engng., MIT, Cambridge, Massachusetts. CHIN, C.T. (1986) "Open-Ended Pile Penetration in Saturated Clays", Ph.D. Thesis, MIT, Cambridge, Massachusetts. CLAYTON, C.R.!., SIMONS, N.E. and MATTHEWS, M.C. (1982) "Site Investigation", Granada Publishing Limited, London. CLAYTON, C.R.!. (1986) "Sampling Disturbance andzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPO BS 5930", Site Investigation Practice: Assessing BS 5930, Edited by A.W. Hawkins, Geological Society, Engng. 320 Geology Special Publication No.2, pp. 33-40. CLAYTON, C.R.I. and KHATRUSH, S.A. (1986) "A New Device for Measuring Local Axial Strains on Triaxial Specimens", Geotechnique, Vol. 36, No.4, pp. 593597. CLAYTON, C.R.I. and KHATRUSH, S.A. (1987) Discussion on "A New Device for Measuring Local Axial Strains on Triaxial Specimens", Geotechnique, Vol. 37, No. 3, pp. 415-417. CLAYTON, C.R.I. and KHATRUSH, S.A. (1988) "The Use of an Automated Triaxial System With Local Axial and Radial Strain Measurement to Investigate the Yielding of a Sand", Conf. on Novos Conceitos em Ensaios de Campo et de Loaborotorio em Geotecnia, Rio de Janeiro, Vol. 1, pp. 57-70. CLAYTON, c.R.I., KHATRUSH, S.A., BICA, A.V.D. and SIDDIQUE, A. (1989) "The Use of Hall Effect Semiconductors in Geotechnical Instrumentation", Geotechnical Testing Journal, ASTM, Vol. 12, No.1, pp. 69-76. CONLON, R.I. and ISAACS, R.M.F. (1971) "Effect of Sampling and Testing Techniques on the Shear Strength of a Glacial-Lacustrine Clay from WeIland, Ontario", Symposium on Sampling of Soil and Rock, ASTM STP 483, pp. 10-29. COSTA FILHO, L. de M. (1980) "A Laboratory Investigation of the Small Strain Behaviour of London Clay", Ph.D. Thesis, Imperial College, University of London. COSTA FILHO, L. de M. (1985) "Measurement of Axial Strains in Triaxial Tests on London Clay", Geotechnical Testing Journal, Vol. 8, No.1, pp. 3-13. DARAMOLA, C. (1978) "The Influence of Stress History on the Deformation of a Sand", Ph.D. Thesis, Imperial College, University of London. DARLEY, P. (1973) Discussion on "Apparatus for Measuring Volume Change Suitable for Automatic Logging", Geotechnique, Vol. 23, No.1, pp. 140-141. 321 DAVIS, E.H. and POULOS, H.G. (1967) "Laboratory Investigations of the Effects of Sampling", Transactions of the Institution of Engineers, Australia, CE9, No. I, pp. 86-94. DIETZLER, D.P., MOOSE, D.A. and SCHUH, J.C. (1988) "Effects of Sampling Disturbance on Shear Strength of Glacial Till and Compacted Fill", Symposium on Advanced Triaxial Testing of Soil and Rock, ASTM STP 977, pp. 628-641. EDEN, W.J. (1971) " Sampler Trials in Overconsolidated Sensitive Clay", Symposium on Sampling of Soil and Rock, ASTM STP 483, pp. 132-142. GENS, A. (1982) "Stress-Strain and Strength Characteristics of a Low Plasticity Clay", Ph.D. Thesis, Imperial College, University of London. GRAHAM, J., KWOK, C.K. and AMBROSIE, R.W. (1987) "Stress Release, Undrained Storage, and Reconsolidation in Simulated Underwater Clay", Canadian Geotechnical Journal, Vol. 24, No.2, pp. 279-288. GRAHAM, J. and LAU, S.L.-K. (1988) "Influence of Stress Release Disturbance, Storage, and Reconsolidation Procedures on the Shear Behaviour of Reconstituted Underwater Clay", Geotechnique, Vol. 38, No.2, pp. 279-300. GERMAINE, J.T. and LADD, C.C. (1988) "Triaxial Testing of Saturated Cohesive Soils", Symposium on Advanced Triaxial Testing of Soil and Rock, ASTM STP 977, pp. 421-459. HENKEL, D.l and GILBERT, G.D. (1952) "The Effect of the Rubber Membrane on the Measured Triaxial Compression Strength of Clay Samples", Geotechnique, Vol. 3, pp. 20-29. HIGHT, D.W. (1982) "A Simple Piezometer Probe for the Routine Measurement of Pore Pressure in Triaxial Tests on Saturated Soils", Geotechnique, Vol. 32, No.4, pp. 396-401. HIGHT, D.W. (1983) "Laboratory Investigations of Sea Bed Clays", Ph.D. Thesis, 322 Imperial College, University of London. HIGHT, D.W., GENS,zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA A. and JARDINE, R.J. (1985) Discussion on "The Reaction of Clays to Sampling Stress Relief', Geotechnique, Vol. 35, No. I, pp. 86-88. HIGHT, D.W., GENS, A. and JARDINE, R.J. (1985) "Evaluation of Geotechnical Parameters from Triaxial Tests on Offshore Clay", Proc., Int. Conf. on Offshore Site Investigation, Society for Underwater Technology, London, England, pp. 253-268. HIGHT, D.W. (1986) "Laboratory Testing: Assessing BS 5930", Site Investigation Practice: Assessing BS 5930, Edited by A.W. Hawkins, Geological Society, Engng. Geology Special Publication No.2, pp. 44-52. HIGHT, D.W., JARDINE, R.J. and GENS, A. (1987) "The Behaviour of Soft Clays", Chapter 2 in Embankments on Soft Clays, Special Publication, Bulletin of the Public Works Research Center, Athens, pp. 33-158. HIGHT, D.W. and BURLAND, J.B. (1990) "Review of Soil Sampling and Laboratory Testing for the Science and Engineering Research Council", Summary Report, Science and Engineering Research Council, England. HINTON, E., OWEN, D.R. (1979) "An Introduction to Finite Element Computations", Pineridge Press Limited, Swansea, U.K. HO, E.W.L. (1985) "Undrained Compression and Extension Tests on Reconstituted Speswhite Kaolin Consolidated under K, Conditions with a Particular Reference to the Effect of Perfect Sampling", Geotechnical Engng. Research Centre, Research Report GE/85/17, City University, London, England. HODGSON, J.D. (1976) Discussion on "A Device for Measuring Volume Change", Geotechnique, Vol. 26, No.1, pp. 243-244. HOPPER, R.J. (1988) "Continuous Consolidation Testing of London Clay", B.Sc. Final Year Project Report, University of Surrey, England. 323 HOLM, G. and HOLTZ, R.D. (1977) "A Study of Large Diameter Piston Samplers", Proc., Specialty Session No. 2 on Soil Sampling, 9th- Int. Conf. Soil Mech. and Found. Engng .• Tokyo, pp. 73-78. HVORSLEV, M.J. (1940) "Preliminary Draft Report on the Present Status-of-the-zyxwvutsrqponm Art of Obtaining Undisturbed Samples of Soils", Supplement to Proc., Purdue Conf. on Soil Mechanics and its Applications, Purdue University, Lafayette. Indiana, U.S.A. HVORSLEV, MJ. (1949) "Subsurface Exploration and Sampling of Soils for Civil Engineering Purposes", Waterways Experimental Station, Vicksburg, U.S.A. International Society for Soil Mechanics and Foundation Engineering (1965) "Report of the Subcommittee on Problems and Practices of Soil Sampling", Proc., 6th Int. Conf. Soil Mech. and Found. Engng., Montreal, Vol. 3, Appendix II, pp. 64-71. IRWIN, M.J. (1972) Discussion on "Use of Servo Mechanisms for Volume Change Measurement and K, Consolidation", JAKOBSON, B. (1954) "Influence of Sampler Type and Testing Method on Shear Geotechnique, Vol. 22, No. I. pp. 186-187. Strength of Clay Samples", Proc., Royal Swedish Geotechnical Institute, No.8, pp. 1-58. JARDINE, RJ., SYMES, M.l and BURLAND, lB. (1984) "The Measurement Soil Stiffness in the Triaxial Apparatus", Geotechnique, Vol. 34, No.3, of pp. 323- 340. JARDINE, R.J. (1985) "Investigations of Pile-Soil Behaviour With Special Reference to the Foundations of Offshore Structures", Ph.D. Thesis, Imperial College, University of London. KALLSTENIUS, T. (1958) "Mechanical Disturbances in Clay Samples Taken With Piston Samplers", Proc., Royal Swedish Geotechnical Institute, No. 16, pp. 1-75. KALLSTENIUS, T. (1961) "A Standard Piston Sampler Prototype", Swedish Geotechnical Institute, No. 19, pp. 38-44. 324 Proc., Royal KALLSTENIUS, T. (1963) "Studies on Clay Samples Taken With Standard Piston Sampler" Proc., Royal Swedish Geotechnical Institute, No. 21, pp. 1-207. KALLSTENIUS, T. (1971) "Secondary Mechanical Disturbance Effects in Cohesive Soil Samples", Proc., Specialty Session on Quality in Soil Sampling, 4th Asian Conf. Int. Soc. Soil Mech. and Found. Engng., Bangkok, pp. 30-39. KHATRUSH, S.A. (1987) "The Yielding of a Fine Sand in Triaxial Stress Space", Ph.D. Thesis, University of Surrey, England. KIMBALL, W.P. (1936) "Settlement Records of the Mississippii River Bridge at New Orleans", Proc., 1st Int. Conf. Soil Mech. and Found. Engng, Vol. I, pp. 85-92. KIMURA, T. and SAITOH, K. (1982) "The Influence of Disturbance Due to Sample Preparation on the Undrained Strength of Saturated Cohesive Soil", Soils and Foundations, Vol. 22, No.4, pp. 109-120. KIRKPATRICK, W.M. and KHAN, AJ. (1984) "The Reaction of Clays to Sampling Stress Relief" , Geotechnique, Vol. 34, No.1, pp. 29-42. KIRKPATRICK, W.M., KHAN, A.J. and MIRZA,zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONM A.A. (1986) "The Effects of Stress Relief on Some Overconsolidated Clays", Geotechnique, Vol. 34, No.4, pp. 511-525. KJELLMAN, W., KALLSTENIUS, T. and WAGER, O. (1950) "Soil Sampler With Metal Foils", Proc., Royal Swedish Geotechnical Institute, No. I, pp. 7-75. I. (1974) "Lever-type Apparatus for Electrically Measuring Volume KLEMENTEV,zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA Change", Geotechnique, Vol. 24, No.4, pp. 670-671. KOUTSOFTAS, D.C. (1981) "Undrained Shear Behaviour of a Marine Clay", Symposium on Laboratory Shear Strength of Soil, ASTM STP 740, pp. 254-276. KUBBA, L.M. (1981) "The Effect of sampling Disturbance on the Deformation of Clay", Ph.D. Thesis, University College, Cardiff, University of Wales, U.K. 325 LACASSE, S., BERRE, T. and LEFEBVRE, O. (1985) "Block Sampling of Sensitive Clays", Proc., 11th Int. Conf. Soil Mech. and Found. Engng., San Fransisco, Vol. 2, pp. 887-892. LACASSE, S. and BERRE, T. (1988) "Triaxial Testing Methods for Soils", Symposium on Advanced Triaxial Testing of Soil and Rock", ASTM STP 977, pp. 264-269. LADD, C.C. and LAMBE, T.W. (1963) " The Strength of Undisturbed Clay Determined From Undrained Tests", Symposium on Laboratory Shear Testing of Soils, ASTM STP 361, pp. 342-371. LADD, C.C. and VARALLYAY, J. (1965) "The Influence of Stress System on the Behaviour of Saturated Clays During Undrained Shear", Research Report R65-11, No. 117, Dept. of Civil Engng., MIT, Cambridge, Massachusetts, 263 pages LADD, C.C. and FOOTI, R. (1974) "New Design Procedures for Stability of Soft Clays", Journal of the Geotech. Engng. Div., ASCE, Vol. 100, NO. OTI, pp. 763786. LAMBE, T.W. (1961) "Residual Pore Pressures in Compacted Clay", Proc., 5th Int. Conf. Soil Mech. and Found. Engng., Paris, Vol. 1, PP. 207-211. LAMBE, T.W. (1967) "The Stress Path Method", Journal of the Soil Mech. and Found. Div., ASCE, Vol. 93, No. SM6, pp. 309-331. LAMBE, T.W. and MARR, W.A. (1979) "Stress Path Method: Second Edition", Journal of the Geotech. Engng. Div., ASCE, Vol. 105, No.6, pp. 727-738. LAMBE, T.W. and WHITMAN. R.V. (1969) "Soil Mechanics", John Wiley and Sons, New York. LA ROCHELLE, P. and LEFEBVRE, O. (1971) "Sampling Disturbance in Champlain Clays", Symposium on Sampling of Soil and Rock, ASTM STP 483, pp. 143-163. 326 LA ROCHELLE, P. (1973) Discussion on the State-of-the-Art Report to Session 4, "Problems of Soil Mechanics and Construction on Soft Clays", Proc., 8th Int. Conf. Soil Mech. and Found. Engng., Moscow, Vol. 4, pp. 102-106. LA ROCHELLE, P., SARRAILH, J. and TAVENAS, F. (1976) "Effect of Storage and Reconsolidation on the Properties of Champlain Clays", Symposium on Soil Specimen Preparation for Laboratory Testing, ASTM STP 599, pp. 126-146. LA ROCHELLE, P., SARRAILH, J., TAVENAS, F., ROY, M. and LEROUEIL, S. (1981) "Causes of Sampling Disturbance and Design of a New Sampler", Canadian Geotechnical Journal, Vol. 18, No.1, pp. 52-66. LA ROCHELLE, P., LEROUEIL, S., TRAK, B., BLAIS-LEROUX, L. and TAVENAS, F. (1988) "Observational Approach to Membrane and Area Corrections in Triaxial Tests", Symposium on Advanced Triaxial Testing of Soil and Rock, ASTM STP 977, pp. 715-731. LEFEBVRE, G. and POULIN, C. (1979) "A New Method of Sampling in Sensitive Clay", Canadian Geotechnical Journal, Vol. 16, No.1, pp. 226-233. LEROUEIL, S., TAVENAS, F., LA ROCHELLE, P. and TREMBLAY, M. (198S) "Influence of Filter Paper and Leakage on Triaxial Testing", Symposium on Advanced Triaxial Testing of Soil and Rock, ASTM STP 977, pp. 189-201. LEVADOUX, J.N. and BAUGH, M.M. (1980) "Pore Pressures During Cone Penetration in Clays", Research Report No. RSO-15, Order No. 666, Dept. of Civil Engng., MIT, Cambridge, Massachusetts, 310 pages. LEWIN, P.1. (1971) "Use of Servo Mechanisms for Volume Change Measurement and x, Consolidation", Geotechnique, Vol. 21, No.3, pp. 259-262. LOWE,1. and JOHNSON, T.C. (1960) "Use of Back Pressure to Increase the Degree of Saturation of Triaxial Test Specimens", Research Conf. on Shear Strength of Cohesive Soils, ASCE, University of Colorado, Boulder, pp. 819-836. 327 LOWE, J., ZACCHEO, P.F. and FELDMAN, H.S. (1964) "Consolidation Testing With Back Pressure", Journal of the Soil Mech. and Found. Div., ASCE, Vol. 90, No. SM5, pp. 69-86. MAGUIRE, W.M. (1975) "The Undrained Strength and Stress-Strain Behaviour of Brecciated Upper Lias Clay", Ph.D. Thesis, Imperial College, University of London. McMANIS, K.L. and ARMAN, A. (1979) "Evaluation of Design Parameters Obtained by Conventional Sampling", Proc., 7th Int. European Conf. Soil Mech. and Found. Engng., Brighton, England, Vol. 2, pp. 81-86. MAYNE, P.W. and HOLTZ, R.D. (1985) "Effect of Principal Stress Rotation on Clay Strength", Proc., 11th Int. Conf. Soil Mech. and Found. Engng., San Fransisco, Vol. 2, pp. 579-582. MENZIES, B.K. (1975) "A Device for Measuring Volume Change", Geotechnique, Vol. 25, No. I, pp. 133-134. MESRI, G. and ROKHSHAR, A. (1974) "Theory of Consolidation for Clays", Journal of the Geotech. Engng. Div., ASCE, Vol. lOO,No. GT8, pp. 889-904. MILOVIC, D.M. (1971a) "Effect of Sampling on Some Soil Characteristics", Symposium on Sampling of Soil and Rock, ASTM STP 483, pp. 164-179. MILOVIC, D.M. (1971b) "Effect of Sampling on Some Loess Characteristics", Specialty Session on Quality in Soil Sampling, 4th Asian Conf., Int. Soc. Soil Mech. and Found. Engng., Bangkok, pp. 17-20. MITACHI, T., KOHATA, Y. and KUDOH, Y. (1988) "The Influence of Filter Strip Shape on Consolidated Undrained Triaxial Extension Test Results", Symposium on Advanced Triaxial Testing of Soil and Rock, ASTM STP 977, pp. 667-678. MITCHELL, R.I. (1970) "On the Yielding and Mechanical Strength of Leda Clays", Canadian Geotechnical Journal, Vol. 7, No.3, pp. 297-312. 328 NAKASE, A., KUSAKABE, O. and NOMURA, H. (1985) "A Method for Correcting Undrained Shear Strength for Sample Disturbance", Soils and Foundations, Vol. 25, No. I, pp. 52-64. NELSON, J.D., BRAND, E.W., MOH, Z.C. and MASON, lD. (1971) "The Use of Residual Effective Stress to Define Sample Quality", Proc., Specialty Session on Quality in Soil Sampling, 4th Asian Conf.,Int. Soc. Soil Mech. and Found. Engng., Bangkok, pp. 82-87. NOORANY, I. and SEED, H.B. (1965) "In-Situ Strength Characteristics of Soft Clays", Journal of the Soil Mech. and Found. Div., ASCE, Vol. 91, No. SM2, pp. 49-79. OLSON, R.E. and KIEFER, M.L. (1963) "Effect of Lateral-Filter Paper Drains on the Triaxial Shear Characteristics of Soils", Symposium on Laboratory Shear Testing of Soils, ASTM STP 361. pp. 482-491. OKUMURA, T. (1971) "The Variation of Mechanical Properties of Clay Samples Depending on its Degree of Disturbance", Proc., Specialty Session on Quality in Soil Sampling, 4th Asian Conf., Int. Soc. Soil Mech. and Found. Engng., Bangkok, pp. 73-81. PARRY, R.H.G. (1960) "Triaxial Compression and Extension Tests on Remoulded Saturated Clay", Geotechnique, Vol. 10, No.4. pp. 116-180. PARRY, R.H.G. and NADARAJAH, V. (1973) "A Volumetric Yield Locus for Lightly Overconsolidated Clay", Geotechnique, Vol. 23, No.4, pp. 451-453. PARRY, R.H.G. and NADARAJAH, V. (1974) "Observations on Laboratory Prepared Lightly Overconsolidated Specimens of Kaolin", Geotechnique, Vol. 24, No.3, pp. 345-358. PARRY, R.H.O. and WROTH, C.P. (1981) "Shear Stress-Strain Properties of Soft Clay", ChapterzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA 4 in Soft Clay Engineering, Elsevier Scientific Publishing Company, The Netherlands. 329 RANDOLPH, M.F., STEENFELT, J.S. and WROTH, C.P. (1979) "The Effect of Pile Type on Design Parameters for Driven Piles", Proc., 7th European Conf. Soil Mech. and Found. Engng., Brighton, England, Vol. 2, pp. 107-114. RAO, S.S. (1989) "The Finite Element Method in Engineering", 2nd Edition, Pergamon Press, Oxford. RAYMOND, G.P., TOWNSEND, D.L. and LOJK.ASEK, M.J. (1971) "The Effect of Sampling on the Undrained Soil Properties of a Leda Soil", Canadian Geotechnical Journal, Vol. 8, No.4, pp. 546-557. RICHARDSON, A.M. and WHI1MAN, R.V. (1963) i'Effect of Strain Rate Upon Undrained Shear Resistance of a Saturated Remoulded Fat Clay", Geotechnique, Vol. 13, No.4, pp. 310-324. ROWE, P.W. (1972) "The Relevance of Soil Fabric to Site Investigation Practice", 12th Rankine Lecture, Geotechnique, Vol. 22. No.2, pp. 195-200. ROBINSKY, E.I. and MORRISON, C.F. (1964) "Sand Displacement and Compaction Around Friction Piles", Canadian Geotechnical Journal, Vol. I, pp. 81-93. ROWLANDS, 0.0. (1972) "Apparatus for Measuring Volume Change Suitable for Automatic Logging", Oeotechnique, Vol. 22, No.3, pp. 525-526. RU1LEDGE, P.C. (1944) "Relation of Undisturbed Sampling to Laboratory Testing", Transactions, ASCE, Vol. 109, Paper No. 2229, pp. 1155-1183. SARRAILH, J. (1975) "Contribution a L'etude des Methodes D'echantillonnage des Argiles Sensibles", M.Sc. Thesis, Department de Genie Civil, Universite Laval, Quebec, P.Q. SCHJETNE, K. (1971) "The Measurement of Pore Pressure During Sampling", Proc., Specialty Session on Quality in Soil Sampling, 4th Asian Conf., Int. Soc. Soil Mech. and Found. Engng., Bangkok, pp. 12-16. 330 SCHMERTMANN, J.H. (1955) "The Undisturbed Consolidation Behaviour of Clay", Transactions, ASCE, Vol. 120, Paper No. 2775, pp. 1201-1227. SCHMERTMANN, J.H. (1956) Discussion on "Effect of Sample Disturbance on the Strength of a Clay", Transactions, ASCE, Vol. 121, pp. 940-950. SCOTT, R.F. (1963) "Principles of Soil Mechanics", Addison-Wesley Publishing Company, Inc., London. SEED, H.B., NOORANY, I. and SMITH, I.M. (1964) "Effects of Sampling and Disturbance on the Strength of Soft Clay", Research Report TE-64-1, University of California, Berkley. SHACKEL, B. (1971) "Some Aspects of Sampling Disturbance Observed Using a Nuclear Method", Proc., Specialty Session on Quality in Soil Sampling, 4th Asian Conf., Int. Soc. Soil Mech. and Found. Engng., Bangkok, pp. 7-11. SIMONS, N.E. and SOM, N.N. (1970) "Settlement of Structures on Clay With Particular Emphasis on London Clay", Construction Industry Research and Information Association (CIRIA), Report No. 22, pp. 1-51. SKEMPTON, AW. (1961) "Horizontal Stresses in an Over-Consolidated Eocene Clay", Proc., 5th Int. Conf. Soil Mech. and Found. Engng., Paris, Vol. 1, pp. 351357. SKEMPTON, A.W. and SOWA, V.A. (1963) "The Behaviour of Saturated Clays During Sampling and Testing", Geotechnique, Vol. 13, No.4, pp. 269-290. SONE, S., TSUCHIYA, H. and SAITO, Y. (1971) "The Deformation of a Soil Sample During Extrusion from a Sample Tube", Proc., Specialty Session on Quality in Soil Sampling, 4th Asian Conf., Int. Soc. Soil Mech. and Found. Engng., Bangkok, pp. 3-6. SYMES, M. and BURLAND. lB. (1984) "Determination of Local Displacement on 331 Soil Samples", Geotechnical Testing Journal, Vol. 7, No.2, TERZAGHI, pp. 49-59. K. (1943) "Theoretical Soil Mechanics", John Wiley and Sons, New York, pp. 265-285. TAVENAS, F., JEAN, P., LEBLOND, Permeability of Natural Soft Clays. Geotechnical VOLD, Journal, Vol. 20, No.4, R.C. Geotechnical (1956) "Opptagning P. and LEROUEIL, S. (1983) "The PartzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPO II: Permeability Characteristics", Canadian pp. 645-660. av Uforstyrrede Jordprover", Norwegian Institute, Publication No. 17, pp. 1-14. WATTS, K.S. (1980) "A Device for Automated Logging of Volume Change in Large Scale Triaxial Tests", Geotechnical Testing Journal, Vol. 3, No. I, pp. 41-44. YUEN, C.M.K., LO, K.Y. and PALMER, IH.L. (1978) "A New Apparatus for Measuring the Principal Strains in Anisotropic Clays", Geotechnical Testing Journal, Vol. 1, No.1, pp. 24-33. 332 APPENDIX· A zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIH LISTING OF COMPUTER PROGRAMS FOR STRAIN PATII COMPUTATION 333 C PROGI C C C C THIS PROGRAM PRINTS A LISTING OF ELEMENT TOPOLOGY C NODE COORDINATES AND FLOW VELOCITY IN Y-DIRECTION C FROM THE LUSAS OUTPUT FILE C C C PARAMETERS DEFINITIONS C C NAMEl = NAME OF LUSAS OUTPUT FILE C CzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA NAME2 = NAME OF NEW OUTPUT FILE THAT CONTAINS A LISTING C OF ELEMENT TOPOLOGY, NODE COORDINATES AND C NODAL VELOCITY IN Y-DIRECTION OF THE C FINITE ELEMENT MESH C C N2 = NUMBER OF NODES IN EACH ELEMENT C C CHARACTER*200 A CHARACTER *20 NAMEl,NAME2 C DIMENSION IC(10000,1O) COMMON IC C PRINT*,'ENTER NAME OF LUSAS OUTPUT FILE' READ(*,'(A)') NAMEI C PRINT*,'ENTER NAME OF NEW OUTPUT FILE' READ(*, '(A)') NAME2 C OPEN(5,FILE=NAMEl) OPEN(6,FILE=NAME2) C PRINT*, 'ENTER NUMBER OF NODES IN EACH ELEMENT' READ*, N2 C C C 10 SEARCH FOR 'E L E MEN L=30000 DO 10 I=l,L READ(5,'(A)',END=20) A IF(A(11 :23).EQ.'E L E MEN CONTINUE T' IN THE LUSAS OUTPUT FILE 'NAMEl' T') GOTO 30 C 20 C 30 C C C PRINT*, 'CAN NOT FIND E L E MEN STOP T' CONTINUE SKIP SIX LINES CALL SKIP(6) C C READ ELEMENT TOPOLOGY FROM FILE 'NAMEl' 334 AND C C 40 C SO C C C 60 C C C 70 C SO C 90 C C C PRINT IN THE NEW OUTPUT FILE 'NAME2' DO 40 l=l,L READ(S,'(A),) A IF(A(1:1S).EQ.' ') GOTOzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA SO READ(A, "') L l,(IC(L I,J),J= l,N2) WRITE(6,'(10I7),) Ll,(IC(Ll,J),J=l,N2) CONTINUE CONTINUE PRINT A BLANK LINE WRITE(6,60) FORMAT(, SEARCH FOR NOD ') E IN THE LUSAS OUTPUT FILE 'NAME1' DO 70 l=l,L READ(S,'(A)',END=SO) A IF(A(11:17).EQ.'N ODE') CONTINUE GOTO 90 PRINT"', 'CAN NOT FIND NOD STOP E' CONTINUE SKIP FIVE LINES CALL SKIP(S) C C C C SEARCH FOR'S PAC lNG' AND '''''''*WARNING***' LUSAS OUTPUT FILE 'NAME1' IN THE DO 100 l=l,L READ(5,'(A),) A IF(A(11:23).EQ.'S PAC I N G') GOTO 100 IF(A(1 :20).EQ.' ') GOTO 100 IF(A(2:14).EQ.'***WARNING"'**') GOTO 110zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLK C C C READ NODE COORDINATES FILE 'NAME2' READ(A, "') L1,X,Y WRlTE(6,'(I7,2F15.4)') L1,X,Y too CONTINUE C 110 CONTINUE C C PRINT A BLANK LINE C FROM FILE 'NAME1' WRITE(6,60) SEARCH FOR 'F I E L D' IN THE LUSAS OUTPUT FILE 'NAME1' DO 1201=t,L READ(S,'(A)', END=130) A IF(A(11:19).EQ.'F I E L D') OOTO 140 120 AND PRINT IN CONTINUE 335 C 130 PRINT*,'CAN NOT FIND FIELD' STOP C 140 CONTINUE C C SKIPzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA NINE LINES C CALL SKIP(9) C C READ FLOW VELOCITY FROM FILE 'NAMEl' AND PRINT VELOCITY C IN Y-DIRECTION IN FILE 'NAME2' C READ(S,zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA '(A)') AzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA 150 A(I:I2)=' READ(A,*) LI,D,E.F,O WRITE(6,'(l7,FI2.7)') LI,O DO 160 I=I,{N2-I) READ(S, '(A)') A READ(A,*) Ll,D,E,F,O WRITE(6,'(I7.FI2.7)') LI.O 160 CONTINUE C C SKIP TWO LINES C CALL SKIP(2) C READ(S,'(A)') A ') OOTO 170 IF(A( 1:20).EQ. ' OOTO 150 C C PRINT A BLANK LINE C 170 WRlTE(6,60) PRINT*.' JOB HAS BEEN COMPLETED' C CLOSE(S) CLOSE(6)zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA END C C C SUBROUTINE TO SKIP LINE(S) SUBROUTINE SKIP(N) CHARACTER*I00 B DO 180 I=l,N READ(S,'(A)') B 180 CONTINUE RETURN END 336 C PROG2 C C CzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA C THIS PROGRAM CALCULATES THE VALUES OF AVERAGE VELOCITY C IN Y-DIRECTION AND STREAM FUNCTION OF EACH NODE OF THE C FINITE ELEMENT MESH C C C PARAMETERS DEFINITIONS C C NAME3 FILE CONTAINING THE ARRAY OF NODE NUMBERS C OF THE FINITE ELEMENT MESH C C NAME2 = FILE THAT CONTAINS LISTING OF ELEMENT TOPOLOGY. C NODE COORDINATES AND NODAL VELOCITIES OF THE C FINITE ELEMENT MESH C C NAME4 = NEW OUTPUT FILE CONTAINING A LISTING OF THE C VALUES OF X AND Y-COORDINATES. AVERAGE VELC OCITY IN Y-DIRECTION AND STREAM FUNCTION OF C EACH NODE OF THE FINITE ELEMENT MESH C C = TOTAL NUMBER OF COLUMNS IN THE ARRA Y CONTAINING Nt C NODE NUMBERS OF THE FINITE ELEMENT MESH C C TOTAL NUMBER OF NODES IN EACH ELEMENT N2 C C Lt = TOTAL NUMBER OF ROWS IN THE ARRAY CONTAINING C NODE NUMBERS OF THE FINITE ELEMENT MESH C C TOTAL NUMBER OF ELEMENTS IN FINITE ELEMENT MESH L2 C C Ml TOTAL NUMBER OF NODES IN FINITE ELEMENT MESH C C CHARACTER*500 A CHARACTER*12 NAME3.NAME2.NAME4 C DIMENSION IB(10000.tOO).IC(10000.10).D(20000.2).E(20000.2) DIMENSION F(20000).S(O:20000) C COMMON IB.IC COMMON /SHAM/D.E.F.S = = = = c PRINT PRINT READ( 'ENTER NAME OF THE FILE CONTAINING THE ARRAY' 'OF NODE NUMBERS OF THE FINITE ELEMENT MESH' ·(A)·) NAME3 c c c PRINT*. 'ENTER NAME OF FILE CONTAINING ELEMENT TOPOLOGY' PRINT .... 'NODE COORDINATES AND NODAL VELOCITIES' READ(*.·(A)·) NAME2 PRINT .... 'ENTER NAME OF NEW OUTPUT FILE' READ( -. '(A)') NAME4 337 OPEN(5,FILE=NAME3) OPEN(6,FILE=NAME2) OPEN(7,FILE=NAME4) C PRINT*, 'ENTER NO. OF COLUMNS IN THE ARRAY CONTAINING' PRINT*, 'NODE NOS. OF THE FINITE ELEMENT MESH' READ*, NI C PRINT*, 'ENTER NUMBER OF NODES IN EACH ELEMENT' READ*, N2 C C C 10 C 20 C READ NODE NUMBERS AT DIFFERENT ROWS FROM THE FILE 'NAME3' M=30000 DO 10 I=l,M READ(5,'(A)') A IF (A(l:1O).EQ.' ') GOTO 20 READ(A,*) (IB(I,J),J=l,Nl) CONTINUE CONTINUE Ll=I-l C 30 C C C 40 C C C 50 C 60 C WRITE(7,30) L 1 FORMAT('TOTAL NUMBER OF ROWSzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPO IN THE ARRAY OF NODE NOS.=',I3) PRINT A BLANK LINE WRlTE(7,40) FORMAT(, ') READ ELEMENT TOPOLOGY FROM THE FILE 'NAME2' DO 50 I=l,M READ(6,'(A)') A IF (A(1:1O).EQ.' ') GOTO 60 READ(A, *) L,(IC(L,J),J= l,N2) CONTINUE CONTINUE L2=I-l C 70 WRITE(7,70) L2 FORMAT(,TOTAL NUMBER OF ELEMENTS=',I5J) C C CALCULATE TOTAL NUMBER OF NODES IN THE FINITE ELEMENT MESH C 90 80 C Ml=O DO 80I=l,L2 DO 90 J=l,N2 MAX=IC(I,J) IF (MAX.GT.Ml) CONTINUE CONTINUE Ml=MAX 338 WRITE(7,100) Ml 100 FORMAT(,TOTAL NUMBER OF NODES=',I5j) C WRITE(7,110) 110 FORMAT('NODE NO.',5X,'X-COOR.',8X,'Y-COOR.',5X,'AVG. VELOCITY', *3X, 'STREAM FUNCTION'zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA j) C C READ NODE COORDINATES FROM THE FILE 'NAME2' C DO 120I=I,M READ(6, '(A)') A IF (A(I:10).EQ.' ') GOTO 130 READ(A,*) J,D(1,I),0(1,2) 120 CONTINUE C 130 CONTINUE C C READ FLOW VELOCITY IN Y -DIRECTION CORRESPONDING C TO EACH NODE FROM THE FILE 'NAME2' C DO 140 I=I,M READ(6, '(A)') A IF (A(1:lO).EQ.' ') GOTO 150 READ(A,*) K,V E(K,I)=E(K,I)+V E(K,2)=E(K,2)+ 1 140 CONTINUE C C CALCULATE AVERAGE FLOW VELOCITY IN Y-DIRECTION C 150 DO 160 I=I,Ml F(I)=E(I, 1)/E(I,2) 160 CONTINUE C C SEARCH FOR THE CONNECTIVITY BETWEEN NODES C DO 230 I=I,Ll S(IB(I,I»=O DO 220 J=I,(Nl-1) DO 200 Il=I,L2 DO 170 I2=1,N2 IF (IB(I,J).EQ.IC(Il,I2» GOTO 180 170 CONTINUE GOTO 200 180 CONTINUE DO 190 12=I,N2 IF (IB(I,J+l).EQ.IC(Il,I2» GOTO 210 190 CONTINUE 200 CONTINUE C S(lB(I,(J+ l»)=S(IB(I,J» GOTO 220 210 CONTINUE C CONNECTIVITY BETWEEN NODE B(I,J) AND B(I,J+l) C HAS BEEN CONFIRMED. C C CALCULATE STREAM FUNCTION AT EACH NODE 339 C A 1=2.0944*(F(lB(I,(J+ 1)))-F(lB(I,J))) A2=(D(IB(I,J), 1)**2+D(lB(I,(J+ 1)),1)**2)zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCB A2=A2+ D(IB(I,J), 1)*D(lB(I,(J+ 1)), 1) PI=3.141592653 A3=(F(IB(I,J))*D(IB(I,(J+ 1)), 1)) A3=A3-F(IB(I,(J+ l)))*D(IB(I,J), 1) A3=A3*PI A4=D(IB(I,(J+ 1)), l)+D(IB(I,J), 1) S(IB(I,(J+ l)))=S(IB(I,J))+A 1*A2+A3* A4zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFE C 220 230 C C C C C CONTINUE CONTINUE PRINT THE VALUES OF COORDINATES, AVERAGE FLOW VELOCITY IN Y-DIRECfION AND STREAM FUNCfION CORRESPONDING TO EACH NODE IN THE FINITE ELEMENT MESH DO 250 I=1,M1 WRlTE(7,240) 1,(D(I,J).1= 1,2),F(I),s(I) 240 FORMAT(I5,2F15.4,F16.7,F20.5) 250 CONTINUE C PRINT*, 'JOB HAS BEEN COMPLETED' C CLOSE(5) CLOSE(6) CLOSE(7) C ENDzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA C ----------------------------------------------------------- C PROG3 C ----------------------------------------------------------- C C ----------------------------------------------------------- C C C THIS PROGRAM CALCULATES THE VALUES OF STREAM FUNCfION CORRESPONDING TO DIFFERENT STREAM LINES AND THE COORDINATES AT DIFFERENT DEPTHS ALONG THE STREAMLINES C ----------------------------------------------------------- C C C C C C C C C C C C C C PARAMETERS NAME3 DEFINITIONS = FILE THAT CONTAINS AN ARRAY OF THE NODE NUMBERS OF THE FINITE ELEMENT MESH NAME2 = FILE THAT CONTAINS ELEMENT TOPOLOGY, NODE COORDINATES AND NODAL VELOCITIES OF THE FINITE ELEMENT MESH NAME5 = NEW OUTPUT FILE CONTAINING A LISTING OF THE MAGNITUDES OF THE STREAM FUNCfION OF DIFFERENT STREAM LINES AND THEIR CORRESPONDING COORDINATES AT DIFFERENT DEPTHS C 340 C C C C C C C C C C C C C C C C C C C NIzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA = TOTAL NUMBER OF COLUMNS IN THE ARRAY CONTAINING NODE NUMBERS OF THE FINITE ELEMENT MESH N2 = TOTAL NUMBER OF NODES IN EACH ELEMENT N3 = NUMBER RI = INTERNAL RADIUS OF CUTTING SHOE OF THE SAMPLER LI = TOTAL NUMBER OF ROWS IN THE ARRAY CONTAINING NODE NUMBERS OF THE FINITE ELEMENT MESH L2 = TOTAL NUMBER OF ELEMENTS L3 = LEVEL NUMBER CORRESPONDING Mt OF COLUMNS IN THE ARRAY CONTAINING NODE NUMBERS COUNTING FROM THE CENTRELINE TO THE INSIDE EDGE OF THE SAMPLER = TOTAL IN FINITE ELEMENT MESH TO BOTTOM OF THE SAMPLER NUMBER OF NODES IN FINITE ELEMENT MESH C CHARACTER*500 A CHARACTER*12 NAME3,NAME2,NAME5 C DIMENSION DIMENSION IB(1 OOOO,IOO),IC(1 OOOO,lO),D(20000,2),E(20000,2) F(20000),s(O:20000) C COMMON IB,IC COMMON/SHAM/D,E.F,S PRINT*, 'ENTER NAME OF FILE CONTAINING THE ARRAY OF' PRINT*, 'NODE NUMBERS OF THE FINITE ELEMENT MESH' READ(*,'(A)') NAME3 PRINT*, 'ENTER NAME OF FILE CONTAINING ELEMENT TOPOLOGY' PRINT*, 'NODE COORDINATES AND NODAL VELOCITIES' READ(*, '(A) ') NAME2 PRINT*, 'ENTER NAME OF NEW OUTPUT FILE' READ(*,'(A)') NAME5 C OPEN(5,FILE=NAME3) OPEN(6,FILE=NAME2) OPEN(7,FILE=NAME5) C C PRINT*, 'ENTER NO. OF COLUMNS IN THE ARRAY CONTAINING' PRINT*, 'NODE NOS. OF THE FINITE ELEMENT MESH' READ*, Nt PRINT*, 'ENTER NO. OF COLUMNS IN THE ARRAY CONTAINING' PRINT*, 'NODE NOS. COUNTING FROM THE CENTRELINE' PRINT*, 'TO THE INSIDE EDGE OF THE SAMPLER' READ*, N3 PRlNT*, 'ENTER NUMBER OF NODES IN EACH ELEMENT' READ*, N2 PRlNT*, 'ENTER INSIDE RADIUS OF THE CUTTING EOOE' PRlNT*, 'OF THE SAMPLER' READ*, RI PRINT*, 'ENTER THE ROW NO. CORRESPONDING' PRINT*, 'TO BOTTOM OF THE SAMPLER' READ"', L3 READ NODE NUMBERS AT DIFFERENT ROWS FROM FILE 'NAME3' 341 C 10 C 20 M=30000 DO 10 I=l,M READ(5,'(A),) A IF (A(1: 1O).EQ.' ') OOTO 20 READ(A,*) (IB(I,J),J=l,Nl) CONTINUE CONTINUE Ll=L-I C C C 30 40 C C C 60 50 C C C 70 C 80 C C C C 90 C C C 100 READ ELEMENT TOPOLOGY FROM FILE 'NAME2' DO 30 I=l,M READ(6, '(A) ') A IF (A(l:IO).EQ.' ') GOTO 40 READ(A,*) L,(IC(L,J),J=1,N2) CONTINUE CONTINUE L2=I-l CALCULATE TOTAL NUMBER OF NODES IN THE FINITE ELEMENT MESH Ml=O DO 50 I=1,L2 DO 60 J=1,N2 MAX=IC(I,J) IF (MAX.OT.Ml) CONTINUE CONTINUE Ml=MAX READ NODE COORDINATES FROM FILE 'NAME2' DO 70 I=l,M READ(6,'(A)') A IF (A(1: 1O).EQ.' ') GOTO 80 READ(A,*) J,D(1,1),D(1,2) CONTINUE CONTINUE READ FLOW VELOCITY IN Y-DIRECfION TO EACH NODE FROM FILE 'NAME2' DO 90 I=l,M READ(6,'(A)') A IF (A(1:IO).EQ.' READ(A, *) K, V E(K,l)=E(K,l)+ V E(K,2)=E(K,2)+ 1 CONTINUE CALCULATE ') GOTO 100 AVERAGE FLOW VELOCITY DO 110 I=1,M1 F(I)=E(I,l )IE(I,2) 110 CORRESPONDING CONTINUE 342 IN Y-DIREcrIONzyxwvutsrqponmlkjihgfedcbaZYXWV C C C C 120 130 140 150 160 C C C C SEARCH FOR CONNECfIVITY ELEMENT MESH BETWEEN NODES IN THE FINITE DO 180 1=I,Ll S(18(1,I»=O DO 170 J=I,(Nl-l) DO 150 Il=I,L2 DO 120 12=I,N2 IF (IB(I,J).EQ.IC(Il,12» GOTO 130 CONTINUE GOTO 150 CONTINUE DO 140 12=I,N2 IF (IB(I,J+l).EQ.IC(Il,I2» GOTO 160 CONTINUE CONTINUE S(18(1,(J+ 1»)=S(18(I,J» GOTO 170 CONTINUE CONNECfIVITY BETWEEN NODE B(I,J) AND B(I,J+ 1) HAS BEEN CONFIRMED. CALCULATE STREAM FUNCfION AT EACH NODEzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQ C A 1=2.0944*(F(lB(I,(J+ 1»)-FC18CI,J») A2=(D(lB(I,J),I)**2+D(18(I,(J+ 1»,1)**2) A2=A2+D(IB(I,J),I)*D(IB(I,(J+ 1»,1) PI=3.141592653 A3=(F(18(I,J»*D(18(I,(J+ 1),1» A3=A3-F(18(1,(J+ 1»)*D(IB(I,J),I) A3=A3*PI A4=D(IB(I,(J+ 1»,I)+O(IB(I,J),I) S(IB(I,(J+ 1»)=S(IB(I,J»+A 1*A2+A3* A4 C 170 180 C C C C C CONTINUE CONTINUE CALCULATE AND PRINT VALUES OF STREAM FUNCfION CORRESPONDING TO DIFFERENT STREAM LINES AND THEIR RESPECflVE COORDINATES AS WELL K=l X2=0.I*Rl PRINT 230, X2 DO 240 J=l,(N3-1) B2=S(IB(L3,J»-S(IB(L3,(J+ 1») B2=B2/(0(IB(L3,J),I)**2-0(IB(L3,(J+l»,I)**2) C2=S(IB(L3,J»- B2 *O(lB(L3,J),1 )**2 190 IF (X2.LE.(D(IB(L3,(J+I»,1)+1.0E-5» GOTO 200 GOTO 240 200 SI=C2+B2*X2**2 DO 220 I=I,Ll B I=S(IB(I,J»-S(IB(I,(J+ 1») B I=B 1/(0(IB(I,J),I)**2-0(lB(I,(J+ 1»,1)**2) Cl=S(lB(I,J»-B 1*D(IB(I,J),I)**2 XI=SQRT«S l-CI)/B I) 343 WRITE(7,21O) K,St,Xt,D(IB(I,J),2) FORMAT(IS,2X,FI2.8,FIS.4,FI7.6) CONTINUE K=K+I X2=X2+0.I*RI PRINT 230, X2 230 FORMAT(F14.8) IF (X2.GT.(RI+1.0E-S» GOTO 2S0 GOTO 190 240 CONTINUE 2S0 PRINT*, 'JOB HAS BEEN COMPLETED' C CLOSE(S) CLOSE(6) CLOSE(7) C END 210 220 C PROG4 C CzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA C C C C C C C C C C C C C C C C C C C C C C C THIS PROGRAM CALCULATES THE MAGNITUDES OF RADIAL AND AXIAL STRAIN AT DIFFERENT DEPTHS ALONG THE STREAMLINES PARAMETERS DEFINITIONS NAMES = FILE NAME6 = NEW Kt = TOTAL NUMBER OF STREAMLINES t.r = TOTAL THAT CONTAINS A LISTING OF THE MAGNITUDES OF THE STREAM FUNCfION OF DIFFERENT STREAMLINES AND THEIR CORRESPONDING COORDINATES AT DIFFERENT DEPTHS OUTPUT FILE THAT CONTAINS A LISTING OF THE MAGNITUDES STREAM FUNCTION, X AND Y-COORDINATES AND RADIAL AND AXIAL STRAINS NUMBER OF ROWS IN THE ARRAY CONTAINING NODE NUMBERS OF THE FINITE ELEMENT MESH CHARACTER*I00 A CHARACTER·12 NAMES,NAME6 C DIMENSION D(1000,1O) C PRINT·, 'ENTER NAME OF INPUT FILE' READ(* ,'(A),) NAMES C PRINT·, 'ENTER NAME OF NEW OUTPUT FILE' READ(*,'(A)') NAME6 CzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA 344 OPEN(5,FILE=NAME5) OPEN(7,FILE=NAME6) C PRINT*, 'ENTER TOTAL NUMBER OF STREAM LINES' READ*, Kl C PRINT*, 'ENTER TOTAL NUMBER OF ROWS IN THE ARRAY' PRINT*, 'CONTAINING NODE NUMBERS' READ*, LI C C C 10 20 C C C C 30 40 C READ DATA FROM THE FILE 'NAMES' K=l DO 20 I=I,LI READ(S,'(A)') A READ(A, *) L,(D(I,J),J= 1,3) CONTINUE CALCULATEzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA AND PRINT THE MAGNITUDES OF RADIAL AND AXIAL STRAINS PLUS THE RESULTS IN THE FILE 'NAMES' DO 40 I=l,Ll RS=(D(Ll,2)-D(I,2»1D(L1,2)*lOO.0 S=( -l.0)*2.0*RS WRITE(7,30) K,(D(I,J),J= 1,3),RS,S FORMAT(I5,2X,F12.8,FlS.8,FlS.6,2Fl7.6) CONTINUE K=K+l IF (K.GT.Kl) GOTO 10 C 50 GOTO 50 PRINT*, 'JOB HAS BEEN COMPLETED' C CLOSE(5) CLOSE(7) C END c C C C C C C C C C C C C C C PROGS THIS PROGRAM CALCULATES THE MAGNITUDES OF TOTAL AXIAL DEFORMATION AND DEPfH TO DIAMETER RATIO AT DIFFERENT DEPfHS ALONG THE STREAMLINES PARAMETERS DEFINITIONS NAME6 = FILE THAT CONTAINS A LISTING OF THE MAGNITUDES OF STREAM FUNCTION, X AND Y-COORDINATES AND RADIAL AND AXIAL STRAINS C 345 NAME7 = NEW OUTPUT FILE THAT CONTAINS ALL THE RESULTS IN C FILE 'NAME6' PLUS MAGNITUDES OF TOTAL AXIAL C DEFORMATION AND DEPTH TO DIAMETER RATIO AT C DIFFERENT DEPTHS ALONG THE STREAMLINES C C C TOTAL NUMBER OF STREAMLINES Kl C Ll TOTAL NUMBER OF LEVELS IN FINITE ELEMENT MESH C CzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA 01 DEPTH AT THE BOTTOM OF THE SAMPLER C = = = C C C 02 = EXTERNAL RADIUS OF THE SAMPLER CHARACTER"'IOO A CHARACTER"'12 NAME6.NAME7 C DIMENSION D(I000.lO).E{lOOO) C c PRINT .... 'ENTER NAME OF INPUT FILE' READ("','(A),) NAME6 PRINT .... 'ENTER NAME OF NEW OUTPUT FILE' READ(* ,'(A)') NAME7 C OPEN(5.FILE=NAME6) OPEN(7.FILE=NAME7) C PRINT*. 'ENTER TOTAL NUMBER OF STREAMLINES' READ*, Kl C PRINT*. PRINT PRINT READ*. 'ENTER TOTAL NUMBER OF ROWS IN THE ARRAY' 'CONTAINING NODE NUMBERS IN THE' 'FINlTE ELEMENT MESH' L1 C PRINT*. 'ENTER DEPTH AT THE BOTTOM OF SAMPLER' READ*.Dl PRINT .... 'ENTER EXTERNAL DIAMETER OF THE SAMPLER' READ*.02 C 10 WRITE(7.10) FORMAT(, STREAM' .4X. 'STREAM' ,lOX, ·X-COOR.· .9X,·Y -COOR.' ,7X, ""RADIAL STRAIN·,4X.·AXIAL STRAIN·.6X,·TOTAL AXIAL'. *6X. ·DEPTH/DIA.·) C 20 C C C WRITE(7.20) FORMATCLINE NO.·.2X.·FUNCTION·.8X.·(MM)·,12X,·(MM)',lOX, *'(PERCENT), .8X, '(pERCENT), ,9X, ·DEFORM.(MM)' ,6X, 'RATIO') PRINT A BLANK LINE WRITE(7,80) C C C 30 READ DATA FROM FILE 'NAME6' K=l DO 40 I=l,Ll 346 40 C C C C C 50 60 70 C C C 80 C READ(5,'(A)') A READ(A, *) L,(D(I,J),J= 1,5) CONTINUE CALCULATE AND PRINT THE VALUES OF TOTAL AXIAL DEFORMATION AND DEPI'H TO DIAMETER RATIO PLUS RESULTS IN THE FILE 'NAME6' E(Ll)=O.O DO 50 I=(Ll-l),I,-1 E(I)=E(I+ 1)+«0(1,5)+0«1+ 1),5»/2)*(0(1,3)-0«1+ 1),3»*.01 CONTINUE DO 70 I=I,L1 03=(0(1,3)-01)/02 WRlTE(7,60) K,(O(I,J),J= 1,5),E(I),03 FORMAT(15,2X,FI2.8,FI5.8,F15.6.2FI7.6,FI7.8,F16.6) CONTINUE PRINT A BLANK LINE WRITE(7,80) FORMAT(, K=K+l IF (K.GT.KI) GOTO 30 C 90 C PRINT*,' ') GOTO 90 JOB HAS BEEN COMPLETED' CLOSE(5) CLOSE(7) C END 347 APPENDIX - BzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLK BASIC ALGORITHM FOR SCANNING SIGNALS FROM VARIOUS MEASURING DEVICES 348 20 ON ERROR GOTO 50 30 LOADBIN "UTIL/1" 40 OFF ERROR 50 REM ******** READ SIGNALS FROM SGA BOX ******** 60 DIM R(13), S(13), U$(13), D$(13) 70 U$(O)="NOT USED " 80 U$(1)="BACK PRESS" @ U$(2)="CELL PRESS" @ U$(3)="AXIAL PRESS" 90 U$(4)="H.E.CALIPER" @ U$(5)="H.E.GAUGE13" @ U$(6)="H.E.GAUGE12" 100 U$(7)="L.S.C.D.T. " @ U$(8)=VOL.CHANGE "@ U$(9)="PORE PRESS " 110 GOTO 230 120 FOR JJ=l TO 9 130 SEND 7 ; UNL UNT MLA TALK 9 SCG JJ 140 ENTER 7 USING #,B,B ; SI, S2 150 S3=BINAND (S2,15) 160 S(JJ)=Sl+256*S3 170 IF BINANDzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA (S2,32)=0 THEN S(JJ)=-S(JJ) 180 R(JJ)=S(JJ) 190 D$(JJ)=VAL$ (R(JJ» 2001=JJ 210 FAST LABEL 80, 30+1*10, D$(I)&" ",0 220 NEXT JJ @ GOTO 120 230 GCLEARzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA @ GRAPH 240 FAST LABEL 5,0, "OUTPUT READINGS FROM SGA 100",1 250 FOR 1=1 TO 9 260 FAST LABEL 0, 30+1*10, "CHANNEL NO. "&VAL$ (1)&" 270 FAST LABEL 90, 30+1*1O,"BITS",1 280 NEXT I 290 GOTO 120 300 END 349 "&U$(I),1 APPENDIX· C zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONML SUMMARY OF STRESS AND STRAIN PATH TEST RESULTS 350 1.2r----------------r-----------------------, Solid symbol represents presheor conditionzyxwvutsrqponmlkjihgfedcbaZYXWVUTS TESTS I and 2 .8 .4 '0:.-zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA 'd:- -. 4 -.8L-----~~----~------~------~~----~ -. 8 -. 4 O. 0 •4 •8 1. 2 1.0~-----------~-------------------------------------------, TESTS I end 2 Solid symbol represents preshear condition .8 .6 .4 .2 -.2 _.4~ -. 4 L_ L_ -. 2 O. 0 L_ •2 L_ .4 351 L_~ •6 L_ •8 ~ 1. 0 ,50 ~----------------------------------------------------~ ,45 TEST 1 .40zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA .35 ----- EXTERNAL ................ LOCAL AX J AL SHEAR ,30 ,25 .20 ,150L-------~------~2------~3~------~4------~5------~6~------:7 STRAIN GO ,5 .4 ,3 .2zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA •1 U ,> 0 0.0 '"'er -,1 -.2 -.3 -,4 -,5 0 2 4 6 8 EXTERNAL AXIAL STRAIN 352 12 10 (7.) 14 16 .80 r-----------------------------------------------------------, TEST 1 .75 //_..---_"-zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA .70 .65 ,,- ,, ,, .60 , ,, " " --- EXTERNAL AXIAL -------- LOCAL SHEAR .55 .50 .450~------~------~2------~3~------~4--------5~------~6------~7zyxwvutsrqponmlkjihgfedcba STRAIN .75 (7.) r-----------------------------------------------------------, TEST 2 .50 .25 ---- EXTERNAL AXIAL -------- LOCAL SHEAR 0.00 ~~.-~------_r------;_------r_----_;------_+------,_------1 '0.. ,'? -. 25 -.50 -.75 -1.00 -1.25 ~----~~----~------~------~----~~----~------~----~ o 2 6 4 8 10 12 STRAIN 353 00 14 16 2. 2 ,-... U1 U1 TEST 2. 1 --- 1 EXTERNAL AXIAL W -------LOCAL SHEAR et:zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJ IU1 zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA 2.0 _J < ...... Cl < 1.9 et: u, lL. W -o,J zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA 1.8 ,-... U1 U1 W et: 1.7 IU1 _J -c ...... x -e 1.6 u, u, w 1.5 o,J 1.4 3 2 0 4 STRAIN (7.) 1.6 ,-... U1 U1 w TEST et: EXTERNAL IU1 _J -e ...... Cl -e . et: u, lL. W .._. 2 1.4 1.2 -------- LOCAL AXIAL SHEAR 1.0 -..... ,..... U1 U1 w .8 et: IU1 _J -c ..... .6 x < ~ u, w .._. .4 STRAIN 354 (Yo) 5 6 7 .200 .175 .150 .125zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA uzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA ,> - b .100 ::I <l .075 .050 .025 0.000 2 0 EXTERNAL AXIAL STRAINzyxwvutsrqponmlkjihgfedcbaZYXWVUTS (X) .01 ~--------------------------------------------------~ -.01 u ,> t5 -.02 ~ -.03 -.04 -.05 ~ 0.0 ~ 2.0·· ~ 4.0 ~ -L B.O B.O ~.-..__~ 10.0 EXTERNAL AXIAL STRAIN (X) 355 12.0 ~----~ 14.0 16.0 .8r-----~~------------------------------------~ TEST 3 .6 Soljd symbol represents preshear condjtion uzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA '~ u .4 ~+JzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA .2 -.2~----~------~-------L------_L------~----~ .6 .8 1.0 .2 0.0 -.2 .30 TEST 3 .25 ..............................................zyxwvutsrqponmlkjihgfedcba .20 ----- "0\ EXTERNAL ) ................ LOCAL . 15 u \:) ....: ,> .10 ......... ............. ':J ., . 05 ........................ . .>: . ' 0.00 ... ............................................. -.05 -.10 -1. 5 -1. 0 -.5 .5 0.0 AXIAL STRAIN 356 1.0 (7.) 1.5 2.0 2.5 .30 TEST 3 .25 .20 • 15zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA u ,> \:) .10 ':J .05 0.00 -.05 -.10 -1.25 -1.00 -.75 -.50 -.25 0.00 .25 .50 .75 (7.) LOCAL RADIAL STRAINzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPON .35 ~------------------~----------------------------------~ TEST 3 .30 --.25 EXTERNAL LOCAL .20 u '> Q. s • 15 . 10 .05 O.O~l~.S-----_-l~.-O-· -----~.-5-----0~.~O------.~5------1~.0------1~.-5-----2~.~0----~2.5 AXIAL STRAIN 357 (7.) ·50,----------------------------- .45 ---- EX TERNAL -------- LOCAL --,zyxwvutsr AX I AL TEST 3zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA SHEAR ... .40 --- ~--...... ---- zyxwvutsrqponmlkjihgfedcbaZYXWVUT /,/'/'/~~ .35 .:zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA ,,, "zyxwvutsrqponmlkjihgfedcbaZYXWVUTS ,, I I .30 , ,, , , , ,, , r .25 ,,/ " .20 .15 ~ _L ~ o ~ 2 ~~ 3 STRAIN .9 ~ 5 4 ~ ~ 6 OD r----------------------------------------------------- EXTERNAL 7 AXI AL TEST --, 3 ----- •B ~, /~ /,/" .7 ,, ,, ,I' , I '0.. 'if ,, , , ,, , , ,, , ,, I .6 I .5 " , , .4 .3 ~------~------~------~------~------~------~~ o 1 2 3 4 STRAIN 358 (7.) 5 6 ~ 7 2.2 ,.... __ - _---TEST 3zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA 2. 1 Ul Ul lJJ c::: ~,... f0Ul ,zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA 2.0 _j <: ...... Cl , I -c -zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA ".tI' """ zyxwvutsrqponmlkjihgfedcbaZYXWVU I c::: 1.9 , I I u.: u, I I , I lJJ I '.J , Ul Ul lJJ EXTERNAL AXIAL ,,I 1.8 ',.... I I / c::: -------- LOCAL SHEAR , 1.7 f0Ul I I I I _j I -c ...... x I I I 1.6 -c I I I u.: u, lJJ 1.5 '.J 1.4 2 0 5 4 3 STRAIN 6 7 OD .200 TEST 3 · 175zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA .150 · 125zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA u '> \:5 · 100 - ::l ~ .075 .050 'O25~ 0.000 0 I ,. 2 4 3 EXTERNAL AXIAL STRAIN 359 5 (1.) 6 7 .8r-----~----------------------------------~ TEST 4 Solid symbol represents .6zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA preshear condition o '~ v SJ .4 .2 - :_.---.-::;:::---- -.2~----~~----~------~------~------~------~ -. 2 0.0 •2 .4 .6 .8 1. 0 zyxwv S'/()~C .30 ~----------------------------~--------------------------~ TEST 4 .25 .20 ---.....•.......... \ .... EXTERNAL LOCAL / zyxwvutsrqponmlkjihgfe zyxwvutsrqponml :/zyxwvutsrqponmlkj /zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQ :.:l •15 u ,'> \5 ..... .10 ...........•.... .j..l ........ -: .' . 05 0.00 i . -.05 -.10 ~ -1. 5 L- ~ ~ -1. 0 -. 5 O. 0 STRAIN 360 ~ .5 O!) ~ 1. 0 ~ 1. 5 .30 TEST 4 .25zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA .20 .15zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA u '> 0 SJ .10 .05 0.00 -.05 -. 10 -.75 -.50 .50 .25 0.00 -.25 LOCAL RADIAL STRAIN .75 1. 00 00zyxwvutsrqponmlkjihgfedcbaZYXW .25 ~ zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLK -r ~ TEST 4 .20 --- EXTERNAL ................ LOCALzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA ........... ,. ............... • 15 ........... u -o '> ..... ... :J <1 .10 :/.... • OS AXIAL STRAIN 361 (;0 .50 r-------------------------------------------------------~ TEST 4 .45 ~~-zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA .40zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA .35 --- EXTERNAL -----~-- LOCAL AXIAL U ,> \j ::-er SHEAR .30 .25 .20 .15~------~------~------~------~------~------~~----~ o 2 3 4 STRAIN 5 6 7 (7.) .9 ~------------------------------------------------~zyxwvutsrqponmlkjihgfedcbaZYXWVUT TEST 4 .8 ----- .7 '0.. 'er .6 --- EXTERNAL -------- LOCAL AX I AL SHEAR .5 .4 .30L-------~1------~2------~3~----4~------~5------~6------~7 STRAIN 0:) 362 2. 2 ,..... U) U) 2. 1zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA TEST 4 W ._ ..zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJ --- 0:: U) 2.0 _J < ...... Cl < 0:: u, u, 1.9 W ....., ...._ ,..... 1.8 U) U) W 0:: ._ --- EXTERNAL AXIAL -------- LOCAL SHEAR 1.7 If) _J < ...... x 1.6 < u.: u, w ....., 1.5 1.4 0zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA 2 3 STRA IN 4 5 6zyxwvutsrqponmlkjihgfedcbaZ 7 (7.) .200 TEST 4 • 175 · 150 • 125 u '> \5 · ] 00 :J <l .075 - .050 V 02S • 0.000 ~ ~ o l' -L ~~ 2 3 EXTERNAL ~ -L 4 5 ~ ~ 6 7 AXIAL STRAINzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPON (I.) 363 .8r-------T-----------------------------------~--__,zyxwvutsrqponmlkjih TEST 5 •6 Solid symbol represents preshear conditionzyxwvutsrqponmlkjihgfedcbaZYXW u '~ u .4 SJ •2 -.2~------~------~------~-------.------~~----~ -.2 0.0 .2 .4zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPON .6 .8 1.0 s'l(j~c .30 TEST 5zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFE .25 .20 ---................ EXTERNAL LOCAL ') .0\ ~~ .. ..::zyxwvutsrqponmlkjihgfedcbaZY zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHG •15 o ,> '6 ,...., ...l ... . ...................... / .o.l . 10 .............,...,..,/ ..... •05 .... . .' 0.00 ..... ... ' ........................................ -.05 -.10 -1. 5 -1.0 -.5 .5 0.0 STRAlN 364 OD 1.0 1.5 .30zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA TEST 5 .25zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA .20 . 15 uzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA ,> Cl . 10 ,..,J .05 0.00 -.05 -.10 -.75 -.50 .50 .25 0.00 -.25 .75 1. 00 (r.) LOCAL RADIAL STRAINzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHG .25 ~---------------------------r--------------------------~ TEST 5 ----- .20 EXTERNAL LOCALzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA · 15 u ,> - '6 · 10 :J <1 • OS ....... -.05 L- -1.5 ~ ~ -1.0 -.5 L_ ~ 0.0 AXIAL STRAIN 365 .5 Cr.) ~ __ ----~ 1.0 1.5 .50 ~------------------------------------------------------~zyxwvutsrqponmlkjihgfedcbaZ TEST 5 .45 .40 ---EXTERNAL AX1AL .35zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA u ,> I::) -------- LOCAL SHEAR ~ --J_ __ '0=" .30 .25 .20 .15 ~ o 2 -----L-------L 4 3 STRAIN ~-- 5 ~ --~ 6 7 (7.) .9 ~------------------------------------------------------~ TEST 5 .8 .7 ,'0- c- .6 ----- EXTERNAL -------- LOCAL AXIAL SHEAR .5 .4 .3 ~ o ~ J_ f 2 -L ~ 4 3 STRAI N 366 (7.) ~ ~ __----~ 5 6 7 2.2zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJI r-. TEST 5zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONM 2. 1 zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA tn _ ...... - ..... ----- U1 W ~ l- zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA 2.0 tn _J -< ...... D -c 1.9 ~ u: -... u, W '-J 1.8 --- EXTERNAL -------- LOCAL AXIAL r-. U1 tn SHEAR W ~ 1.7 IU1 _J -c ...... x -< 1.6 w '-J 1.5 u:u, 1.4 3 2 0zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA 5 4 7 6 STRAlN GO .200 TEST .175 5 .150 .125 u ,> '6 .100 :J <I .075 - .050 .025 0.000 0 1. 2 5 4 3 EXTERNAL AXIAL STRAIN 367 GO 6 7 .B~------~--------------------------------------~ TEST 6 Soljd symbol represents preshear condjtjonzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHG .6 -.2~------~------~------~------~------~----~ .8 1.0 .6 .4 0.0 .2 -.2zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA s/C5vc .30 TEST 6 .25 .20 EXTERNAL ................ LOCAL . 15 u ,> o ,~ .. ' .' ................. •10 .'... .... .......•........ ...' ... •05 . . ! 0.00 -.05 -.10 ~ -1. 5 ~ -1. 0 ~ -.5 -L ~ 0.0 .5 STRAI N 368 (7.) ~------~ 1.0 1.5 .30 TEST 6 .25 .20 · 15zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA uzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA ,> \:) ....-:;:r · 10 .05 0.00 -.05 -.10 -.75 -.50 -.25 .25 0.00 .50 .75 1. 00 LOCAL RADIAL STRAINzyxwvutsrqponmlkjihgfedcbaZYXWVUT (r.) .25 r---------------------------~--------------------------~ TEST 6 ----- .20 EXTERNAL ................ LOCAL •15 u - '> \:) .10 :J <I .05 ~------~--------~--------~------~--------~------__j -.5 -.05 -1. 5 -1. 0 .5 0.0 AXIAL STRAIN 369 (7.) 1.0 1.5 r-----------------------------------------------------------~zyxwvutsrqp .50 TEST 6 .45 ------ ------ .40zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA .35 U ,> R_ --- EXTERNAL AXIAL -------- LOCAL SHEARzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQ 'er .30 .25 .20 .15 ~ L- o ~ 2 _L ~ 3 ~ 4 ~ 5 ~ 6 7 STRAINzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPO (r.) .9 ~-----------------------------------------. TEST 6 .8 -_zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA .. -- .7 ,_ '0.. er .6 ---- EXTERNAL AXIAL -------- LOCAL SHEAR .5 .4 .3 ~----~------~------~------~------~------~----~ o 1- 2 3 4 STRAIN 370 (i0 5 6 7 2. 2zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA ,....zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA 2. ] TEST 6zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA tn tn ---- -----zyxwvutsrqponmlkjihgfedcbaZYXW W Cl:: l- tn 2.0 ....J -c ,_, 0 < 1.9 Cl:: u, u, W """ ........... 1.8 ,.... tn tn W Cl:: l- 1.7 tn --- EXTERNAL AX 1 AL -.:------ LOCAL SHEAR ....J < ..... 1.6 x -e u, u, W 1.5 """ 1.4 5 4 3 2 0 7 6 STRAIN OD .200 TEST 6 •]75 · ]50 .125 u '> - b · ]00 :J <1 .075 .050 .025 0.000 0 2 :3 6 4 EXTERNAL AXIAL STRAIN 371 (1.) 7 .8zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA TEST 7zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONML Solid symbol represents preshear condition .6 U ,> .4 R ....~ -.2L-----~~----~------~------~------~------~ 1.0 .8 .6 .4 .2 0.0 -.2 .30 TEST 7 .25 EXTERNALzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA ................ LOCAL .20 i::) u ,> '6 • 15 ~ .. ... ...... ...../ . 10zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA ......... . 05 .................. AXIAL STRAIN 372 (7.) . ' .30 ~---------------------------r---------------------------. TEST 7 .25 .20zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA ,u '> \5 .15 +J .10 .05 0.00zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIH L~ ~ ~~ _L ~----~~~~~~--~ -.500 -.375 -.250 -.125 0.000 .125 .250 .375 .500 LOCAL RADIAL STRAIN (7.) .200 .175 TEST 7 .150 --- EXTERNAL ................ LOCAL .125 u ,> o .100 ::J <I .075 .050 .025 0.000 ~----_.------~---~-----~------~----~ -1.00 -.75 -.50 -.25 0.00 AXIAL STRAlN 373 .25 (7.) .50 ~ .75 ~ 1. 00 .50zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA TEST 7 .45 ---zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDC ~~~~ ,_, , , ,, , , .40 , , ,zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONML ,, I I b .... er EXTERNAL AXIAL .35 u '> -------- LOCAL SHEAR zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA .30 .25 .20 .15 ~------~------~--------~------~------~------~~----~ 4 3 2 o STRAIN 5 7 6 GO .9 ~----------------------------------------------------------, TEST 7 •8 ...... ,~, •7 ,, I I , ~~~~ ~~~~ ,- "" ... " I I ! '0... ' U=" •6 , EXTERNAL AXIAL I -------- LOCAL SHEAR .5 .4 1·' 2 4 3 STRAIN 374 (r.) 5 6 7 2.2 ,... 2. 1zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDC TEST 7 lJ) lJ) W 0:: I- zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA ~~~ 2.0 ~~~zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDC -c ~~ ~~ lJ) _J .. ,,< II' -: 0 < 1.9 0:: , ,, ,,zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA WzyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLK u, u, -'-J 1.8 ,... EXTERNAL AX1AL I' I I lJ) lJ) W 0:: I- ,/ ,, 1.7 -------- LOCAL SHEAR lJ) _J < ...... 1.6 x < u, u, w 1.5 '-J 1.4 2 0 3 4 STRAIN 5 6 7 5 6 7 (7.) .200 TEST 7 .175 .150 .125 u ,> - '0 • lOO ::l <l .075 .050 .025 0.000 0 1· 2 3 4 EXTERNAL AXIAL STRAIN 375 (i.) .8~------~------------------------------------~zyxwvutsrqpo TEST 8 Solid symbol represents preshear condition .6 u '~ v ~.jJ .4 .2 _.2~ -. 2 ~~-O. 0 __ _J -L__--__~-- __--~ --~-- .2 .4 •6 •8 1. 0 s1cr~c .30 ~----------------------------r---------------------------~ TEST 8 .25 ---- EXTERNAL ................ LOCAL .._ ...... ) .20 ...... u ,> o ,:; .15zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDC ................... . 10 ..... (... ' . 05 ....... 0.00 ~----~------~------~------~----~------~------~----~ -.125 0.000 . 125 -.375 -.250 -.500 AXIAL STRAIN 376 (;0 .250 .375 .500 .30 r-----------------------------.-----------------------------, TEST 8 .25 .20zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA o ,> '0 .15zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA ... +J .10 .05 0.00 ~ -.2500 ~ -.1875 ~ -.1250 ~ -.0625 L- 0.0000 ~ .0625 LOCAL RADIAL STRAIN .100 ~ ~ .1250 .1875 ~ .2500 (7.) ~----------------------------r-----------------------------. TEST 8 .075 EXTERNAL ................ LOCAL ......... ................ .050 .. .....' / ' ...... - o ,> .. .. '0 .025 ' <I 0.000 .. ' zyxwvutsrqponmlkjihgfedcbaZYXWVUT ...........zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHG .. ,0"" .. ) .",:' ' :J .. ........ ,..:/ ........ ' -.025 -.050~~--~~----~~----~~--~~~--~~----~------~----~ -.500 -.37~zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA -.250 -.125 0.000 .125 .250 .375 AXIAL STRAIN 377 (7.) .500 ·50 ~----------------------------------------------------------,zyxwvutsrqponmlkjihgfed TEST 8zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA ---- ,,-zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDC .45 " """zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA ," I I .40 I I I .35 ---- EXTERNAL -------- LOCAL AXIAL SHEAR .30 .25 .20 .15 ~------~------~------~~------~------~------~------~ o 2 345 STRA 1N .9 6 7 (7.) r-----------------------------------, TEST B •B .7 '0- '"C? 16 -------- LOCAL SHEAR .5 .4 1" 2 3 4 STRAIN 378 (7.) 5 6 7 2.2zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA ,.....zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA TEST 8 2. 1 UJ UJ lLJ et:: IUJ 2.0 ,,,,,- .. ;-' ~~ ~~~~ _J -c ...... 0 -c /,//~~~ 1.9 et:: ,zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPO .: zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDC l.J... u, L1J - 1.8 et:: 1.7 'oJ ,..... IUJ " /' I EXTERNAL AX1AL I UJ UJ L1J I ,/ -------.LOCAL SHEAR ...J -e ...... x 1.6 l.J... l.J... lLJ ....., 1.5 -c 1.4 5 6 7 4 3 2 0zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDC STRAIN (i.) .200 TEST 8 .175 .150 .125 u ,> I:) • lOa :::> <1 .075 .050 .025 0.000 0 2 5 4 3 EXTERNAL AXIAL STRAIN 6 (1.) 379 UNIUERSI iV UI i»Umtl:Y zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA U R 7