0612046: M.Eng. Dissertation

An Investigation into the Undrained Shear Strength of Cohesive Materials.

A Dissertation

by

Inderjit Sidhu

A dissertation submitted in fulfilment of the requirements for the degree of Master of Engineering,

Brunel University.

Department of Engineering & Design

Brunel University of West London

March 2010

Major Subject: Civil Engineering with Sustainability

Declaration of Originality

This is to confirm that the work done for this project is entirely my own and not of any other person,

unless explicitly acknowledged (including citation of published and unpublished sources). This work

has not previously been submitted in any form to Brunel University or to any other institution for

assessment for any other purpose.

Signed _________________________________________________

Date ___________________________________________________

Acknowledgements

I would like to thank Supervisor Dr Phil Collins, Malcolm Austen, and Paul Szadorski. I have benefited significantly from their guidance, support and undivided attention throughout the completion of my project. Secondly, I would like to thank and all those people who gave some of their time to suggest modifications and improvements.

Thirdly, I would like to thank Simon Burke, Joanne Sturges, John H Masters, and all the rest of the staff at GeoLabs, BRE Watford for assisting to complete the tests.

Inderjit Sidhu, March 2010.

Nomenclature

Term Symbol Units

Total Major Principal Stress σ1/p1 kPaTotal Minor Principal Stress σ3/p3 kPaNormal Stress σn kPaEffective Major Principal Stress σ1’ kPaEffective Minor Principal Stress σ3’ kPaMaximum Deviatoric Stress (σ1 – σ3) kPaPore – Water Pressure u kPaShear Strength (Cu) kPaShear Strength at Failure f kPaNormal Stress at Failure σf kPaUndrained Shear Strength cu kPaCohesion Intercept (Total Stresses) c kPaCohesion Intercept (Effective Stresses) c’ kPa

Angle of Shearing Resistance (Total Stresses) ϕ degrees

Angle of Shearing Resistance (Undrained – Total Stresses) ϕu degrees

Angle of Shearing Resistance (Effective Stresses) ϕ’ degrees

Angle of Shearing Resistance (Total Stresses) ϕ* radiansMidpoint MP kPaRadius R kPaShear Stress produced in the Soil cmob kPaApplied Bearing Pressure σmob kPaFounding Breadth B mFounding Depth D mUnit weight of Soil removed γ kg/m3

Ultimate Bearing Capacity qf kPaAllowable Bearing Capacity qa kPaNet Bearing Pressure  qn kPaBearing Capacity Coefficient Nc

Bearing Capacity Pressure Factor Nq

Bearing Capacity Density Factor Nγ

Skempton’s Bearing Capacity Factor (Undrained) Ncu

Shape Factor sc

Depth Factor dc

Safety Factor Fs

Triaxial Compression TCConventional Triaxial Compression CTCQuick Undrained QUUnconsolidated Undrained UUConsolidated Undrained CUConsolidated Drained CD

List of Tables

3.1 Specimen description of densities, mass and moisture content

4.1 Summary of Experiments performed

5.1 Determination of Shear Strength and Normal Stress at Failure

5.2 Undrained (Immediate) Shear Strength of Cohesive Soils

5.3 Presumed Bearing Values under Static Loading

List of Figures

1.1 Stability of shallow foundation or excavation in soft clay (Berre, 1981)1.2 Development of Shear Failure beneath a Foundation (Coduto, 1994)2.1 Estimation of the Friction Angle2.2 Plastic Deformation Mechanism (Chen, 1975)2.3 The three modes of Bearing Capacity Failures (Das, 1995)2.4 (a)Triaxial Setup and (b) Typical Stress Paths or Triaxial Tests (Wood, 1990)2.5 Qualitative comparison of shear strength results for UU, CU and CD tests2.6 Results of a triaxial test on specimens of a homogenous soil.2.7 Mohr circles showing stress state of three different specimens of the same........................ cohesion-less soil (c = 0) when subjected to different confining pressures.2.8 Mohr circle determinations using combinations of axial and confining pressurefor non-granular materials (Vickers, 1983).2.9 Mohr circles for the Total and Effective Stress for tested materials (Vickers, 1983)2.10 Mohr circles comparison between Theoretical (a) and ......................................................Actual Results (b) for most soils (Vickers, 1983)2.11 Illustration of the stresses and forces acting on a CTC subjected specimen2.12 The confining, shearing, and failure modes of a specimen under Triaxial ........................Compression (Vickers, 1983)2.13 Failure plane of a specimen under TC (Vickers, 1983)2.14 Triaxial compression test showing test pressure and assumed plane of failure of AB.......(Duncan, 1998)3.1 (a)Buff School Clay in its natural form. (b) Stoneware Clay shown as large clump and ....clusters3.2 (a)Specimen on weighing scale. (b) Two specimens of both clays correctly ......................weighed3.3 (a) 2.5 kg Rammer for Compaction. (b) 1 L Mould Cylinder to 0.01 g3.4 (a) Sequence of first four impacts. (b)Sequence of successive blows after first four ..........(3.3, BS 1377, 1990)3.5 (a) Crucible containers used to weigh specimens. (b) Balance readable to 0.01 g4.1 Loading cell with Pressure Gauge and Loading ring; used for the confining pressure tests4.2 The Brunel VJTech Advanced Triaxial testing system (not functioning)4.3 (a) A Latex membrane stretched along the inside of the mould. (b)Compacted specimen...within the mould 4.4 A prepared specimen in the cell ready to be tested4.5 General setup of Quick Undrained system ……………………………………………….(VJTech Soil and Rock Testing Manual, 2009)4.6 (a) General features of Wykeham Farrance Triaxial System. (b) Loading Cell5.1 Graphs of Shear Strength over Confining Pressure of Specimen C_X1 (a) and specimen...C_X2 (b)5.2 Graphs of Shear Strength over Confining Pressure of Specimen C_Y1 (a) and specimen...C_Y2 (b)5.3 Annotated Mohr Circles of Specimen C_X1 and Specimen C_X25.4 Annotated Mohr Circles of Specimen C_Y1 and Specimen C_Y25.5 Mohr’s circle to determine the shear strength τ f and the normal stress σ f at failure5.6 (a) Deformation changes at failure of C_X1 specimen. (b) Deformation changes at failure of C_X2 specimen5.7 (a) Deformation changes at failure of C_Y1 specimen. (b) Deformation changes at failure of C_Y2 specimenA.1 Quick Undrained Triaxial test results for X1 Clay SpecimenA.2 Quick Undrained Triaxial test results for X2 Clay SpecimenA.3 Quick Undrained Triaxial test results for Y1 Clay SpecimenA.4 Quick Undrained Triaxial test results for Y2 Clay SpecimenA.5 Quick Undrained Triaxial summary results for all Clay Specimens

Abstract

This study will look at different cohesive soils working in axial compression, observing this behaviour, and its effect on the undrained shear strength. An investigation was lead into the additional parameters affecting test results; for this purpose, smaller samples of soils are acquired in order to conduct characteristic analysis of specimens prior to experimentation. The experimental work was carried out on a Triaxial Compression apparatus at GeoLabs, BRE Watford testing centre. The triaxial device was able to contain a cylindrical soil sample of height 76 mm, diameter 38 mm, where these tests were co-ordinated with the staff at GeoLabs. Modifications to the apparatus to perform a multistage undrained test resulted in a significant improvement in the performance of the apparatus and a better representation of soil stresses.

The relationship between the undrained shear strength and the moisture content of cohesive soils considers soil type and size of clay grains on the moisture content at uniform shear strength. Previous studies have examined different soils, mainly sands and gravels, but not to great depth, whilst a lack of focus is given to the impact of confining pressure on foundations built in clay. This paper focuses on an investigation of the relationship between moisture content and undrained shear strength in clay samples of different structures at different confining pressures, and their workability in shallow foundation construction. This study examined the effect of this behavioural – interaction by varying bulk density of the clays using a compactive effort, choosing a set of appropriate confining pressures for number of clays chosen for their degree of variability i.e. silty to firm clays. These included additional properties of moisture content, particle size, composition (coarse – grain soils), compaction, and structure. All tests were carried out on pure samples of well – graded uniform laying clays with different cohesive properties. Pressures subjected on the clay samples were two sets of three different confining pressures, and this study developed relationships of behaviour to their reaction of increasing pressure.

Stiff brown silty clay and firm grey clay were tested; were compactive effort obtained the required bulk densities. The tests revealed that as the confining pressure increases, the undrained shear strength increases. In addition, an increase the moisture and degree saturation produces a decrease in the angle of shearing resistance angle (internal friction angle) and an overall decrease in undrained shear strength. The deviatoric stress to affected the strain behaviour of the clays; however, a specific trend was not found. These types of tested clays were inappropriate in to which level a shallow foundation. Additionally, the extent of settlement occurring when foundation loads applied to the ground is dependent on the structure rigidity, duration and type of loading, and deformation characteristics of the ground soil. The study showed that consolidation settlement in clays and silts occur for long periods after the structure is finished. The drainage rate from the voids under stresses influence applied is slow; therefore, consolidation settlements must be considered. The shape of the specimens did influence the overall failure mode response and demonstrate elastic distortion (barrelling/bulging) in conjunction with general shear and local shear failure with the axial loading rate and confining pressure.

Table of Contents

Declaration of OriginalityAcknowledgementsNomenclature List of Tables List of Figures Abstract

CHAPTER 1...............................................................................1INTRODUCTION.........................................................................1

1.1: Study Background.........................................................................................................11.1.1: Shallow Foundations..........................................................................................2

1.2: Chapters Overview........................................................................................................41.3: Aim and Objectives.......................................................................................................5

CHAPTER 2...............................................................................6LITERATURE REVIEW.................................................................6

2.1: Introduction....................................................................................................................62.2: Friction...........................................................................................................................62.3: Bearing Capacity of Shallow Foundations....................................................................72.4: Soil Testing....................................................................................................................9

2.4.1: The Triaxial Test................................................................................................92.4.2: The Vane Shear Test..........................................................................................9

2.5: Drained & Undrained Conditions................................................................................102.6: Mohr – Coulomb Criterion..........................................................................................112.7: Shear Stress Parameters...............................................................................................142.8: Loading and Stress Parameters....................................................................................172.9: Recent & Past Advances in Triaxial Testing...............................................................18

2.9.1: Equipment........................................................................................................182.9.2: Methodology....................................................................................................182.9.3: Errors and Interpretation..................................................................................192.9.4: New Test Varieties...........................................................................................19

CHAPTER 3.............................................................................20MATERIAL CLASSIFICATION & PREPARATION.............................20

3.1: Introduction..................................................................................................................203.2: Material Description....................................................................................................203.3: British Standards for Experimental Work...................................................................213.4: Preparation of a Remoulded Specimen........................................................................22

3.4.1: Compaction Method for Soils using a 2.5 kg Rammer....................................223.4.1.1: Compaction Criteria.........................................................................223.4.1.2: Compaction Equipment...................................................................233.4.1.3: Compaction Procedure.....................................................................23

3.5: Determination of Moisture Content of Specimens .....................................................253.5.1: Oven – drying Method.....................................................................................253.5.1: Procedure.........................................................................................................25

CHAPTER 4.............................................................................27EXPERIMENTAL WORK ............................................................27

4.1: Introduction..................................................................................................................27

4.2: Equipment Description................................................................................................274.3: Preparation of Specimen into Loading Cell.................................................................284.4: Equipment Preparation................................................................................................294.5: Testing Procedure........................................................................................................30

4.5.1: Multistage Quick Undrained Procedure...........................................................304.6: Tests Performed...........................................................................................................33

CHAPTER 5.............................................................................34EXPERIMENTAL RESULTS.........................................................34

5.1: Introduction..................................................................................................................345.2: Data Obtained..............................................................................................................345.3: Effects of Confining Pressure and Undrained Shear Strength.....................................345.4: Specimen Mohr Circle Analysis..................................................................................395.5: Deformational Changes at Failure...............................................................................425.6: Effects of Moisture Content ........................................................................................455.7: Theoretical / Actual Results Comparison ...................................................................465.8: Degree of Saturation on Shear Strength .....................................................................46

CHAPTER 6.............................................................................47CONCLUSIONS & RECOMMENDATIONS......................................47

6.1: General Conclusions....................................................................................................476.1.1: Confining Pressure Modification.....................................................................476.1.2: Triaxial Testing Results...................................................................................476.1.3: Shear Stress & Normal Stress at Failure..........................................................476.1.4: Saturation, Moisture Content and Shear Strength............................................476.1.5: Specimen Shape...............................................................................................47

6.2: Recommendations........................................................................................................486.2.1: Recommended Changes in Testing Procedure................................................49

6.3: Future Work.................................................................................................................496.4: Summary......................................................................................................................49

APPENDIX A............................................................................51Specimen Test Results........................................................................................................51Specimen Test Summary Sheet..........................................................................................55

READING LIST.........................................................................56

CHAPTER 1 – INTRODUCTION

1.1 Study Background

This study examines the characteristics of cohesive materials, focusing on undrained shear strength. A cohesive material is an accumulation of particles sticking together through moisture; the mechanical behaviour of cohesive materials is a function of the materials particle size, shape, particle-to-particle friction, their arrangement, saturation, moisture, and the associated pore voids. Especially with in – situ cases, there must also consideration on the loading and confining pressure acting on the material. The deformation in cohesive materials is due to external forces; these create internal matrix changes caused by particle interlocking, matrix suction, sliding and rolling. Understanding these material responses can largely influence the importance of design of structures such as foundation systems, pavements constructed with clay, natural earth dams, and the improving the stability of the foundations in construction. These kinds of materials used in the development of shallow/deep foundation construction, landscaping and design, where shallow foundations must follow through certain soils in order to be less influenced by the behaviour of the soil. The analyses of these structures is based on the strength and deformational behaviour of the material adjacent or beneath to the foundation system. High strength soils, granular and cohesive materials will be able to work with the foundation it surrounds, decrease the chance of rapid settlement, and in the hope to support a structure without failing. Many studied reports of soil – interaction in literature, and the vast majority of these are concerned with the use of interaction and properties under loading. Understanding this develops measures to improve ground conditions, offer insight into working ground materials and attempts to understand the nature of soils, particularly clays and sands. For example, the economic and successful application of new shallow foundation depends on the soils immediate bearing capacity, as well as supporting the weight of slabs; the foundation provides a level on which to build. A structures weight keeps a foundation in place, in the case of taller buildings; foundations anchor a building to the ground. A properly designed foundation or gravity – based structure will limit settlement, i.e. the tendency of sink. The requirement of foundations is that it has to be working in coordination with the material beneath it. This can be shown in Figure 1.1 that demonstrates the zones of soil affected by foundations.

Figure 1.1: Stability of shallow foundation or excavation in soft clay. (Berre, 1981, [5])

Builders and non – professionals have consistently documented the requirement of good foundations. A foundation is an element of a structure that works in coordination of a structure to the adjacent zone of soil/rock beneath it. The main principle of having a foundation is to transfer the structure loads to the soil/rock underlying with no overstress exerted on the soil or rock.

There are three basic requirements for a foundation to be satisfactory (Sowers, 1962, [36]):

A properly situated foundation should respect any future influences, which could affect performance.

A foundation, including the soil below it, must be stable and safe from failure. A foundation must not settle adequately to damage the structure.

1.1.1 Shallow Foundations

Foundations have two main categories, shallow, and deep. This part of the study focuses on shallow foundations rather than deep foundations. Shallow foundations are common in construction when the soil formation has sufficient strength for a safe bearing support. If the soil is very compressible, or has minor shear strength, shallow foundations are sometimes inadequate to these problems. The loads applied transmit to greater depths then to a stiffer layer. A necessity of foundations is the ability to carry structural loads without movement to cause damage. Most soils supporting foundations do not overstress beyond their limit strength, as the deformation caused by this compressed soil cannot be extreme. The pressure that a soil can withstand without shear failure (overstress) is the soil bearing capacity (McCarthy, 2007, [32]). The maximum bearing capacity of soils for a foundation relates to the soil properties and foundation characteristics. Three principal failure modes of soil exist, defined as general shear failure, local shear failure and punching shear failure, of which Figure 1.2 demonstrates.

Figure 1.2: Development of Shear Failure beneath a Foundation (Coduto, 1994, [10])

General shear failure is inclined to have a brittle stress – strain relationship. A distinct area of wedging is under the foundation where slip surfaces elongate diagonally from the footing edges down through the soil then up toward the ground surface. This is where the ground surface bulges upward and displacement is visible by foundation. Punching shear failure involves plastic material characteristics. There is important compression that occurs beneath the foundation and vertical shear occurs under the foundation edges. Zones beyond these edges are not generally affected and no real bulging occurs. Local shear failure involves general shear and punching shear characteristic failure modes. A distinct wedge and slip surface is formed below the foundation and slight bulging occurs at the ground surfaces. Local shear states represent an intermediary condition between general and punching shear failure (McCarthy, 2007, [32]). Shallow foundation design considers general shear failure to happen in dense granular soil and in firmer saturated cohesive soils subjected to undrained loading. Punching shear case usually applies to compressible soil (sands) that have low to medium relative densities, and for cohesive soils subjected to slow axial loading (McCarthy, 2007, [32]). Shallow foundations transmit the loads of the structure to the adjacent soil under it. The types of shallow foundation available are pad foundation, strip foundation and raft foundation.

Pad Foundation

These support column point loads. There are different types of pad foundations accessible, like mass –concrete for steel column, plain reinforced concrete, and balanced pad foundations. Shallow mass – concrete pads are for varying states of soil layers where appropriate load bearing soils exist at shallow depths (Curtin, Shaw, Parkinson & Golding, 1994, [13]). Deep mass concrete pads (cast at depths of 1.5 m – 2 m) are used when piling substitutes are more unrestrained. Shallow reinforced concrete pads are like mass – concrete pads except of a smaller thickness because of the practice of reinforcement on the tensile face of the pad that improves resistance to bending moments.

Strip Footing

Strip footings lie under consistent point loads. Strips distribute load concentration indirectly into an increased thickness of sub – strata to reduce the settlement and bearing stress to a limit (Curtin, Shaw, Parkinson & Golding, 1994, [13]). The structure will distribute load into longitudinal directions when the loading is non – uniform. Strip width is chosen according to the bearing stress limit and bucket size of the excavator (Curtin, Shaw, Parkinson & Golding, 1994, [13]).

Raft Foundation

Raft foundations spread the structural loads over a larger area to reduce the bearing pressure, as it is more rigid and lowers the potential for excessive differential settlements. Rafts are heavier in weight, able to resist higher uplift loads, and distribute lateral loads into soils more consistently (Curtin, Shaw, Parkinson & Golding, 1994, [13]). Sensible foundation design offers suitable safety factors for shear failure of the soil and excessive settlement.

Once the strength of the soil is determined through laboratory shear testing, provided all data regarding the in – situ conditions are sufficiently obtained, the strength can be used to determine the Bearing Capacity on that same soil for a specific shallow foundation. The maximum load that a foundation can support may be calculated using bearing capacity theory. In the case of preliminary design, there can be theoretical bearing values (presumed) used to indicate and estimate the pressures that would normally result in a satisfactory safety factor. The Ultimate Bearing Capacity (qf) is the bearing stress value that causes a sudden settlement of the foundation due to shear failure. The Allowable Bearing Capacity (qa) is the ultimate bearing stress applied to a foundation that is safe against instability by shear failure,  and the non – exceed maximum tolerable settlement. The allowable bearing capacity derives from the ultimate bearing capacity using a safety factor (Fs). The net bearing pressure (qn) is the rise in soil stress. Excavation for a foundation relives the stress at founding level when the weight of the soil removed.

Most soils contain clay, sand, gravel, air and water. Exerted loads on soils permit consolidation, and air and water drain out. Non – cohesive soils experience consolidation during construction phases. In cohesive soils, clays and silts, this occurs over years or rapidly depending on the rate of consolidation and material. Sand or gravel based materials are heavily explored for these kinds of analyses. Nonetheless, there are many questions concerned to the basic understanding of the shear failure and bulging/barrelling phenomena affects in cohesive materials such as clays. The current literature lacks an experimental analysis that can address these concepts.

Water within the construction unfavourably affects the performance of foundations. The soil moisture content can affect the shear strength and as a result the bearing capacity. The execution of design and construction of foundations are with the intention of keeping the construction saturated to a limit to prevent slip and sinking. A water source can usually range from direct precipitation or through surrounding water. Casagrande & Shannon, 1951, [8] recognised two significant sources of water from which base courses in foundations can become saturated. The first being frost action and infiltration through the ground soil. This was initially based on theoretical analysis of crack width and

assumption of laminar flow through the base; they second source eventually concluded that base courses could speedily become saturated through surface cracks upon rainfall. A better understanding of this type of behaviour performed through laboratory tests where there is adequate replication of in – situ stress conditions and loads. The two most common systems used to test cohesive soils are the Triaxial Compression Test and the Unconfined Compression Test.

Determining soil stability is one of the few important and challenging aspects of geotechnical engineering. The concern, including past knowledge, on soil stability has driven the important advances and constant renewed understanding of the intricate behaviour of soils. This has also lead to improvement of most testing equipment used to investigate soil behaviour. Broad research has gone into the study of providing an agreement of soil mechanics principles and the approach to solve practical soil problems regarding stability (Duncan & Wright, 2005, [17])

These advances matured into the evaluation and prediction of ground, slope stability, foundation, pavement, runway and road design, and excavation in soft clay, here experience, judgment, and rational methods have combined to better understand and improve the level of achievable confidence in the use of observation, testing, and analysis. With these numerous advances, effective prediction of soil stability still requires essential judgment, as it remains a difficult field. The problem is that even when geology and soil conditions have been evaluated and keeping with good practice standards, the overall stability has been assessed with methods that been efficient in previous projects. It is a problem because it without proper understanding of behaviour, the theory behind stability is unjust it is hard to determine the reasons for failure.

For the vast majority of structures built upon certain materials (cohesive and cohesion-less soils) the main consideration is that of foundation loads, shallow or otherwise, and their compressive short/long – term effect on the type of soil below and how this possibly affects performance. However, with the specific techniques to improve materials that used to support loads, there is uncertainty with respect to the forces leading to failure and their magnitudes. In the case of integrated soil foundations there is a great deal of ambiguity in ignoring the type of existing soil on site and the overall contribution it may have on the structure, even if some tests warrant a good practical use of the soil. It is apparent therefore that research is required into the particular area in which an improved assessment in soil stresses regarding the type of common soil used on site and the possible reasons why that would benefit/adverse the type of structure built upon.

Over time, there has been significant experience with the behaviour of shallow foundations. With respect to their function, this has led to the development of better understanding of changes in soil properties that can occur over time, the importance of the requirements, limitations of laboratory, in – situ testing for evaluating soil strengths, and development of more effective types of instrumentation to predict potential soil behaviour. This is in coordination with improved understanding of the principles of soil mechanics related to overall soil behaviour, improved analytical procedures amplified by broad examination of the mechanics (large scale testing), detailed comparisons with field behaviour, and finally with the use of computers to perform analyses. A better understanding of cohesive behaviour performed through laboratory tests is necessary where there is effective reproduction of in - situ stress conditions and loads.

1.2 Chapters Overview

This study begins with a concise literature review of the equipment used to measure shear strength, the criterion used for failure and analysis, failure modes, drained and undrained conditions and their relevance to soil testing, theory of plastic/elastic mechanisms in shallow foundations, loading and stress parameters, internal soil friction, and strength properties are presented in Chapter 2. Chapter 3 will discuss the preparation and methods to change the density of the specimens, compaction, to determine the bulk and dry density, moisture content, relevant Eurocodes and standards along with a description of the cohesive soils tested. This study performed undrained soil strength testing behaviour using the Wykeham Farrance 38 mm diameter Triaxial Testing System at GeoLabs, BRE Watford. In which observations were made of the specimens during compactive and actual testing procedures. This includes the additional help of the Brunel University VJT5110 TriSCAN100

Advanced Automated Triaxial Testing System. The Brunel device itself is not functioning but the explanation of the system provides understanding of the process. This apparatus allows three kinds of tests to be performed, of which can be performed on cohesive cylindrical soil specimens to be encased in rubber membranes with confining pressure present, then the soil specimen in the chamber, under chosen lateral pressure, is subjected to an increasing axial load until the specimen fails and is described fully in Chapter 4. Chapter 4 introduces the specific test and criteria used for actual experimentation, describing details of the specific triaxial compression system used in this experimental study, with specimen preparation method, testing procedure, along with a description of tests to investigate cohesive soil behaviour.

Chapter 5 is the representation of results and analysis of test data, these being the undrained shear strength, maximum deviator stress, angle of shearing resistance (angle of internal friction), the shear and normal stress at failure, and apparent cohesion. This section describes the confining stress to the shear strength response of the specimens, followed by the discussion of the impacts of compressive load, confining pressure, deformational changes (specimen shape and failure modes), comparing theoretical and actual results, and the effect of moisture and saturation. Each specimen prepared will have a degree of variability and then compared with those of other specimens tested in order to establish the effectiveness of the experiment, repeatability and reliability of the results. This will include detailed comparison between the clays in order to determine their parameters.

Chapter 6 is essentially a concluding review of what was determined, the behavioural interaction of the specimens, and the parameters used in them. Discussion into varying the effect of confining pressure, effect of moisture and saturation, specimen shape, values of shear and normal stress on each sample, and reliability of triaxial tests. Additionally, Chapter 6 describes the potential benefit of soil shearing resistance values in wider applications. Recommendations and improvements for future research, future work in the same geotechnical field, coordinating with data from literature, laboratory and field are also compared with the proposed method to provide an independent check. This will additionally include changes in experimental work with what possible improvements to future to ensure results that are more reliable. This chapter also consists of a summary of the research findings and the general contribution this study has made. In addition, with documented conclusions derived from the study.

1.3 Aim and Objectives

The real aim of this study is to demonstrate which of the selected cohesive soils provide the greatest shear strength and the lowest shear strength when subjected to different confining pressures and whether the effect of moisture does play a part in the overall undrained shear strength. Additionally to what extent this relationship plays in shallow foundations. To develop a means of determining the relationship between undrained shear strength and confining pressure from a single specimen subjected to a set of confining pressures, whilst better understanding overall cohesive soil theory, shear strength properties, and confining pressure in certain in – situ conditions. Additionally with an observation on how specimen shape affects the overall failure mode of a cohesive material. This should discuss which cohesive materials are appropriate in shallow foundation design, provide useful data for immediate bearing capacity, perhaps road, and clay – pavement design. Especially for sites under excavation with soft clay and receiving channelised traffic like construction vehicles.

CHAPTER 2 – LITERATURE REVIEW

2.1 Introduction

This chapter comprises of a literature review conducted on the soil shear tests, particularly their potential and constraints, representation of stresses in soils, as this is required in order to provide an appropriate level of understanding for the successful application of the shear test apparatus. Additionally, this also consists of the latest advances made in the field of soil testing primarily focussing on the parameters of the Triaxial Test. Added documentation on the behaviour of cohesive materials under shallow foundation loading. Additionally presented are the factors affecting the determination of these properties, with description of types of experimental determination of these same properties, and concluding on recent test procedure developments. This section also mentions the criterion used to determine the current existing models that are widely used with a summary of the literature of the review finalising this chapter.

Clays are accountable for a large percentage of problems with stability due to complex water – interaction. Certain strength properties of clays are complex and are subject to changes over time through consolidation, permeability, swelling, weathering, slickensides, and creep. Undrained strengths of clays are important for short – term loading conditions, and drained strengths are important for long – term conditions. In nature, most soils and rocks are viscous – elastic materials. In current literature, there are developed theories that are usually categorised into the linear viscous – elasticity field, while soils and rocks are known to have highly nonlinear stress – strain behaviour with a known dependency on permeability and time. Consequently, the time – independent elastic – plastic theory is commonly used to describe the stress – strain relationships of natural materials; the material is said to be linearly elastic up to the yielding point, after this it is achieves a perfectly plastic condition (Holtz and Kovacs, 1981, [25]). In other cases, some materials are known to be brittle and demonstrate a small amount of stress when under strain, this is usually the case for rock materials; whereas others are known to demonstrate work – hardening (i.e. compacted clays, and some loose sands) or even the opposite, known as work – softening.

2.2 Friction

Determining the shear resistance between any two particles is by measuring the force needed to cause movement between the particles. The two most common methods of expressing frictional resistance are to use the friction coefficient, or the friction angle. Figure 2.2 can help explain this friction system. The friction angle ϕu (undrained), can be gained from a series plot of relations of a normal force, N versus a shear force τ , which acts on a body. These values help to create the sliding of that same body.

Figure 2.1: Estimation of the Friction Angle

There are two main rules that run this principality; the first is the shear resistance between two bodies is proportional to the normal force between them, and the second being the shear resistance between two bodies is independent of the size of the bodies. In most materials, different sources add to the frictional resistance, this includes sliding and rolling of particles, resistance to volume change, interlocking particles and crushing of particles. The mechanisms of these sources relates to the effect of volume change. Particles are interlocked and in contact with each other. This interlocking between particles is directly related to material density, the denser material the greater the interlocking of particles. If the shear stress is applied, the first act is resistance to volume change, and then followed by particle sliding that is relative to each – other. For a dense specimen, the particles will roll upward and over each other creating an increase in volume of the assembly of particles. In loose specimens, the particles roll downward resulting in a decrease in volume.

2.3 Bearing Capacity of Shallow Foundations

Geotechnical engineers have two problems regarding the design of shallow foundations. The first problem is the bearing capacity failure, and second is the excessive settlement. The bearing failure is confirmed by using plasticity theory, while the excessive settlement is commonly checked using the theory of elasticity. The determination for settlement in saturated clay has two mechanisms, the first is the immediate settlements due to the deformation occurring at a constant volume, and the second is the amount of consolidation settlement associated with the dissipation of pore – water pressure (Skempton & Bjerrum, 1957, [39]). The excessive total settlements are the main sources of substandard building performance. The issues addressed of which caused by unexpected consolidation, with or without the presence of water, and lack of linear elasticity to understand the earlier stages of undrained settlement that leads to serious doubts. The stress – strain of most soil is non – linear from very small strains. The non – linear stress – strain characteristics can have high influence on the form and the amount of the displacement distribution of soft clay structures. A solution known as the Prandtl mechanism, (Figure 2.2) for plane strain indentation, offers a plastic region of continuous deformation below a stiff circular punch. The strain is negligible outside this region (Osman & Bolton, 2004, [33]). The solution includes three zones of spread shear, alleged to shear and deform compatibly and continuously with no virtual sliding at their boundaries. There are many factors that can affect the performance of shallow foundations such as permeability (drained/undrained), compressibility or relative density, shape and stability of strips, adjacent foundation contact and other structures, relative soil stiffness and footing, occurrence and relative extent of horizontal loadings/moments, and stiffer or weaker underlying layers present. The soil strains and deformations created according to the shear stress keep the foundation in equilibrium. Shear stresses in the soil are linked to the external loading by the bearing capacity coefficient (Nc).

σmob = N c (cmob) Equation (2.1)

Where σmob is the applied bearing pressure, and cmob is shear stress produced in the soil.

Figure 2.2 demonstrates the certain deformation pattern where there are no displacement discontinuities.

Figure 2.2: Plastic Deformation Mechanism (Chen, 2007, [9]).The ultimate bearing capacity of a shallow foundation i.e. strip footing can be a three – term expression integrating the bearing capacity factors that are non – dimensional: Nc, Nq and Nγ, of which are linked to the angle shearing resistance ϕ’ (Terzaghi, 1943, [43]).

qf = (c.Nc) + (qo.Nq) + ½(γ.B .Nγ) Equation (2.2)

In Drained loading, the terms for calculations are for effective stresses; where ϕ´ > 0, and terms Nc, Nq and Nγ are > 0. In Undrained loading, calculations are in terms of the total stresses; where the undrained shear strength (cu); Nq = 1.0 and Nγ = 0

Figure 2.3: The three modes of Bearing Capacity Failures (Das, 1995, [14]).a) General Shear, b) Local Shear, c) Punching Shear.

The most common mode of foundation shear failure is General Shear Failure. This failure happens in moderately incompressible and strong (Rel. Density > 70 %) rock, and in normally consolidated and saturated clays, of which are loaded quickly so that undrained conditions are overcome (Coduto, 2001, [12]). Failure is sudden, and there is a clearly defined surface of failure. The formed bulge emerges to the surface of the ground and around the edges of the foundation. There can be considerable tilting of the foundation if the structure does not prevent footing rotation. (Vesić, 1973, [44]).

Local Shear Failure displays shear at the surfaces clearly beneath the foundation, which then become indistinct near the ground surface (Vesić, 1973, [44]). This happens with a footing that rests on moderately dense sand (36 % < Rel. Density < 70 %). Here, a small bulge may form, however considerable settlement (0.5 foundation width) is required before the formation of a clear shear surface develops near the ground (Coduto, 2001, [12]). The foundation will carry on sinking into the ground without the likelihood of sudden failure.

The last shear is known as Punching Shear Failure, typically common in very loose sands where the Rel. Density < 36 % of a thin layer of strong soil underlain with a weak soil, or in very weak clays in slowly loaded drained conditions. The high compressibility soil causes large settlements, and weakly defined vertical shear surfaces with these types of soils. There is little to no bulging at the ground surface due to the low soil density that is not enough to propagate the failure surface, the failure then forms gradually.

2.4 Soil Testing

2.4.1 The Triaxial Test

The term “conventional” describes the triaxial test; though, this ends up being unclear. In text, the system is defined better as an axially symmetric testing apparatus that has two degrees of freedom, the word “triaxial” being a misnomer, with an induced axial stress and confining cell pressure (Figure 2.4). The soil specimen is cylindrical and is enclosed in an impermeable latex membrane, which prevents direct contact with a sustained pressurised fluid (water), surrounding the specimen within the pressure cell. A modification to this cell pressure allows a desired pressure to confine the specimen. Afterwards, a lowered piston via the top of the cell applies the axial load. The end platens (top and bottom) are rigid and made to function in the capacity of being porous in order to allow for drainage of the water (optional), which can be used to measure pore – water pressure. The typical measurements taken during a triaxial test are the axial load (σ1), change in volume (δV), cell pressure (σ3), and change in specimen length (δl). In drained tests, there is an extension of measuring change in pore – water pressure (δu) in place of volume change.

(a) (b)

Figure 2.4: (a) Typical Triaxial Setup. (b) The typical Stress Paths for Triaxial Tests (Wood, 1991, [46])

There are two methods the axial stress can be applied: one by applying dead weights or a hydraulic pressure that is implemented in equivalent increments until failure, also known as stress - control. The other method is by application of axial deformation at a fixed rate using a geared or hydraulic loading press, this is known as an axial displacement rate or loading rate. The triaxial test is performed under one of the following settings: undrained or drained. In a drained test, the volume change in the specimen is measured by the amount of water loss via the drain lines. In an undrained test, the changes in pore water pressure in the specimen are not measured. Certain materials work under certain drainage conditions and therefore these need to be considered prior to choosing a test profile. These tests attempt to replicate full – scale behaviour with regard to both loading rates and drainage conditions.

2.4.2 The Vane Shear Test

The vane shear test determines the undrained shear strength of clay soil. The apparatus for this test includes four blades at the end of a rod. The height of a vane is twice the diameter. The vane is rectangular, or conical, and pressed into the soil. The vane shear rod is covered in a protective membrane to prevent soil adhesion during rotation. The soil tested is undisturbed by the pushing. Standard vanes will rotate at a criterion rate of 0.1°/sec. The soil fails in a cylindrical shape

surrounding the vanes. Once failed, the maximum torque applied for rotation that causes failure is then measured. (Das, 2004, [14])The undrained shear strength value attained is too large due to the increase of strength via the high rate of shear straining and soil anisotropy. Corrections are made for this using the correction factors. Additionally, it can be correlated with pre – consolidation pressure and over – consolidation ratios of clay (Das, 2004, [14]). The vane shear test is an economically simple for of test and rapid so that the excess pore – water pressure developed during the testing does not have the time to dissipate. It provides a reasonable results for medium – stiff clay. Nevertheless, this test is time – consuming and is therefore limited for soft to stiff clay. Poor calibration results in errors in torque measurement, damage to vanes and improper vane rate rotation – control.

2.5 Drained & Undrained Conditions

Drained or undrained conditions in a soil depends on the soil type (i.e. fine – grained or coarse – grained), the geological arrangement (sand layers in some clays, fissures, etc), and rate of loading. In the case of the rate of loading coupled with a normal construction activity, saturated coarse – grained soils (e.g. gravels and sands) undergo drained conditions, and saturated fine – grained soils, like silts and clays, undergo undrained conditions. If rate of loading is considerably fast (earthquake conditions), coarse – grained soils can experience undrained conditions leading into liquefaction. Drained conditions take place when there is no change in pore – water pressure via external loading. Pore – water can drain out effectively causing volumetric strains inside the soil. Undrained conditions exist where pore – water is unable to drain out when the rate of load is faster than the rate of which it drains out. In this case, the pore – water with increase in pore – water pressure, takes most the external loading, correlating that a soil’s tendency is to change volume when suppressed during undrained loading.

Quick Undrained (QU) / Unconsolidated Undrained (UU) Tests

In this type of test, no drainage of pore water is allowed at any stage. This involves the specimen shearing at constant moisture content and, if the sampling, storage and preparation of the test have been carried out correctly, the moisture content during testing should compare exactly with the moisture content of the soil in its natural state at the time of sampling on site. Additionally, it also means that if it is a saturated soil, with no drainage during the testing, there is therefore no volume change during testing. Usually undrained tests are referred to as ‘Q’ or Quick tests.

Consolidated Undrained (CU) Tests

In this next category of test, the sample initially consolidates under an effective stress corresponding to the same effective stress in – situ, where the moisture content reduces from its initial value. In other tests where soil behaviour is analysed, consolidation is a process by which pore water expels from a saturated soil under a constant total stress. Therefore, drainage allows the volume of the specimen to decrease and the completion of consolidation noted by a termination of further volume change, or discharge of water. The time needed to complete consolidation depends on the soil and permeability. After the consolidation is finished, shearing is carried out at the moisture content reached at the end of consolidation. Therefore, no recording of and permitted volume change is allowed during shearing.

Consolidated Drained (CD) Tests

The last of the three is the drained test, performed under conditions whereby drainage of the sample is allowed at all times. Afterwards, there is a repeated reduction in the moisture content of the soil specimen from that obtaining at the time of sampling. Coupled with this continually changing

moisture content are the continual volume changes of the sample during both consolidation and shearing processes of the test. Another condition of the drained test is that the loading is applied gradually so that there is no or very little expansion of pore pressure. Each addition of load is applied only when the pore – water pressure within the sample has fallen to the value existing before the application of the previous load addition. Usually drained tests are referred to as ‘S’ or Slow tests.

2.6 Mohr – Coulomb Criterion

In almost every soil testing application, there is a general criteria used to investigate the soil and determine its corresponding values of strength. Mohr – Coulomb theory is usually the criteria used to investigate soils subjected to shear testing. The parameters and theory of this particular criterion represent the shear strength of soil as Mohr circles. These circles show the state of stresses of a specimen in the plane that contains the major (σ 1) and minor (σ 3) principal stresses. Mohr circles can be used to show different specimens of the same material subjected to different confining pressures, then the friction angle (ϕ) of the material can be estimated from the slope of the line tangent to the circles, this is known as the failure envelope (Figure 2.7). Equally, depending on the type of engineering problem presented, either the peak friction angle or the constant volume friction angle is required. These angles can be determined by using Mohr – Coulomb criterion of failure.

The cohesion of a soil refers to the ability of soil particles to stick together. If is possible for the soil to be moulded effortlessly without breaking, it possesses plasticity. These properties depend on the moisture content of the soil. The consistency is referred to as the indicator of cohesive or plastic soil. The consistency varies with water content and it can range from dry - solid to wet - liquid.

The limiting shear stress for effective stress conditions (soil strength) is given by

t = c + σn tan ϕ Equation (2.3)

Where c = cohesion (apparent)ϕ = angle of shearing resistance σn = normal stress

cu and ϕ u are known as the undrained strength parameters for effective stress conditions

t = cu + σn tan ϕ u Equation (2.4)

Figure 2.5: Qualitative comparison of shear strength results for UU, CU and CD tests (McCarthy, 2007,

[32]).

Figure 2.5 shows the typical results of a triaxial test represented in Mohr circles. This additionally denotes the type of results obtained with certain test conditions. In most common tests, Unconsolidated Undrained/Quick Undrained procedures usually do not give angles of shearing resistance (depending on saturation), as they tend to undergo bulging/barrelling deformation rather than some common buckling conditions.

Figure 2.6: Results of a triaxial test on specimens of a homogenous soil. (McCarthy, 2007, [32]).

Figure 2.6 shows a single homogenous soil tested at three confining pressures and the overall “best common tangent for all Mohr circles. This method is more effective as it is generally representative of stress state within the specimen.

Figure 2.7: Mohr circles showing stress state of three different specimens of the same cohesion-less soil (c

= 0) when subjected to different confining pressures.

ϕ

Figure 2.8: Mohr circle determinations using combinations of axial and confining pressure for non-

granular materials (Vickers, 1983, [45]).

Figure 2.9: Mohr circles for the Total and Effective Stress for tested materials (Vickers, 1983, [45]).

Figures 2.7, 2.8 and 2.9 are demonstrating the denoting of the angle of shearing resistance, and using the tangent line of the Mohr circle to determine the slope of the common tangent.

(a)

σ n

(b)

Figure 2.10: Mohr circles comparison between Theoretical (a) and Actual Results (b) for most soils

(Vickers, 1983, [45]).

Figure 2.10 is demonstrating the difference between the expected triaxial results and actual triaxial results. The reason why this is also studied is that traditionally if the Mohr circles are representative of a single specimens stresses, there are still limitations, estimations, and uncertainties brought about in the test results. The graphical representation of the Mohr circles is a good method to demonstrate if anything has not followed according to predictions. The reason for testing these kinds of materials is that the strength properties of silty and firm clay are not yet fully understood, unlike the conditions of pure clay and sand, of which do represent the two limits of soil behaviour due to their mechanical properties and overall composition. Mohr circles for actual tests (Figure 2.9 (b)) are common in all triaxial test profiles (UU, CU & CD).

2.7 Shear Stress Parameters

The most widespread form of triaxial test is the Conventional Triaxial Compression or CTC test. This method involves loading the specimen in the axial direction whilst keeping a constant confining pressure (σc). This method is based on the statement that there are no shear stresses occurring at the end platens, σc and the axial stress: σa = σc + (Fa /A) can be taken as the major (σ1) and minor (σ3) principal stresses. Figure 2.11 demonstrates the stress states on a typical cylindrical specimen.

Figure 2.11: Illustration of the stresses and forces acting on a CTC subjected specimen.

(BS 1377, 1990, [48])

An obtainable and suitable analysis of these results can be executed by applying the following equations and constructing the related relations:

Deviatoric stress, q = σ1 - σ3  Equation (2.5)

Axial strain, ε a=∆ H /H 0 Equation (2.6)

Volumetric strain, ε v=∆ V /V 0 Equation (2.7)

Major Principal Stress, σ1

Minor Principal Stress, σ3

Undrained Strength, σ1 - σ3 = 2 cu Equation (2.8)

Where:

Shear strength, τ = (σ1 - σ3)/2 Equation (2.9)

Hence, τ = cu, the undrained strength.

Where: ∆ H and H 0 are the change in height and initial height of the specimen. ∆ V and V 0 represent the change in volume and initial change in volume.

Out of this, the following strains can be calculated from these measurements: l0 = initial length of specimen

δl = change in length of specimen, compression is positiveV = initial volume of specimenδV = change in specimen volume, volume increase (expansion is negative)The axial strain rise,

δε r=−δl

l0 Equation (2.10)

The volumetric and radial strain rise,

δε r=δεv−δεa

2∴δε v=δεa+2 δε r=

−δVV

Equation (2.11)

The shear strain rise,

δεq=23(δεa−δεv) Equation (2.12)

The subsequent stresses can be calculated from these measurements: σ c= pressure in cell (σ3)Fa = axial force (σ1)A = current cross-sectional area

The deviatoric stress

q ≈Fa

AEquation (2.13)

The total mean stress

p=σc+q3

Equation (2.14)

The conventional triaxial test has traditionally been the most widespread method to determine soil strength properties due to its control simplicity; however, with more complex loading scenarios, various stress paths can be effortlessly followed (Figure 2.2). The cell pressure is held constant while the axial force is increased in a triaxial compression test, this results in the additional total mean stress becoming δ p=δ q/3 which is represented as line AB in Figure 2.2, where A is the initial state of stress. In a conventional triaxial extension test, the loading ram is withdrawn causing the axial stress to decrease while the confining pressure kept constant inducing a negative deviator stress. In this case, the stress path would include the same overall slope as the conventional compression test, but in the negative direction, shown by line AC in Figure 2.1. The next line DE illustrates the stress path for maintaining the axial stress constant while altering the cell pressure simultaneously, which results in a rise of the total mean stress relationship. The line FG in Figure 2.2 can be important where the total mean stress is constant and if the separation of volumetric and distortional features of soil response is required. If the test was to be executed using an undrained triaxial test, the distinction between the total and effective stresses must be considered.

During the triaxial test, the specimen goes through deformational changes depending on the type of material, confining pressure, and amount of applied force. In most sands and gravels tested, these would demonstrate an almost clear plane of slip identical to buckling in slender members with the type of failure expected with that material. This is not usually the typical response of clays or strongly cohesive materials as the type of deformation resembles a bulge or barrel type structure and is hence give the name bulging/barrelling to describe the phenomena. Most specimens of a short nature rather than slender tend to undergo barrelling phenomena, where as slender specimens are more commonly to deform to the buckling type of failure mode. Figure 2.12 shows the type of general deformation of a specimen subjected to a triaxial test.

Figure 2.12: The confining, shearing, and failure modes of a specimen under Triaxial Compression

(Vickers, 1983, [45])

Figure 2.13 shows the overall slip failure of a specimen subjected to triaxial loading. This additionally indicates the major and minor principal planes in correlation to the failure of the test specimen. This includes the relevant normal and shear stresses at the line of slip.

Figure 2.13 Failure plane of a specimen under TC (Vickers, 1983, [45])

Figure 2.14: Triaxial compression test showing test pressure and assumed plane of failure AB. (Duncan, 1998, [16]).

Figure 2.14 demonstrates the potential type of deformation and the plane of failure with a specimen usually tested in the triaxial device, indicating where the angle of shearing resistance is determined. The specimen can usually deform to buckling if it is slender, however shorter specimens in diameter and height tend to show more complex failure modes than most common slender specimens.

2.8 Loading and Stress Parameters

In laboratory tests, the desired loading rate is for the producing no excess water pressures and the specimens are free to drain from the cell. The parameters known as ‘effective stresses’ can be found from the applied total stresses, and the known /determined pore – water pressure on the specimen. Only the effective strength parameters, c’ and ϕ’, have application only to drained tests. With these values, it is possible to construct a series of total stress Mohr circles (Figure 2.9) but the inferred total stress (undrained) strength parameters become insignificant.

When dealing with undrained laboratory tests, there will be no drainage from the specimen, along with no moisture redistribution within the specimen. There can be slower axial displacement rates in triaxial tests as the conditions are uniform, and no allowance of drainage from the sample. When there is pore – water pressure measurement in a triaxial test, the effective stresses can be determined (effective parameters noted with a ’) and the effective strength parameters c’, ϕ’ are assessed. As aforementioned, these can be used to assess long – term stability. Undrained tests are traditionally used to determine total (or undrained) strength parameters cu, ϕu. If these parameters are to be pertinent to ground parameters, the moisture content is to be the equivalent. This can be achieved by both performing the QU/UU tests or CU tests and consolidating to the in – situ stresses. The total, or undrained, strength parameters consider the short – term stability of soil – based constructions. It is vital that there be no drainage if this approach is to be suitable. This means that Mohr’s circles with total stress analysis would not be appropriate for sands and gravels. In clay soils, a total stress analysis is the likely way to assess stability. Additionally, undrained strengths can be determined for any soil, however they are not usually relevant in practice. In the case of sands and gravels, the pore pressures drive out rapidly, and the effective strength parameters can be used to check the short – term stability. Hypothetically, the effective strength parameters can also be used to check the stability at any time for any soil type; however, the pore pressures in the ground must be identified prior to testing and usually known in the long – term.

When documenting the different total stress Mohr circles against a single effective stress Mohr circle, this indicates that the pore pressure is different for each sample. Increasing the cell pressure without allowing drainage causes an increase in the pore pressure by the same amount (Δu = Δσc) with no change in effective stress. During shearing of the specimen, the change in pore – water pressure is a function of the initial effective stress and the moisture content. These would be identical for the say three samples with an identical strength.

2.9 Recent & Past Advances in Triaxial Testing

Along with discussing the principles of the triaxial test, an additional aim of this literature review is to also understand and state the recent, past, and new advances made in triaxial testing. There have been many updates and modifications on methodology, instrumentation, and measurement since the publication of ‘The Measurement of Soil Properties in the Triaxial Test’ (Bishop & Henkel, 1962 [6] / 1979 [7]). This document is used as underpinning for obtaining better existing, and developing new, triaxial testing standards for this study and explores the validity of current British Standard/American Society for Testing and Materials. This review is a collection of state – of – the – art papers separated into four key areas; these include (I) Equipment, (II) Methodology, (III) Errors and Interpretation, and (IV) New Test Varieties.

2.9.1 Equipment

The first part of this review describes examples of specific equipment and the systems for triaxial testing on both soil and rock. The first paper (Tatsuoka, 2001, [42]) reviews the testing equipment for both the static and cyclic loading tests. Additionally, this describes considered plane strain and simple shear devices (direct shear). In this paper, many devices described can be applied to clays and sands; yet, the illustrations given are only for the testing on non – cohesive soils. Three papers in this

category use computer – controlled testing systems, these include two in soils and one considering soft rock.

The next four papers describe the specifically designed triaxial cells for various uses; this includes the study of static and dynamic behaviour, the real – time measurement of different parameters, soft rock testing and specimens of larger diameters regarding in – situ conditions. The last four papers express the using effects of some components of the triaxial tests. These last papers include the use of a flow pump, which permits the measurement of tensile soil strength, the use of filter paper, practices to reduce the leakage in long – duration tests, and finally a comparison of local and global deformations and their measurements. These papers, and various compiled reports indicate that equipment in this field possess test devices capable of following a stress path of a soil specimen almost mimicking loading encountered in – situ. Nonetheless, it also demonstrates a lack of appreciation in testing along given strain paths that could be noteworthy in engineering applications. One example of this is the simple form of the controlled strain path test, where the constant cross – sectional area (A0) test is referred, but the more generalised and common controlled strain path tests are not mentioned, when thoroughly browsing though these texts that the technology exists for that application.

2.9.2 Methodology

In discussing the next category, two papers have significance above the rest. The first paper (Baldi, Hight & Thomas, 1988, [3]) re – examines some features of the conventional methods in the triaxial tests based on the initial work of the first publication (Bishop & Henkel, 1962 [6] / 1979 [7]). This integrates advances in sample preparation practices and measurement, sample disturbance, various stress paths, and use of UU (Unconsolidated Undrained) tests are conversed. The second significant report (Lacasse & Berre, 1988, [27]) reduces the common 1986 triaxial testing preparation at the Norwegian Geotechnical Institute. The next three papers explain the detailed use of multistage drained triaxial tests on highly variable soils, where the other two papers then go on to focus on experience in dealing with contaminated soils and soils of high gas content. The remaining four papers then discuss the specific test preparations on relaxation, weak rock extension tests, high cell pressure testing, and practice of recomposed sand specimens.

2.9.3 Errors and Interpretation

With the third category, there are over 20 papers and records all thus indicating the significance of this area and implementing test results to engineering designs. The first paper (Germaine & Ladd, 1988, [20]) reassesses the various types of errors involved in measuring the behaviour of the soil specimen and suggests that (UU) unconsolidated undrained and (CU) consolidated undrained tests do not give reasonable undrained strength values. This paper additionally includes the assumption that tests like the anisotropically consolidated (in – situ pressure) undrained compression and extension (CA0U) tests and oedometer tests should commonly and preferably used more.

The next compilation of papers deal with the influence of the test conditions, procedures, preparation of specimens, and the geometry on the test results. Most of these are considered on previous experimental studies, while others are based on more commonplace theoretical analysis. This third category would contain more details if there had been more information presented that carefully exemplified the application of these influences on actual problems in engineering design. The most important point of this area is the strong supporting statement to reject the UU test for design purposes, when generally practiced on clays to develop behavioural relationships. However, a premature assertion is made to perhaps discard the CU test (Germaine & Ladd, 1988, [20]). As aforementioned, the CA0U is suggested as a substitute for the reason of it being more theoretically valid, however, difficulty lies in measuring the correct A0 value and the overall process being time consuming.

2.9.4 New Test Varieties

The last category consists of six papers devoted to test devices that allow independent control of stresses/strains in three principal directions. The first paper in this category (Arthur, 1988, [1]) discusses the adaptability and limitations of various devices for testing cubical specimens to develop a true triaxial apparatus for soils. The second paper (Saada, 1988, [35]) reviews the advantages and constraints of hollow cylinder torsional devices for both static and cyclic loadings. The next three papers describe the devices when using entirely rigid loading boundaries, the remaining three papers then describe devices using a combination of flexible and rigid loading boundaries. These papers also illustrated tests on soil (undisturbed or reconstituted) and rock specimens. This area consistently highlights the difficulty that arises in conducting a true triaxial test for producing a representation of a stress/strain relationship in a soil. Nonetheless, highly specialised and complex machinery do not guarantee synchronised stress, and strain fields.

There is still a lack of concern regarding the need of relationships via test results from these devices with those observed in practical situations, and therefore there will be a certain amount of time before these tests are accepted and integrated into geotechnical engineering and design. Additionally with all these methods to determine soil strength properties, there are also particular tests that must be conducted in order to establish the variability of specimens and what constants will be kept in order to determine a relationship. The next chapter will discuss the preparation and classification of the materials in order to better establish the degree of variability among them, and what possible relationships could those particular materials yield.

CHAPTER 3 – MATERIALS CLASSIFICATION & PREPERATION

3.1 Introduction

This chapter describes the particular materials used in the experimental work (Chapter 4). The procedure used to modify the density, using compaction methods, of the clay material specimens; along with a description of the used materials, analyses that were performed using visual description and pre – tested moisture content are reported. Additionally, the chapter presents a description of the physical properties and the tests performed to quantify those physical properties of the particulate material used.

3.2 Material Description

The materials used in the experimental testing were cohesive materials obtained from GLS Supplies UK. They are widely used due to their availability. In the case of clays, for the eventual study of shear strength, the practical use of these kinds of clays is primarily in modernised construction and landscaping. Colours describe the soil appearance. Organic soils are dark brown or dark green in colour, while Peat soil is dark brown or black in colour. The description of colour is used with caution as the soil mass colour can change with a change in moisture content or chemical composition. (Lambe and Whitman, 1969, [29])

Clay X

Buff School Clay - (grogged) - an excellent multipurpose clay with a wide firing range, a fired colour of light creamy gray buff getting darker towards the top end of the temperature range 1120 – 1280°C.

Features the same characteristics as the school buff smooth clay but with added grog giving extra strength for larger pots and modelled work. Dull – like in appearance and colour, elastic as well as being very moist for clay. Figure 3.1(a) shows the elastic and firm nature of the Buff clay. This material has Hue 7.5Y/Chroma of 7/1or “light gray” soil colour. Clays such as this are common in sub – humid environments, where the evaporation is greater than the precipitation, and there is an upward movement of water and soluble salts present in the soil

Clay Y

Toasted Stoneware Clay - A lightly grogged clay firing to a warm brown toasted colour. Very good handling characteristics. Firing temperatures 1150 – 1260°C. The clay itself is very crumbly when taken away from a larger cluster. It can present itself as aggregates, clumps and even smaller particles. Figure 3.1(b) shows the clump yet silty like nature of the Stoneware clay. This material has Hue 5YR/Chroma of 4/4 or “dull reddish brown” soil colour. The colour suggests that the clay had, from its location, high annual precipitation amounts; a soil that is high in soil moisture, and/or litter from coniferous trees favour an accumulation of organic matter. The material is toasted; therefore, this reduces the amount of moisture present and changes the overall properties.

These clays are found in deposits at depths roughly about 1.5 m to 2 m in the UK. They are excavated then processed for general use and industrialised use. The colours were determined using the Munsell Soil Colour Chart. Shallow foundations are situated where the depth below finished ground level is less than 3 m and include strip, pad and raft foundations (3.1 of BS 8004, 1986, [47]). Silts average particle size is between 0.002 – 0.06 mm, where Clays have a particle size of < 0.002 mm. Particle Size Distribution indicates that the clays consisted largely of SILT and SOFT CLAY. The chemical tests done indicated that the soil is of 4.5 to 8 pH values and the amount of organic present is less than 0.01%, which is insignificant. The total sulphate and chloride content are less than 0.01%, and the cohesive materials are considered non – aggressive.

(a) (b)

Figure 3.1 (a): Buff School Clay in its natural form. (b): Stoneware Clay shown as large clump and

clusters.

3.3 British Standards in Experimental Work

With most materials being tested, there are standards for the specifications of the conditions. Below are the British Standards that are taken in conjunction with tests run on soils or clays of this nature and referenced from publication Eurocode 7.

Eurocode 7, Part I, BS 1997-1This standard is the general basis for geotechnical aspects in the design of buildings and civil engineering works, geotechnical data evaluation, ground improvement and reinforcement, dewatering

and fill. This additionally includes geotechnical design of spread foundations, piles, retaining structures, embankments and slopes. Calculation rules for actions originating from the ground.

Eurocode 7, Part II, BS 1997-2This part is the requirements for execution, analysis and use of results of laboratory and field tests to assist in the geotechnical design of structures. This leads into the Standard BS 1377: 1990.

BS 1377, 1990 This British Standard has nine parts. Part 1 is general information relevant to the other Parts. Parts 2 to 8 explain methods of soil tests for civil engineering purposes where samples are taken for testing in a laboratory. The laboratory test procedures as follows Part 2: Classification tests; Part 3: Chemical and electro-chemical tests; Part 4: Compaction-related tests; Part 5: Compressibility, permeability and durability tests; Part 6: Consolidation and permeability tests in hydraulic cells and with pore pressure measurement; Part 7: Shear strength tests (total stress);and Part 8: Shear strength tests (effective stress).

BS 8004, 1986This British Standard is the code of practice and provides recommendations for design and construction of foundations for the normal range of buildings and engineering structures. Section two covers the general principles of design; sections three, four, five and seven are detailed considerations of design and installation of main types of foundations. Sections six, eight and nine narrate site operations and construction processes involved in foundation engineering and section ten describes the factors affecting the durability of the various materials used in foundation structures. Section eleven covers safety precautions. The standard does not cover foundations for special structures.

3.4 Preparation of Remoulded Specimens (Compacted)

Each tested specimen was weighed to 200 g to ensure the proper fitting into the 38 mm diameter aluminium split mould. The weight of specimen would change if the specifications of the tested specimen diameter were to change or increase. Usually larger specimens require more of the sample from the undisturbed or remoulded source. (3.0 of BS 1377 – 4: 1990, [48])

Figure 3.2 (a): Specimen on weighing scale. (b) Two specimens of both clays correctly weighed.

In general, the specimen in a triaxial cell usually has a height equivalent to about two times the diameter (H/D ratio), with the plane ends normal to the axis. This means in most cases the size of the biggest soil particle cannot exceed one-fifth of specimen diameter. If any large particles found their size and mass is noted. In some cases, undisturbed specimens cohere to field or specified conditions.

3.4.1 Compaction Method for Soils using a 2.5 kg Rammer

Generally, this test covers the determination of the dry density of soil passing a 20 mm test sieve when it is compacted in a specified manner over a range of moisture contents. The range includes the optimum moisture content at which the maximum dry density for this degree of compaction is obtained. In this test, a 2.5 kg rammer is used falling through a height of 300 mm to compact the soil in three layers into a 1 L compaction mould. The amount of actual specimens is smaller than the 1 L mould can comply with so only the bottom layer of the compaction mould is used.

3.4.1.1 Compaction Criteria

The amount of compaction is related to two main criterion, this being 1) Compactive Effort, in which the soil is compacted at a specified moisture content into a mould under the application of a specified compactive effort (3.3 and 3.5 of BS 1377 – 4: 1990, [48]), the last being 2) Dry Density, in which the soil is compacted to a specified moisture content into a mould in order to gain a specified dry density. Any of the techniques is used and is congruent to the following types. i) A sample that is bigger than the required test specimen, in which one or more test specimens of smaller size are to be taken (7.7.4 in BS 1377 – 1, 1990, [48]), ii) the specimen tested, when dealing with large diameters, e.g. 100 mm (7.7.5 in BS 1377 – 1, 1990, [48]).

3.4.1.2 Compaction Equipment

A cylindrical British Standard compaction mould was obtained, having a nominal internal volume of 1 L. The mould is fitted with a detachable base plate and a removable extension. The essential dimensions are in Figure 3.3, which also indicates the shape of the mould. The internal faces are smooth, clean and dry before each use. A metal rammer has a circular face with 50 ± 0.5 mm diameter, weighing approximately 2.5 kg ± 25 g. This rammer has a suitable arrangement for controlling the height of drop to approximately 300 ± 3 mm. Additionally, there is a balance readable to 1 g, a palette knife, a straightedge steel strip 300 mm, by 25 mm, and 3 mm thick, one bevelled edge. Test sieves, orifice sizes 37.5 mm and 20 mm, and a receiver. Lastly, a corrosion – resistant plastic tray with sides 80 mm deep.

(a) (b)

Figure 3.3 (a): 2.5 kg Rammer for Compaction. (b): 1 L Mould Cylinder to 0.01 g

3.4.1.3 Compaction Procedure

Firstly, the mould is weighed with base plate to the nearest gram (1 g). The initial dimensions are measured to the nearest 0.1 mm. The extension is attached to the mould and the mould placed onto a solid base. This would be the concrete floor, and kept in place to avoid disturbance. A) An amount of the moist soil was placed into the mould so that at compaction, it occupies just over a third of the height of the full mould. B) 27 consecutive impacts applied to the specimen within the mould via the rammer, dropped at a height of 300 mm above the soil, controlled by a guiding tube. The soil is compacted by 27 blows for the 1 L mould, where the first four (4) impacts correspond to the pattern of Figure 3.4 (a). After the first four blows, the rammer is then moved, in the direction according to Figure 3.4 (b) between the numbers of successive blows. This allows the blows to uniformly distribute over the whole area. The impacts are performed uniformly over the surface of the specimen and the rammer was allowed to fall freely, uninterrupted by soil in the guide tube.

(a) (b)Figure 3.4 (a): Sequence of first four impacts. (b): Sequence of successive blows after the first four.

Processes A) and B) are repeated twice more so that the amount soil used is enough to fill the mould, and this ensures that the surface is not more than 6 mm proud of the upper edge of the mould. It is also necessary to control the amount of total volume of compacted soil as it has been discovered that the amount of soul struck off after removing the extension is to large, the results become inaccurate. The extension is then removed, and any excess soil is stroked off and the surface is levelled off the soil compacted, the top of the mould is carefully cleared of any excess using the straightedge. Any course particles that were taken off in the levelling process, including fine material, is replaced from the original sample. Weighing of the soil with the mould and base plate is completed to the 1 g (m2). C) The compacted soil then removed from the mould and then placed into a metal tray. A selective representative sample of the soil is taken in order to determine the moisture content (3.2 of BS 1377 – 2: 1990, [48]). The remaining soil is broken up, and chafed through a 20 mm test sieve and the rest of the prepared test sample is mixed. Process D) involves adding suitable additions of water and mixing thoroughly into the soil (in general, additions of 2 % to 4 % for cohesive soils) since this has not been prescribed, nor are the specimens original moisture contents modified, it was ignored. Processes A) through C) are repeated in order to gain a total of least five determinations. The internal volume of the 1 L mould is then calculated, V (cm3). Determination of the bulk density, ρ (Mg/m3), of each compacted specimen from the equation:

Bulk Density Formula ρ=m2−m1

V Equation (3.1)

Where:

m1 is the mass of mould and base plate (g);

m2 is the mass of mould, base plate and compacted soil (g).

Then, calculate the dry density, ρd (Mg/m3), of each compacted specimen from the equation:

Dry Density Formula ρd=1 00ρ

100+w Equation (3.2)

Where:

w is the moisture content of the soil (%).

(Head, K.H, 1982, [23])(BS 1377 – 4: 1990, [48])

3.5 Determination of Moisture Content of Specimens

Soils have water present within the structure, through numerous voids. This amount of water, expressed as a mass of the dry particle, is the moisture content. This has an effect on the soils mechanics and behaviour. In this circumstance, a soil is only dry when there is no further possible removable of water at a temperature that does not exceed 110° C. The moisture content is needed as a guide to classify certain natural soils and be used as a method of criterion for control in re – compacted soils. It is usually measure on samples used for common field and laboratory tests. The most common method in determining the moisture content is the oven-drying method. (3.2 of BS 1377 – 2, 1990, [48])

3.5.1 Oven-drying Method

This section covers the method to determine the moisture of content of a specimen of soil expressed as a percentage of the dry mass. The equipment used is as follows: a drying oven, capable of keeping a temperature of 105° C to 110° C. A microwave is unsuitable as it is difficult to use a microwave to maintain temperatures. A glass – weighing bottle, fitted with a ground cap, or a suitable airtight corrosion - resistant metal container, a balance readable to 0.01 g, and desiccator containing anhydrous silica gel.

(a) (b)

Figure 3.5 (a): Crucible containers used to weigh specimens. (b): Balance readable to 0.01 g.

3.5.2 Procedure

The weighing container is cleaned and dried and measured to the nearest 0.01 g (m1). A sample of 30 g is then taken from the soil, crumbled and placed loosely into the weighed container, and the lid placed on top. The contents and the container are weighed to the nearest 0.01 g (m2). The lid prevents any loss of moisture prior being placed into the oven. The lid is removed, and the containers are placed into the oven and dried at 105° C to 110° C. The temperature chosen is 110° C. The sample was not checked periodically every certain number or hours and this reduce effectiveness of cooling the sample.

The sample was left for 24 hrs, as this was deemed appropriate for clay samples. After drying, the specimen with the container was removed from the oven and the whole specimen placed into the desiccator to cool. The then dried specimens were left to cool for about 30 minutes to 1 hr. A desiccator is not essential but can reduce any further sources of error. The container and contents is weighed to the nearest 0.01 g (m3).

The moisture content of the soil specimen was then calculated, w, as a percentage of the dry soil mass to the nearest 0.1% using the equation:

w = (m2−m3m3−m1 ).100 (%) (Equation 3.3)

Where:

m1 is the mass of container (g);m2 is the mass of container and wet soil (g);m3 is the mass of container and dry soil (g).

Table 3.1: Specimen description of densities, mass and moisture content.

Sample Ref.

Specimen Ref.

Description Mass (g)

Bulk Density(Mg/m³)

Dry Density(Mg/m³)

W – Moisture Content (%)

C_X1 X1 Stiff brown silty CLAY

200 2.13 1.78 20

C_X2 X2 Stiff brown silty CLAY

200 2.11 1.76 20

C_Y1 Y1 Firm grey CLAY 200 1.99 1.54 29

C_Y2 Y2 Firm grey CLAY 200 1.96 1.52 29

CHAPTER 4 – EXPERIMENTAL WORK

4.1 Introduction

A series of quick undrained/unconsolidated undrained triaxial compression tests were performed on the two types of cohesive materials, of which each had different moisture contents and different sets of multistage confining pressures subjected. This included each type of soil having two sub – samples of which the parameter varied was clay type and confining pressure. This meant that two clay specimens were drier than the other two and of different clay composition. The parameters that were additionally constant in these tests were density, bulk density, and moisture content. This chapter presents a description of the equipment used to perform those tests, the procedure for pre – testing specimen preparation, and a table summary of the tests performed.

4.2 Equipment Description

The testing equipment used for this investigation consists of the following main parts: triaxial cell for 38 mm diameter by 76 mm height specimens, aluminium split mould, loading frame, pressure control gauge with dials, a de – aired water tank with water supply, air filter, air compressor, vacuum pump, an AVC (Automatic Volume Control). This triaxial system can support up to three types of tests as aforementioned, the primary set up of this system being more refined towards a standard quick undrained test. In order to obtain a reasonable assessment of the c and ϕ values, four experiments were performed on two identical specimens of two different clays at three different cell pressures. In these experiments, the Wykeham Farrance Triaxial System is used with the conventional triaxial cell with an added bottom end platen of 38 mm in diameter. The loading frame is has Pressure Gauge Dial for reading the loading stress of the specimen at failure. With this system, there is one main load cell with a maximum load capacity of 1700 kPa used for multistage confining pressure tests, the first clay specimens (C_X1 and C_Y1) at 25 kPa, 50 kPa and 75kPa tests of confining pressure tests. Then, this cell is reloaded with the second clay specimens (C_X2 and C_Y2), and then tested at confining pressures of 30 kPa, 45 kPa and 60 kPa. The loading ring and cell used in experimental work is presented in Figure 4.1, which has a pressure gauge dial.

Figure 4.1: Loading cell with Pressure Gauge and Loading ring; used for the confining pressure tests.

4.3 Preparation of Specimen into Loading Cell

Additionally, the Brunel VJTech Advanced Triaxial testing system is used to explain the procedure of placing the specimen into the base pedestal, not for actual testing, as the equipment in the university laboratory makes the process easier to demonstrate. This procedure is demonstrated in Figures 4.2 and 4.3. After the specimens are compacted, they are placed into a 38 mm diameter two – part aluminium split mould. Each specimen has similar bulk densities at different moisture contents. Hence, a series of tests would run on two sub – specimens of the sampled clay material, at different multistage confining pressures. A cylindrical latex membrane is initially attached to the bottom end platen with the help of an o – ring (38mm), then the mould is placed around that platen and the membrane is tightened along its inside using grease. No porous stones are placed above the bottom platen and top platen of the specimen, prior to attaching the final o – ring. The membrane will remain in the specimen itself with the help of an additional confining correction pressure. Figure 4.4 (a) shows the stretched membrane along the inside of the mould used with grease to keep the membrane fixed to the inside of the mould. The mould is filled with the corresponding specimen of approximately 200 g. After the deposited specimen is secure in the mould, it is then as shown in Figure 4.4 (b); this method was chosen in order to get more consistency in the void ratio values than with the pluviation method. When the mould is completely or mostly filled with the specimen, the top end platen is attached with another o – ring, the edge of the membrane covers the top platen with a o – ring and a membrane correction pressure that is approximately 2.0 kPa to 3.0 kPa will be applied to the outside of the specimen to prevent disturbance. Subsequently, the aluminium mould is then removed and the cell sleeve is set in place along with the top cell plate and the loading ram.

Figure 4.2: The Brunel VJTech Advanced Triaxial testing system (not functioning).

(a) (b)

Figure 4.3 (a): A Latex membrane stretched along the inside of the mould. (b): Compacted specimen

within the mould.

Figure 4.4: A prepared specimen in the cell ready to be tested.

The reason for the experimental work not performed on the Brunel triaxial system is due to lack of instructional literature regarding the set-up, equipment, and time – constraints in order to get it working.

4.4 Equipment Preparation

After the cell is fully assembled, it is then placed into the loading frame. The cross beam that has the loading cell is manually adjusted to the desired height. The cell becomes pressurised with water to create a confining pressure. Figure 4.6 demonstrates the complete set-up of the triaxial cell in the loading frame structure, ready for compression. Afterwards, the desired cell pressure is applied while reducing the vacuum to avoid the confining the specimen to a higher pressure than of the desirable test confining pressure. Then, the vacuum is removed, the specimen is vented and the test can begin. There is no data acquisition system, as the readings are taken manually.

In QU (Quick Undrained) tests, the total stresses are normally measured in a triaxial cell where the sample is subject to an all round confining pressure. The load applied through a piston onto the top platen cap, additionally with the specimen being confined within a rubber membrane so that no drainage in or out of the specimen is allowed. Generally, pore – water pressures are not usually measured and the undrained test is often referred to as the QU – TXL test. An addition of the QU test is the unconsolidated undrained test (UU); this is similar to the QU test but run at a slower rate in order to measure pore water pressure. By comparison, effective stresses when measured in a triaxial cell are more complex in their nature, as numerous parameters can be measured. These include backpressure, pore – water pressure and volume change; all of which can be used to calculate the required engineering properties. The tests used to determine effective stress are usually referred to as consolidated drained (CD) or consolidated undrained (CU). Generally, the CD test is applicable to sands and gravels while either the CU or CD test can be used with clays to further study their drained behaviour.

In the case of the Wykeham Farrance system at GeoLabs, the quick undrained procedure was already prescribed and ready to be used at any time. This eliminated any time needed to prepare the system set at other conditions, and any final modifications regarding the system for the specimens was also made prior to testing.

4.5 Testing Procedure

All the tests were performed under the QU (Quick Undrained) conditions of the Triaxial Compression (TC) conditions, where the radial confining pressure was kept constant while the axial load was amplified at a constant rate. Here the specimen was subjected to three confining pressures in order to get three representative Mohr’s circles for the shear strength and stresses of that one specimen. This is more effective and less time - consuming than single stage tests as it less involves removal, remoulding and preparing the same specimen to a higher confining pressure after failure. Figure 4.5 shows the general setup and features of a quick undrained test using the Brunel Triaxial testing system that has relativistic features as the Wykeham Farrance system. The computer terminal is optional when software for data acquisition is present.

Figure 4.5: General setup of Quick Undrained system. (VJTech Soil and Rock Testing Manual, 2009)

4.5.1: Multistage Quick Undrained Procedure

Generally, this test covers the determination of the undrained compressive strength of a cohesive soil specimen when subjected to a constant confining pressure, and to strain – controlled axial loading, where no change in the total moisture content is permitted. Additionally this method gives a means of finding a relationship between the confining pressure and undrained shear strength via a single specimen. In most cases, this test is done without brittle, non – cohesive or sensitive soils. (9 of BS 1377 – 4, 1990, [48])

As aforementioned, the test in the triaxial system is applied to a cylinder specimen of a height approximately equal to twice the diameter. In most cases, the specimen has the same diameter, about 100 mm; this is usually the formal size as an undisturbed specimen taken from the ground. Since the specimens are taken from a processed clay source, then cleaned, and free of any organic matter; the diameters can range from 38 – 150 mm. In this case, the specimens are 38 mm in diameter and 76 mm in height. The specimens are confined in a transparent impermeable membrane between impermeable end caps in the triaxial cell, with no porous stones, which is then pressurised by a fluid in order to

create confining pressure acting on the specimen. The axial load in then increased in usually three stages by applying a constant rate of strain until the maximum vertical stress is reached, each stage has a different confining pressure.

The tests are performed under the following test conditions, and these are specified before starting the test. The test specimen and the confining pressures for each stage are initially clarified. It is sometimes appropriate to have cell pressures of about 0.5vs, vs, and 2vs, where vs represents the total vertical stress in – situ of a shallow foundation. The pressures are usually chosen in order to mirror the range of likely vertical stress experienced by a soil in – situ. Again, with compacted soils, the cell pressures should be indicative of those estimated total stress likely to occur in common foundation conditions. In this case, the tests are not dealing with in – situ specimens, and this study is investigating a relationship between confining pressure and undrained shear strength. Therefore, the pressures chosen are closer together in order to get a more refined determination of shear strength in relation to increasing confining pressure. This helps to determine the probability of demonstrating a relationship observed as a slow increase or decrease in the strength and pressure values. The specimen densities are remoulded to a NMC standard density at their chosen pressures. In over – consolidated clays, the lowest cell pressure should not usually be less than the total vertical in – situ stress. This involves an undisturbed or remoulded specimen that is ready to be tested. For a remoulded specimen, the moisture content or dry density has been obtained during the compactive effort process.

A Triaxial cell, of dimensions suitable for 38 mm diameter by 76 mm height test specimen was chosen, suitable for use with water at internal working pressures needed to perform the test. A gas is not usually suitable for creating a pressure; that is why this system features a water tank. The main features of the device shown in Figure 4.6 below.

(a) (b)

Figure 4.6 (a): General features of the Wykeham Farrance Triaxial System. (b): Loading Cell

The cell top plate is made of corrosion – resistant material that is fitted with an air – bleed plug and a close-fitting piston guide bushing. Additionally, with a loading piston for applying the axial compression to the specimen, lateral bending of piston during the test is usually negligible and ignored. The piston is cleaned properly before use, and oiled lightly for lubrication. There is friction between the piston, or seal, and the bushing; usually this is small enough to allow the piston slide smoothly under its self - weight when the cell is empty. There is clearance between the piston, and the bushing or seal usually minimises the leakage from the cell. The cell body is always cylindrical, and made of a removable base pedestal and top cylinder; it is then sealed to the top plate and base plate. The cylinder membrane is made of a transparent material, so the specimen can be observed during the test. The cell base is also made of corrosion – resistant material with a built – in connection port.

The devices used for applying and maintaining the desired pressure on the water within the cell run to an accuracy of ± 5 kPa with a gauge of a test grade for measuring pressure. Pressure systems are dependent on air pressure regulators, dead – weight pressure cells, and even oil pressure regulators, which have been used successfully. In most cases, their ability to supply or take in water is adequate to compensate for cell leakage.

A bourdon tube pressure gauge is used part of the instrumentation (4.2 of BS 1377 – 1:1990, [48]) for the tests covered by this standard, and they are of suitable test grade suitable. Electrical pressure transducers can be used as an alternative, the working range of either type of instrument is appropriated to the requirements of the test apparatus. These are readable within 0.5 % of the full scale reading for pressures exceeding 10 % of the current reading. In these tests, a machine capable of applying an axial compression at a uniform rate to the confined specimen at a speed should be within a range of 0.05 mm/min to 4mm/min. In the Wykeham system, the rate is chosen as a standard test rate of between 1.9%/min to 2.0 %/min which complies with the standard. An appropriate strain is selected before commencing the test. In undrained shear, the selected strain is chosen to ensure equalisation of pore pressures throughout the specimen. An extraneous requirement is that it should additionally be able to apply an axial deformation of about one- third of the height of the tested specimen. The Wykeham system also complies with this protocol with the dial gauges readable to about 0.01 mm, as this depends on the range of travel specified. (4.2.1.3.5 of BS 1377 – 1: 1990, [48]).

A calibrated measuring device to determine force is supplied; this is supported by the crosshead of the compression machine so that the self – weight is not transferred onto the test specimen. A load ring compensates for this, additionally a force transducer, or a mounted transducer inside the triaxial cell capable of being underwater is used as an alternative. An alternative range of calibrated force – measuring devices were readily available to suit the specimen strength, however all system parts were functioning. The end caps for the specimen are made of a rigid corrosion – resistant material, with the same diameter. Additionally a self – aligning seating provided placed between both the top cap and loading ram.

Since the multistage test is more effective than performing three individual single stage tests on a homogenous sample separately, this avoids and reduces time to prepare a specimen; additionally it also means that only one latex membrane is used, and an easier representation of the stresses on the specimen is represented. A cylindrical high – density natural latex/rubber membrane surrounds the specimen, and provides resistance against leakage from the cell fluid. The un–stretched internal diameter is not less than or larger than 90% of the specimen diameter. The length is sufficient enough to cover the specimen and both end caps. The membranes thickness is 0.3 mm; this cannot go beyond 1% of the specimen diameter (1 % of 38 mm is 0.38). This can expand in conjunction with the specimen changing shape and on the specimen diameter. If the soil contained angular or coarse particles, the introduction of two or more membranes separated by silicon grease is considered. The membranes are thoroughly checked for imperfections and damaged ones removed. A membrane stretcher of suitable diameter to the specimen is also provided, along with an o – ring stretcher. Two rubber o – rings of suitable 38 mm diameters are added to the specimen, which seals the end of the latex/rubber membrane to the top cap and base pedestal (bottom) cap. These are un–stretched, with diameters between 80 % to 90 % of the specimen diameter. These are additionally free from flaws and necking.

The triaxial cell is then pressurised with any final adjustments made. The water pressure in the cell is modified to the desired value with the loading piston reserved by the load frame. The loading machine is then adjusted to bring the piston within a few millimetres of its seating on top if the specimen top cap. The reading of the force - measuring device is recorded during steady motion as the initial reading develops. The machine is further adjusted in order to bring the piston into contact with the seating on the top cap.

The reading of the axial deformation gauge is noted, and the scale of the gauge is made to zero so that the axial compression is registered directly as the reading of the gauge. The test is then started, and the readings are taken up to the point of failure. This device does not have a system able to determine the stress/strain data at regular intervals, so this is done manually by observation. The impending failure is localised through visual observation. This denotes the shear stress and the values of strain at that the point prior to failure. The cell pressure is then increased to the next pressure value and readings are again taken in the same fashion without terminating the machine. This is then repeated when the next maximum stress is indicated through visible signs in the specimens shape.

After completing the test on the chosen confining pressures, the axial force is disengaged from the specimen and machine platen lowered so that top cap has clears the ram. The machine is then restarted in the upward direction, and the reading from the axial force dial gauge is documented when it is stable. The cell pressure is reduced corresponding to each of the pressures used, this process is then repeated with each pressure, started by the restart of the machine. The cell pressure is reduced to zero, and the water used to create the confining pressure is drained out from the loading cell, the cell is then dismembered and the specimen is then removed. The membrane is then taken off and the specimen is then drawn by hand at its failure state. All of these parts are done progressively and effectively to reduce any loss of moisture content.

4.6 Tests Performed

A series of unconsolidated undrained TC tests were performed on the cohesive materials with different test parameters of confining pressure and moisture content.

Table 4.1 presents the tests performed by their designated name along with the mass of the sample tested, specimen dimensions (mm). The designation of the names is as follows: category (C_X1 = Clay X1; C_X2 = Clay X2; C_Y1 = Clay Y1; C_Y2 = Clay Y2), and the confining pressures (25 = 25 kPa; 30 = 30 kPa, 45 = 45 kPa, 50 = 50 kPa; 60 = 60 kPa, 70 = 70kPa).

Table 4.1 Summary of experiments performed.

Specimen Ref.

Confining

Pressures

(kPa)

Specimen

Dimensions

Diameter (mm)

by Height (mm)

Mass

(g)

Bulk Density

(Mg/

m³)

Dry Density

(Mg/m³)

W –Moisture Content

(%)

C_X1

25

38.0 by 74.7 200 2.13 1.78 2050

75

C_X2

30

38.0 by 77.9 200 2.11 1.76 2045

60

C_Y1

25

37.8 by 75.2 200 1.99 1.54 2950

75

C_Y2 30 37.9 by 78.2 200 1.96 1.52 29

45

60

CHAPTER 5 – EXPERIMENTAL RESULTS

5.1 Introduction

In this chapter, the results of the quick undrained triaxial compression experiments that where conducted to investigate the relationship between confining pressure and other parameters on the strength properties of cohesive materials (i.e. clays). As aforementioned, two triaxial compression tests were executed on two different clays for each combination of test conditions (moisture content, density, material type, and confining pressure) to check the likelihood of repeatability. These results are produced from data to graphical format and are presented in the Appendices as the relevant data regarding the specimen, the pressure parameters and using Mohr’s circles to determine the amount of cohesion present in the specimen, the shear strength, along with the angle of shearing resistance.

5.2 Data Obtained

The results of the QU tests for the four different types of clay soil at the different confining pressures are shown in the Appendix as Figure A.1 to Figure A.4. This additionally includes a summary sheet of the data obtained from each specimen as Figure A.5. Figures A.1 and A.2 are data sheets representing the statistics obtained from specimens C_X1 and C_X2. Figures A.3 and A.4 then represents the same data obtained for specimens C_Y1 and C_Y2. These data sheets additionally contain the determined Mohr circles for that particular specimen calculated for the multistage tests, with the values of angle of shearing resistance ϕ taken from the best common tangent of all circles, the undrained shear strength cu, and the value of cohesion c. These are presented in the form of Shear Strength over Confining Pressure. These Mohr circles were determined directly from the Axial Strain (%) vs Deviator Stress (σ1 – σ3, kPa). Multistage testing concluded when the failure of the specimen was imminent. This was determined by observing when the deviator stress tends a maximum value. The specimens were not subjected to excessive deformation, particularly during the early stages of loading. The specimens were inclined to develop distinct shear failure modes, and the strength could have possibility been reduced from its peak strength. The shear strength measured at the successive stages may have contributed towards an ultimate or residual strength condition. The ultimate (residual) shear strength condition was obtained when the deviator stress has levelled off after reaching its maximum value.

5.3 Effects of Confining Pressure and Undrained Shear Strength

The results shown in the graphical format in Figures 5.1 and 5.2 show a relationship exists between the undrained shear strength. The graphs all show a proportional relationship between these two parameters, where the shapes of the curves is non – linear and mostly parabolic. Due to increasing the confining pressure through a series of increments, the undrained shear strength demonstrates the ability to increase to the change in pressure. In some cases, the changes appear diminutive. In other cases, it shows the doubling of the pressure from 30 kPa to 60 kPa in the C_X2 and C_Y2 specimens, where the strength increases gradually. In Figure 5.2 for specimens C_Y1 and C_Y2 the maximum shear strengths for both specimens is the same for both the second and third confining pressures (Stage 2 and Stage 3). This is either anomalous or an indication that the shear strength is relatively unchanged in a more saturated sample or that this remains unchained in a sample that is closer to that saturation point, when confining pressures are increased. Generally if the clay has less moisture within it, it tends to provide greater undrained shear strength before impending failure. Figures 5.1 and 5.2 demonstrate that the shear strength increases with an increasing confining pressure; that these values of shear are considerably larger, almost three – fold in pressure. This undrained strength is also dependant on the pore – water pressure. The matrix suction in the clays decreases with a rise in the degree of saturation, which is also in attachment of a reduction in the volume. The two sets of identical clay specimens are brought to different initial states because of the changes in the pore –

water pressures under the undrained loading conditions. Clays X1 and X2 are considered more unsaturated (moisture content ‘w’ is < 20%) than C_Y1 and C_Y2; therefore, in the undrained loading conditions for unsaturated soils, the rise in the shear strength caused by a rise in the confining pressure is mostly greater than the reduction of the shear strength coupled with the decrease in the matrix suction. The matrix suction changes in an unsaturated soil, under undrained loading conditions, are comparable to the changes in pore – water pressures in saturated soils that are under similar undrained loading conditions. The volume change in unsaturated soils under undrained loading is largely due to the compression of air. The undrained pore pressures are assumed to be generated immediately after loading.

20 30 40 50 60 70 8089

90

91

92

93

94

95

96

97

98

Stage 1

Stage 2

Stage 3

Shear Strength over Confining Pressure, C_X1

Shear StrengthConfining Pressure σ3 (kPa)

Shea

r Str

engt

h Cu

(kPa

)

(a)

25 30 35 40 45 50 55 60 6572

74

76

78

80

82

84

86

Stage 1

Stage 2

Stage 3

Shear Strength over Confining Pressure, C_X2

Shear StrengthConfining Pressure σ3 (kPa)

Shea

r Str

engt

h Cu

(kPa

)

(b)

Figure 5.1: Graphs of Shear Strength over Confining Pressure of Specimen C_X1 (a) and Specimen C_X2

(b)

20 30 40 50 60 70 8018.4

18.6

18.8

19

19.2

19.4

19.6

19.8

20

20.2

Stage 1

Stage 2 Stage 3

Shear Strength over Confining Pressure, C_Y1

Shear StrengthConfining Pressure σ3 (kPa)

Shea

r Str

engt

h Cu

(kPa

)

(a)

25 30 35 40 45 50 55 60 6518.4

18.6

18.8

19

19.2

19.4

19.6

19.8

20

20.2

Stage 1

Stage 2 Stage 3

Shear Strength over Confining Pressure, C_Y2

Shear StrengthConfining Pressure σ3 (kPa)

Shea

r Str

engt

h Cu

(kPa

)

(b)

Figure 5.2: Graphs of Shear Strength over Confining Pressure of Specimen C_Y1 (a) and Specimen

C_Y2 (b)

Figure 5.3: Annotated Mohr Circles of Specimen C_X1 and Specimen C_X2

Figure 5.4: Annotated Mohr Circles of Specimen C_Y1 and Specimen C_Y2

Figures 5.3 and 5.4 depict the diameter of the Mohr circles increasing with an increase in confining pressure. These diagrams additionally demonstrate, with annotation, the impact of the degree of saturation on the specimen shear strength.

Unsaturated Approaching Saturation

Unsaturated Approaching Saturation

Unsaturated

Unsaturated

Figure 5.3 represents the specimens C_X1 and C_X2 Mohr circles. The undrained conditions work differently to the specimens of C_Y1 and C_Y2, as the diagrams show the Mohr circles are spread wider apart. Similarly, observations demonstrated that the undrained shear results from the remoulded specimens at their natural moisture contents did show an appreciable difference from the specimens that were more or less saturated. Both Figures 5.3 and 5.4 show a clear transition from an unsaturated zone towards an approaching saturation point. Both Figures 5.3 and 5.4 indicate an unsaturated zone. This zone is more apparent in specimens C_Y1 and C_Y2, more pronounced in C_Y2 as the Mohr circles are better spread demonstrating a far clearer transition. C_Y1 and C_Y2 have greater saturation due to higher moisture content, therefore the representative Mohr circles show this change more clearly than of the previous two Clay X specimens. An approaching saturation point is less evident than that of C_X1 and C_X2. This also demonstrates not just the relationship between the undrained shear strength and confining pressure but the also the degree of saturation as parameter in triaxial compression and how this affects the overall expected profile for undrained triaxial tests.

The plotted envelope identifies a curved relationship between the shear strength and total normal stress for unsaturated soils tested under undrained conditions. Once the soil becomes further saturated under the application of confining pressure, a horizontal envelope develops with respect to the shear strength axis. Under saturated conditions, where a single stress state variable controls the strength, an increase in the confining pressure is equally balanced by a pore – water pressure increase. The total stress remains constant in spite of the applied confining pressure, σ3. Once the clay is saturated like C_Y1 and C_Y2, the shear strength behaviour acts like the ϕu = 0 model. In these tests, the relationship for angle of shearing resistance is ϕu ≥ 0 model. The pore – pressure changes due to the application of deviator stress were commonly neglected for these undrained tests and generally for unsaturated soils. Specimens C_Y1 and C_Y2 fit the profile of QU test, because the shearing resistance angle is closer to zero, which is what is expected of QU tests. The cohesion has relatively low values in C_Y1 and C_Y2 and tends to increase continuously at an increases rate that is much more pronounced than that of the angle of shearing resistance (internal friction angle). The friction angle is higher in values of lower cohesion, suggesting that the less cohesive the material, the greater the amount of shearing resistance in the material (Chapter 2). This effect is demonstrated in both C_X and C_Y specimens, in which the cohesion departed from 83 kPa to 54 kPa where the angle of shearing resistance increased from 5° to 13° in the X tested specimens. The same effect, albeit almost insignificant without other tests to prove, illustrate the same trend. In specimens C_Y1 and C_Y2, the shearing resistance angle increases from 2° to 3.5° when the cohesion was 17 kPa and decreased to 16 kPa. In order to determine this relationship, if one exists, is to determine the cohesion and angle of shearing of resistance of every Mohr circle. The deviator stress increases with the axial strain almost continuously. The total stress increment Δσ2 equals Δσ3. The development of pore pressures in the undrained triaxial test is influenced both by the total stress increment, Δσ3, and from the change in the deviator stress, Δ (σ1 - σ3).

Now that the values of undrained shear strength su have been obtained for each clay, the data can be used in the application of determining the Ultimate Bearing Capacity of shallow foundations as mentioned in Chapters 1 and 2 (Skempton, 1951, [38]) using the expression below.

Ultimate Bearing Capacity

qf = (su .Ncu) + qo Equation (5.1)

Ncu = Skempton's Bearing Capacity Factor, determined via a chart or using the following expression:

Ncu = (Nc.sc.dc) Equation (5.2)Where sc is a shape factor and dc is a depth factor.

c = Apparent Cohesion intercept, qo = γ. D (Density × Depth) D = Founding Depth, B = Foundation Breadth γ = Unit Weight of removed Soil.5.4 Specimen Mohr Circle Analysis

Figure 5.5: Mohr’s circle to determine the shear strength τ f and the normal stress σ f at failure.

The condition of the failure of the specimen is generally approximated to by a straight line drawn as a tangent to the circles (best common tangent), the equation of which is t = c + σn tan ϕ. The value of cohesion (c) is read of the shear stress axis, where intersected by the tangent to the Mohr circles, and the angle of shearing resistance (ϕ) is angle between the tangent and a line parallel to the shear stress.

Equation (5.3) Equation (5.4)

Equation (5.5)Equation (5.6)

Table 5.1: Determination of Shear Strength and Normal Stress at Failure.C_X1 C_X2

Stage 1 Stage 2 Stage 3 Stage 1 Stage 2 Stage 3q (kPa) 184 191 194 151 163 168σ1 (kPa) 209 241 269 181 208 228σ3 (kPa) 25 50 75 30 45 60Radius R (kPa) 92 95.5 97 75.5 81.5 84Midpoint MP (kPa) 117 145.5 172 105.5 126.5 144Cu (kPa) 92 95 97 72 76 84f at Failure (kPa) 91.650 95.137 96.631 73.565 79.411 81.847σf at Failure (kPa) 108.98 137.18 163.55 88.516 108.17 125.1Cohesion (c) (kPa) 83 54ϕ * (Radians) 0.0872665 0.2268928

C_Y1 C_Y2Stage 1 Stage 2 Stage 3 Stage 1 Stage 2 Stage 3

q (kPa) 37 40 41 37 40 41σ1 (kPa) 62 90 116 67 85 101σ3 (kPa) 25 50 75 30 45 60Radius R (kPa) 18.5 20 20.5 18.5 20 20.5Midpoint MP (kPa) 43.5 70 95.5 48.5 65 80.5Cu (kPa) 19 20 20 19 20 20f at Failure (kPa) 18.489 19.988 20.488 18.465 19.963 20.462σf at Failure (kPa) 42.854 69.302 94.784 47.371 63.780 79.249Cohesion (c) (kPa) 17 16ϕ * (Radians) 0.0349066 0.0610865

Usually in Quick Undrained tests, half the deviatoric stress, q, is taken as the undrained shear strength (cu), R = σn; therefore σn is taken as the maximum R and maximum possible shear strength, via Stage 3 of every test.

Since this is a multistage test, not every value of cohesion (c) and angle of shearing resistance (ϕ) has been determined for every Mohr circle. Therefore, the assumption is made that the angle of shearing resistance value is an average via the tangent of all the Mohr circles in that test and that the value of c is assumed as average across all the Mohr circles for that test. The angle of shearing resistance is taken in order to determine the shear strength at failure and normal stress at failure in calculation for each Mohr circle of that specimen. However, a best common tangent is useful in obtaining a representation of stresses in the entire specimen.

Softer more – saturated clays are easier to predict as they tend to follow the UU profile, drier less – saturated clays tend to perform almost CU type of profile in UU tests. Both sets of clays exhibit elastic distortion, or bulging, rather than complete plastic deformation. Considering the C_X1 and C_X2 specimens, they are noticeably uneven and have behaviour that is considerably variable then usually expected. From this, the assumption that less saturated more brittle clays have greater angle of shearing resistance is considered. This could surmise a general statement that the drier the material, the greater the angle of resistance.

With the data of C_X1 and C_X2, judging based on the Mohr circles, their profiles fit the type of graphical output expected of a CU test, or perhaps a sample much less saturated then considered that it deviates from the type of relationship expected of a QU test. This means both C_X1 and C_X2 were either allowed to initially consolidate, or that the two specimens are far less saturated and that this these tests reflect that parameter in the form of a higher shearing resistance angle, a more brittle like material, and more variable QU test results. In addition, the degree of saturation (moisture) is most likely to be a parameter that affects the shear values of C_X1 and C_X2 since the Mohr circles are not similar. This is only considered under the assumption that saturation and moisture parameters could affect the results of C_X1 and C_X2, along with the rate of application of the confining pressure (consolidation prior to test) being a lot slower due to the nature of Clay C_X. In QU/UU tests, the rate of consolidation is faster; where in CU tests, the consolidation is applied slowly. QU/UU tests are also a good application of testing almost or fully saturated samples if determining the undrained shear strength. C_X1 and C_X2 are far less saturated than the other clays in this test, therefore it is concluded that the degree of saturation was a varying factor in the determination of the shear stresses of C_X1 and C_X2 and this additionally changed the expected UU profile. To determine this relationship further, more tests would need to be performed in order to examine the rate of consolidation and the degree of saturation (moisture).

The material tested in the C_X1 and C_X2 tests is silty clay. The reason for testing this material is first that the properties of silty clay are not yet fully understood, unlike pure clay and sand, which represents the two extremes of soil behaviour due to their composition and the mechanical properties. Silty clay is very common therefore, it is often necessary to relate to this material during foundation construction works. It is also appropriate to obtain information regarding stresses in silty – clay type of cohesive materials as they found in concentration during foundation design if present when drilling to certain depths or applying foundations through that cohesive layer. This still underlines that the overall behaviour of silty clay requires better understanding when considering shallow foundation implementation. With these values of undrained shear strength in cohesive materials, a comparison is need of

Table 5.2: Undrained (Immediate) Shear Strength of Cohesive Soils (BS 8004, 1986, [47])

ConsistencyUndrained (Immediate) Shear Strength

In accordancewith BS 5930

Widely Used Field Indications (kN/m2)(Kgf/cm2)(tonf/ft2)

Very Stiff Very Stiff or Hard Brittle or very tough > 150 > 1.5

StiffStiff

Cannot be mouldedin the fingers

100 to 150 1.0 to 1.5

Firm to Stiff 75 to 100 0.75 to 1.0

FirmFirm

Can be moulded in the fingers by strong pressure

50 to 75 0.5 to 0.75

Soft to Firm 40 to 50 0.4 to 0.5Soft Soft Easily moulded in the fingers 20 to 40 0.2 to 0.4

Very Soft Very SoftExudes between the fingers when squeezed in the fist

< 20 < 0.2

Table 5.2 permits comparisons between the undrained values obtained in the triaxial testing of this study and the values pre – determined for cohesive materials in British Standard. This table shows the range at which the determined undrained shear strength of Clay X and Y lie according to their immediate shear strength. The undrained shear strengths obtained are comparable to the lower half of the table in the Firm and Soft clay region. Clay X specimens demonstrate to be firmer and stiffer with greater values of shear strength and then the Clay Y specimens. This additionally suggests that firmer and stiffer cohesive materials are less – saturated in general practice of tests and approach an almost brittle condition. These make them more suitable to be encountered with when preparing a shallow foundation.

Table 5.2 and Table 5.3 below; demonstrate the necessity to prove that a foundation is safe from shear failure by taking a presumed bearing value that does not exceed the ratio of the ultimate bearing value over a suitable safety factor. This means that the ratio should be in the order of 2 to 3 and no higher. The likelihood of plastic deformation must be checked, when preparing and discovering a layer of softer clay like Clay X and Clay Y at a depth below the foundation. Additionally, the net increase of pressure in the soft clay due to the loading of the foundation should be addressed, as this net increase of pressure is a limited value that will be a satisfactory safety factor against soft – clay layer shear failure. A vital requirement to certify is that the total and differential settlements of the planned foundations are not too large. Once the presumed bearing value has been assessed, the settlement beneath this pressure is approximated, if too great the foundations will need to be redesigned by reducing the bearing pressure or increasing the depth of the foundation to an area of low settlement. If the statement that Clay X consolidates slower in the triaxial test is assumed correct, it makes it more applicable to the types of clay layers encountered in real construction situations. The extent of settlement occurring when foundation loads are applied to the ground is dependent on the structure rigidity, duration and type of loading, and deformation characteristics of the ground soil. Consolidation settlement in clays and silts can happen for long periods after the structure is finished. The rate of drainage from the voids under influence of the stresses applied is slow; therefore, allowances for slow consolidation settlements must be considered. Founding a structure on a peat or organic soil may be inevitable, therefore consideration on the secondary consolidation mentioned above gives rise to more worrying settlements that can be long – lasting. This settlement is dependent on bearing pressure, compressibility of soil, foundation dimensions, depth prescribed, shape and width. Wide foundations settle more than narrow foundations in non – cohesive soils with the same applied bearing pressure. The allowable bearing pressure on wider foundations is less than that of the bearing pressure allowed on narrow foundations, to keep settlements within suitable limits.

Settlement degree of foundations on cohesive soils can be determined in laboratory tests, where consolidation characteristics and Young’s Modulus are determined. Elastic constraints from in – situ test are usually preferred rather the values from laboratory testing. Estimating settlement is

traditionally based on oedometer tests (one – dimensional consolidation) (Terzaghi, 1943, [43]; Skempton, 1951, [38]). This method overvalues settlement rates in soils that have greater horizontal permeability than vertical permeability. Foundations must carry down through peat and organic soil to a consistent bearing layer below. Peat and organic soils are highly compressible, where lightly loaded foundations are subject to considerable settlements over a long period. Soils such as these are not suitable for carrying the loads from larger structures, and traditional lowering of the groundwater produces a substantial and prolonged settlement.

5.5 Deformational Changes at Failure

In this section, all tested specimens are used to investigate the effect of loading rate, and confining pressure on shear failure mode of the cylindrical clay specimens, the development of shear bands, and the relating the deformation to expected shallow foundation failure modes. In Figures 5.6 and 5.7 are the drawings of the specimens after their respected failure modes were observed at the same axial loading rate and after a set of identical confining pressures. These are presented in two dimensions and three dimensions to evaluate the shear bands visible on the front surface of specimens. The strains at failure for each specimen as follows: C_X1, 12.7 %, 16.1 %, 20.7 %; C_X2, 16 %, 16 %, 19.9 %; C_Y1, 14 %, 17.3 %, 20.6 %; and C_Y2, 12.2 %, 15.3 %, 19.2 %. All specimens underwent strain at failure up from approximately 12 % to 21 % per Stage at each confining pressure. The strain at failure (%) increases as a result of increasing each confining pressure (kPa). This means the relationship is again non – linear and parabolic.

Shear bands were clearly observed in all specimens, as the local shear strain did not exceed 100% near the shear plane. Bulging or barrelling failure mode seems for common for slower loading rates than that of buckling failure mode in faster loading cases. In addition, these specimens confirmed the behaviour of general and local shear failure patterns expected, which is given clear mention in Chapters 1 and 2. The drawing of specimens C_X1, C_X2, C_Y1 and C_Y2, show shear bands acting on the surface of the specimens and in some instances crossing each other with a shape similar to an ‘X’.

Clays, like many other materials tend to exhibit elastic to plastic deformation when subjected to extreme loading and confining conditions. The clays being tested in these triaxial compression tests are generally no different, however, due to slow loading rates to reduce the build up of pore – water pressure and low confining pressures, the behaviour observed is less defined then some extreme buckling cases and more complex. Clays usually behave as viscous – elastic materials, and in triaxial tests, they tend to change volume in order to compensate for the increasing axial load. The most common shape is bulging, in which the compression causes the specimen to compress to a shape that resembles a “bulge” structure, where the diameter in the centre of ten specimen exceeds the diameter of the specimen prior to testing. All tested specimens underwent bulging/barrelling because of a slow axial displacement rate. Most of the specimens were theorised to bulge like that in Figure 1.2 (a), where this mode is common in normally consolidated and saturated clays that are loaded quickly so that undrained conditions are overcome. The specimens are small in diameter and height, and the average axial displacement was 2.0 %/min, therefore the loading could be considered fast acting on a small specimen, compared to a larger specimen of perhaps 50 mm diameter and 100 mm height where this axial displacement rate would be considered slow. As the specimens tend to go bulging, the particular shape varies between two distinct shear failure modes. Therefore, it is stated that the specimens underwent diffuse shear failures as they indicate areas of general and local shear failure. Bulging phenomena is also more likely to occur in specimens of short diameter and length, and at low confining pressures as demonstrated by these tests. The confining pressures have less considerable influence on the outcome of the specimen shape after testing.

These drawings also indicate the shear around the surface but no real observation is made on the internal failure of the specimens. The transparent latex membrane sticks to the specimen due to the membrane correction pressure; therefore, it is removed in order to see the specimen fully. The

drawings have lines representing shear bands created by the compressive process in order to follow any changes in shape and the visible shear. This helps to determine the overall failure mode of each specimen in order to determine a relationship with the specimen and the conditions during the test. The figures demonstrate that Clay X is as susceptible to bulging as Clay Y, despite it being stiffer and drier silty clay. From these tests, the conclusion made is that the effect of similar confining pressures and similar axial displacement rates has no real affect on a short specimen of two different specimens.

The stress – strain relation and the process of shear bands formation for are demonstrated in subsequent figures below. The deviator stress continues to increase until an axial failure strain of the first pressure is reached, although strain localization is clearly seen at the second and third pressures. The cause of this is due to the dilatancy of over consolidated clay increases the deviator stress much more than the softening induced by development of the shear band on the side surface. In the specimens where bands intersect, when the strain behaviour starts, the second shear band generated intersects the first one. The second shear band becomes clearer than the other and more defined then those that have intersected. In this way, the stress - strain relations of the cases of specimen X and Y show strain softening corresponding to the generation of shear bands.

Each specimen showed local shear on the surface that continued to propagate after each increase in confining pressure. C_X1 and C_X2 demonstrated an almost brittle tendency with C_X2 demonstrating buckling in the lower half of the specimen. This states that drier or less moist, and therefore less saturated, specimens in these tests tended to undergo both bulging failure modes in C_X1 and in a particular case some substantial buckling failure modes in C_X2. In most cases, conclusions like this are not warranted without further tested specimens. Therefore, further tests would need to be performed in order to certify some general understanding about unsaturated samples. C_X1 and C_X2 are considered to be less – saturated or unsaturated in general as this quantifies the cohesive material more easily and allows more detailed explanations about their behaviour. Clays C_Y1 and C_Y2 are then considered to be saturated as their behaviour and their Mohr circle representation allows it to make it easier to compare between Clay X and Clay Y. Starting with C_X1, the specimen shows clear shear bands that act along the top right base of the specimen towards the bottom left base. The ‘V’ shaped shear band was formed as two individual shear bands at strains between 12.7 % and 16.1 % and propagated continuously till they intersected in the centre of the specimen. The other two shear bands of smaller length at the top and bottom of the specimen formed between the second and third confining pressure. Specimen C_X2 experienced local bulging in the bottom half of the specimen, not uncommon, yet a clear shear band that shows a behaviour of a buckling nature. The large almost horizontal shear band spread across the specimen considerably during the changing of the first to the second confining pressure and then propagated to a limiting amount after the third pressure was introduced. The almost 45° degree band of shear to the top part of the specimen additionally lengthened during the change in confining pressure

(a) (b)

Figure 5.6 (a): Deformation changes at failure of C_X1 specimen. (b) Deformation changes at failure of

C_X2 specimen.

The axial displacement rates are the same for each specimen, the overall shape changes seem to be equivalent to each other, and therefore it appears that the loading rate has genuine effect on the behaviour of the clays. Although the peak strengths are almost same between two specimens of C_Y1 and C_Y2, this additionally identifies that a slenderer specimen shows the smaller residual strength. The decrease of deviator stress is observed from the more saturated specimens C_Y1 and C_Y2. These specimens have deviator stresses considerably less than that of the Clay X specimens. Clay Y has a greater degree of saturation and these specimens; especially C_X2 has an area of local shear failure mode. Additionally demonstrating the behaviour in stress – strain compared with Clay Y specimens, which are observed to have more common ‘X’ shear bands. These short specimens show the more complex failure mode among the same shape. Two shear bands occur that intersect in the C_Y1 and C_Y2, where in case of C_Y2, two shear bands resemble an inverted ‘Y’. Figure 5.7 (a) shows the C_Y1 specimen, again with the additional lines on the drawing to represent the shear bands. The specimen yet again demonstrates the bulging phenomena. In 5.7 (b), C_Y2 specimen has a less uniform bulging effect, with more considerable buckling failure or compression on the centre of the specimen towards the right.

(a) (b)

Figure 5.7 (a): Deformation changes at failure of C_Y1 specimen. (b) Deformation changes at failure of

C_Y2 specimen.

Currently, triaxial compression tests using rectangular clay specimens can be used to discuss the three – dimensional failure behaviour of clay, where the local shear strain distributions are determined by image analysis (digital photography) taken on a side surface of specimen during the triaxial compression test. This study deals with the traditional cylindrical specimens with the traditional representative drawing that can offer reasonable insight and explanation for their mode of failure due to the loading rate and confining pressure. However, no depth into strain distribution and strain localisation within the specimen. All specimens showed both these relationships and that the strains at failure never seems to exceed 21 %.

Additionally, a slenderer specimen would show behaviour that is more unstable in stress – strain relationships. The failure mode of a slender specimen tends to looks likes buckling of column. Where, as aforementioned, these specimens exhibit the unique bulge or barrel – like mode of failure. The specimens tested are shorter in length and diameter and are inclined to show a more complex shear failure pattern as the stress – strain relationship is which demonstrates more stability in shorter specimens.

With the clay specimens now tested, the Figures of 5.6 and 5.7 show shear bands common in the case of a slow loading rate and low confining pressure, particularly more visible in C_X1 and C_Y2. According to the failure modes that occur in these specimens, the stability of ultimate failure load fluctuates. The slower loading rate used in these specimens shows more stable behaviour due to the failure mode without buckling. There are different failure modes observed under the same test conditions such as the same loading rate and same shape of specimen. There is also a level of uncertainly as to whether large deformation and failure behaviour using the cylindrical specimen can directly link to the practical problem for foundations and whether rectangular specimens should be

investigated to understand clear three – dimensional behaviour. However, it was very important that the minute observation of the typical surface strain localisations and shear patterns were observed in the specimens in order to improve the accuracy of prediction of progressive failure.

The specimens additionally indicate that a set of significantly higher pressures on these specimens could yield in more indicative shear lines and planes of failure and perhaps more buckling – type of deformation observed with slender specimens (Height/Diameter ratio > 2). It is noteworthy to mention that the bulging/barrelling phenomenon directly shows similarities to the elastic bulge formation of general shear in Figure 1.2 (a) and local shear phenomena (b) (Coduto, 1994, [10]). Here the lines show the state of the elastic distortion caused by static loading acting on a shallow foundation in an area of depression. For purely cohesive soils, the local shear failure may be assumed to occur when the soil is soft to medium.

In considering preliminary design purposes, the standard BS 8004 gives presumed bearing values that are the pressures, which would usually result in a sufficient safety factor against shear failure for particular soil types, through no consideration of settlement.

Table 5.3: Presumed Bearing Values under Static Loading (BS 8004, 1986, [47])

Category Types of rocks and soils Presumed bearing value (kN/m²)Cohesive soils Very stiff bolder clays & hard clays 300 to 600   Stiff clays 150 to 300   Firm clay 75 to 150   Soft clays and silts < 75   Very soft clay Not applicable

As highlighted by the cohesive soils in this table to the cohesive clays investigated in this study, the presumed bearing values of cohesive soils increases as the clay becomes stiffer and firmer. For soft clays and silts investigated in this study, the bearing values are in the range indicated in bold in Table 5.2. Clays are far more interesting in this manner as they have presumed values over range, rather than dense or loose sands and gravels that have clear ranging values of presume bearing. This additionally means that clays are susceptible to long – term consolidation settlement. The general principles concerning bearing pressure and settlement characteristics are expressed in Chapters 1, 2.3 and 5.4. This additionally means that certain deep foundations would be inappropriate to use in areas where the bearing values were less than 150 kN/m2, whereas shallow foundations could be made to work sufficiently in areas of sub – standard cohesive materials.

5.6 Effects of Moisture Content

A relationship has been determined between the degree of saturation and undrained shear strength, therefore a similar relationship is determined directly with the amount of moisture present in the specimen, as both saturation degree and moisture content work in unison. Since is realised that the shear strength will become greater (increase) and the conditions become safer with time as the pore – water is allowed to drain. These shear test results indicate that it is possible to improve the shear strength of clay soils by consolidation, if time is available for allowing the required pore – water drainage to occur. This increase in shear strength is because of the fact that during shear, the effective stress increases as the rate of shear increases and pore – water diverts from the plane of shear. This explains why the failure envelopes for undrained tests are above the failure envelopes for drained tests as shown in the Mohr circle profile diagram Figure 2.2.

5.7 Theoretical / Actual Results Comparison

From the Literature review, with Figure 2.9, it is evident that not all theoretical estimations can correctly surmise the expected result of a triaxial test. It is also clear practice that it would more appropriate to do triaxial tests more often in order to obtain better general information on the clays use for foundation construction. There should be no dependency to rely heavily on theoretical predictions made by Mohr circle and they should only be taken as estimates and not actual representative values of the stresses in specific soils. Actual laboratory and in – situ testing is the best and most effective method to examine shear strength of a soil and soil behaviour directly.

5.8 Degree of Saturation on Shear Strength

As observed, the degree of saturation has effects on the strength and slightly the stiffness of the cohesive materials. The degree of saturation and stiffness are related properties that influenced the frictional strength and cohesion of the clays. The relationship of these two properties does suggest some cementing conditions of the material. An important factor controlling the strength of the clays appears to be pore – water. Despite the Y clays being heavily saturated, the X clays have very high matrix suction. Since the Y clays are saturated and subjected to reduction in effective stress, the strains associated with saturation and removal of suction lead to reduction in strength. However, the relation between pore – water pressure and strength needs further study before any certainties can be agreed upon.

CHAPTER 6 – CONCLUSIONS & RECOMMENDATIONS

6.1 General Conclusions

6.1.1 Confining Pressure Modification

The analyses indicated that the purpose of changing the confining pressure between an unsaturated and saturated material was significant in manipulating the value of shear stress. A clearly defined relationship does exist between the undrained shear strength and confining pressure, in accordance with other significant parameters such as shape, moisture content, and degree of saturation. The test results in table and graphical format was additionally helpful in order to demonstrate these relationships. In order to get a better representative state of the stresses in a specimen, the sample or specimen source should come from an in – situ location, in which to understand field conditions more than lab conditions.

6.1.2 Triaxial Testing Results

The results from the triaxial tests were very reliable as this method is more effective in determining undrained shear strength values than the conventional Unconfined Compression test. The actual results varied compared to theoretical predictions but this strengthens the notion that the reason for testing these materials is firstly that the properties of clay are not fully understood, and further analysis is needed on a larger amount of samples. These specimens additionally showed that the clay is slightly anisotropic with respect to strength.

6.1.3 Shear Strength and Normal Stress at Failure

The values of undrained shear strength taken as half the maximum deviatoric stress using Mohr’s solution was found to variant to the actual the values of strength and stress at failure. In either case, the determination using Mohr’s circle can be taken as a maximum value. The difference in calculations could be to the occlusion of minor errors in procedure. For the values of normal stress, it would be wiser to use the values of obtained at failure as this is shown to be more reliable than taking the midpoint of the Mohr circle, which again could be taken as a maximum value. The values obtained are for the indication of total stress, which are limited in studying soil mechanics as they provide the drainage and strength properties of that soil in the short – term rather than the long – term.

6.1.4 Saturation, Moisture Content and Shear Strength

It was determined that both the degree of Saturation and the amount of moisture present in the specimen had a considerable effect on the shear strength of the specimens. The more saturated the sample, the more constant the shear strength value for each Mohr circle, and the smaller the angle of shearing resistance (internal friction angle). This also means that the strength is related not just to the amount of moisture in the specimen, but to the amount of saturation. The level of moisture and saturation evidently has an impact on the amount of apparent cohesion. More moist or ‘wetter’ samples have less apparent cohesion present.

6.1.5 Specimen Shape

It was observed that the specimens tested, short in length and diameter, are more inclined to show a more complex shear failure pattern (barrelling) as the specific stress – strain relationship in each clay

is more stable for shorter specimens. In fact, there is no essential difference between the different cases is shown in Figure 5.6 and 5.7, which presents the shape of the four specimens in the final state. In among the four tests, the specimens remained solid after the unstable deformation phase, retaining the geometry of the final state when the testing machine was stopped.

6.2 Recommendations

Based on the findings of the experimental work performed for this study, the following recommendations are presented:

• Increase the quality control in the moisture procedure in order to obtain uniformity in the moisture content of the specimens.

• Perform conventional triaxial compression tests with a mixture of different clays with modified moisture contents to compare the difference in particles and the mixture.

• Additionally provide further investigation into the effect of slope stability and other engineering predictions.

• All future tests should be conducted at a strain rate of 0.1% or lower.• Select confining pressures of 0.5vs, vs, and 2vs (vs – maximum vertical stress)• The tests that had leaks should be conducted again to verify the adjustments made for the leaks

were correct.• Specimens should be consolidated prior to triaxial shear testing in order to determine if the

process improves shear strength.• Conduct tests on a wider range of clays and silty materials to investigate the trends in shear.• Conduct tests on rectangular specimens of different diameter to height ratios to investigate the

effect of specimen shape under triaxial compression.• Conduct verification tests using a non-compressible or “dummy” specimen in order to verify the

compression corrections and volume change measurements.• Conduct verification tests using an easily compressible “dummy” specimen to verify the minimal

effects of the vertical load.• Conduct tests with the horizontal direction as the major principal axis.• Addition of LVDT’s (Linear Variable Differential Transducer) to the other side of the pressure

cell to better follow the specimen and shear band response.• Specimens of different densities may be tested to investigate the affects of density on the results.• Tests may be conducted at different higher confining pressures to investigate how the increase in

confining pressure affects the results.• Software additions may be made, additionally to use a computer using the LabView or WinCLISP

software. • An automated – pressure – control device for the more complex loading conditions and varying

minor principal stress. • Perform Consolidated Undrained/Consolidated Drained (CD/CD) tests on clays, with more

samples, to investigate the effect of the testing conditions. • Use a more up-to-date Triaxial Testing system.• Measure the pore – water pressure as it helps to determine the effective stresses used for long-

term stability information.• Measuring pore – water pressure can additionally provide relationships between the drained shear

strength and confining pressure. These tests are performed at much slower speeds so suitable amount of time should also be considered.

• When varying moisture content in compaction process, to increase the accuracy of the test it is often advisable to reduce the increments of water in the region of the optimum moisture content.

• Focusing in a one type of soil condition and determine what is the foundation system that best suites the soil condition.

• Establish the relationship between the foundation system and the site soil condition.

6.2.1 Recommended Changes in Testing Procedure

• Document the reading of the axial deformation gauge. • Recording readings of the force gauge and compression gauge at a regular interval of the latter, so

that at least 30 – 40 sets instead of 15 sets of readings are taken up to the point of failure. • Readings should be at close intervals when approaching maximum load.• Plot calculator deviator stress against axial strain and force gauge reading against strain so that the

impending failure is localised.

6.3 Future Work

The following recommendations are made for future research and continuation of the present study:

• More research focusing in one type of soil condition and determine what kind of foundation system best suites that particular soil condition. Perhaps investigate in – situ soils in a construction site in order to better comprehend the function of the soils in real construction cases.

• A comprehensive study on the soil available in any country, the types of foundation system currently used, the factors governing the selection of a safe and economical foundation system, the design approaches adopted by the design engineers and lastly, the problems related to the foundation system.

• Investigate cases about the development of medium and high – rise buildings for commercial or residential purpose. The study would comprise of the site conditions, selection of foundation system, the foundation system used and some of the main issue in the cases. At later stage, comparisons between these case studies could be made.

• Determine the strength of soil at various depths below the foundation from the site investigation to find the safe bearing capacity.

• Establish the invert level of the foundation by either using the minimum depth below the ground level, which is unaffected by temperature, moisture content or by the depth of basement.

• Resolve the foundation area from the characteristics working loads and allowable pressure. This will determine the type or combination of types of foundation. A selection based on economic consideration, speed and ability to build.

• A building may, to the extent, collapse due to the failures in its foundation system. Therefore, it is important to identify the factors that can cause a foundation system to fail and also the ways to minimize these problems. The major cause for foundation failures especially in expansive soil is due to water. An investigation into the variation of water level can indentify settlement and disturbance problem.

• Establish the relationship between the foundation system and the site soil condition. Here an investigation could lead into the effects of a shallow/deep foundation shape and how this behaves in certain soil conditions.

• The water content in cohesive soils changes at which the consistency changes from one stage to the next stage. This is known the Atterberg Limits. An investigation into soil moisture content should be conducted as it is critical in drained conditions where there is permitted drainage in a foundation and if it contributes to settlement.

6.4 Summary

This study has summarised that mechanical behaviour of cohesive soil materials is largely influenced by the stress path to which it is subjected. In order to study the complex behaviour of cohesive soils in widespread stress space, many single element – testing methods have been developed in past several decades noted in literature, which involve a variety of specimen shapes and loading/boundary conditions. Using these common testing methods, the cohesive specimens were subjected to similar stress paths (quick undrained triaxial compression tests) and the results were presented in this study. As a result, it remains essential to carefully understand the influence of specimen shape and

loading/boundary conditions while analysing the experimental data to use in geotechnical design. Essentially, the triaxial results showed the test is considered as a good method to investigate the undrained soil behaviour, though some studies like this in particular have indicated that results of conventional triaxial tests are reliable only when measured soil strains are approximately larger than 0.001 (0.1 %). This study has additionally shown that soil - strength and the deformation behaviour is significantly influenced by the stress – strain behaviour of soils at small strain percentages, that is obviously much similar than those measured in the conventional triaxial test.

As aforementioned, modelling soil strength and other properties in the analysis and design of shallow foundations is very significant. The choosing of soil parameters for analysis and testing is , as shown in this study, difficult as there is a great deal of uncertainty as to which parameters are more pertinent to remain constant and which are good for variability. Since the properties of shallow and deep soil elements are quite different. The selection of a distinguishing stress – strain curve to represent a cohesive soil is necessary in design, but is difficult to settle on. In the case of design purposes of square or circular footings in homogenous cohesive soils, displacements produced can be assumed to be inhibited by the average soil stiffness in the deformation zone. Additionally, an extension of bearing capacity theory should include the plastic deformation mechanisms with distributed plastic strains. This can provide a unified solution for most design problems. This application is unlike the conventional applications as it can please approximately both safety and serviceability needs with prediction of stresses and displacements under working conditions.

As the kinematics, characterisation is concerned, all four clay specimens experiences bulging modes of failure and no real strain localisation was observed. A more detailed method of the triaxial system is needed to investigate the conditions where there may be a loss of resistance or total liquefaction. It is also difficult to identify whether a mode of failure is localised or dispersed. The specimens displayed general and local shear failures of that stated in Chapters 1 and 2, again, the triaxial test does not always provide clear result for a failure mechanism, as the observation of the specimen failure is limited to the outer boundary or surface state. Strain localisation is in fact hidden in such test conditions. Failure modes can be more easily featured in additional tests using biaxial compression.

Engineering and mechanical properties of soil at a construction site will influence the design and construction of shallow foundation systems. This study has stated how critical an investigation of soil strength properties are essential in modern day construction and the wide variety of soil testing methods available nowadays. The soil testing available that investigate soil properties and strength are triaxial test, particle size distribution, moisture content, Atterberg Limit test, soil chemical test and others. Considering the buildings erected nowadays, they are either medium or high – rise, and therefore deep foundations such as piles is recommended. Therefore, more studies into this should be carried out using more complex loading scenarios and with more non – uniform materials. Piles transfer loads to a greater depth, which work better against shearing resistance for greater support. The choice of a safe and cost effective foundation system for a particular structure is governed by many factors of the site itself, particularly the soil strength. These factors should be greatly considered to avoid damage and other harmful attributes and shall not be omitted just to make the design and construction of foundations faster.

APPENDIX A

Figure A.1: Quick Undrained Triaxial test results for X1 Clay Specimen

Figure A.2: Quick Undrained Triaxial test results for X2 Clay Specimen

Figure A.3: Quick Undrained Triaxial test results for Y1 Clay Specimen

Figure A.4: Quick Undrained Triaxial test results for Y2 Clay Specimen

Figure A.5: Quick Undrained Triaxial summary results for all Clay Specimens.

READING LIST

A list of the references used throughout in the format specified by the Brunel University Ref System.

1. Arthur, J.F.R. (1988). Cubical Devices: Versatility and Constraints. Advanced Triaxial Testing of Soil and Rock, ASTM STP 977, R.T. Donaghe, R.C. Chaney, and M.L. Silver, Eds., American Society for Testing & Materials, Philadelphia, 1988, ISBN: 9780803109834, pages 743-765.

2. Aysen, A (1 Jan 2002). Soil Mechanics: Basic Concepts and Engineering Applications, Taylor & Francis; 1st Edition, ISBN: 978-9058093585.

3. Baldi, G, Hight, D.W, and Thomas, G.E (1988). A Re-evaluation of Conventional Triaxial Test Methods. Advanced Triaxial Testing of Soil and Rock, ASTM STP 977, R.T. Donaghe, R.C. Chaney, and M.L. Silver, Eds., American Society for Testing & Materials, Philadelphia, 1988, ISBN: 9780803109834, pages. 219-263.

4. Baldi, G (31 Dec 1991). Faults in Clays - Their Detection and Properties, European Commission, ISBN: 978-9282625163.

5. Berre, T (1981). Triaxial testing at NGI. Geotechnical Testing Journal, Vol. 5, 1982, No. 1/2, pages 3-17.

6. Bishop, A.W, and Henkel, D.J. (1962). The Measurement of Soil Properties in the Triaxial Test.

7. Bishop, A.W, and Henkel, D.J. (1 Jan 1976). The Measurement of Soil Properties in the Triaxial Test, Hodder Arnold; 2nd Revised Edition, ISBN: 978-0713130041.

8. Casagrande, A. and Shannon, W.L. (1951). Base course drainage for airport pavements. Proceedings of the American Society of Civil Engineers 77 (75): 1–23.

9. Chen, W-F. (15 May 2007). Limit Analysis and Soil Plasticity, J. Ross Publishing, ISBN: 978-1932159738

10. Coduto D.P, (1994). Foundation Design - Principles and Practices, Prentice Hall Inc

11. Coduto, D. P. (1999). Geotechnical Engineering Principles. Prentice-Hall, Inc., Upper Saddle River, New Jersey. pages 759.

12. Coduto, D.P. (2001). Chapter 6: Shallow Foundations- Bearing Capacity. Foundation Design: Principles and Practices. 2nd Edition, pages 170-206.

13. Curtin, W.G, Shaw, G, Parkinson, G.I & Golding, J.M, (1994).Structural Foundation Designers’ Manual, Blackwell

14. Das, B.M. (1995).Principles of Foundation Engineering. 3rd edition. PWS Publishing, Boston.

15. Donaghe, R.T, Chaney, R.C. and Silver, M. L. Published (31/12/1988). Advanced Triaxial Testing of Soil and Rock, American Society for Testing & Materials (ASTM), ISBN: 9780803109834.

16. Duncan, C.I. Jr. (1998).Soils and Foundations for Architects and Engineers, Second Edition; Kluwer Academic Publishers; Boston/London/Dordrecht.

17. Duncan, J.M, and Wright, S.G. (22 Feb 2005). Soil Strength and Slope Stability, John Wiley & Sons; 1st Edition, ISBN: 978-0471691631.

18. Fratta, D, Aguettant, J, and Roussel-Smith, L (14 May 2007). Introduction to Soil Mechanics Laboratory Testing, CRC Press; 1st Edition, ISBN: 978-1420045628.

19. Fredlund, D.G, and Rahardjo, H (Aug 1993).Soil Mechanics for Unsaturated Soils, Wiley – Interscience, ISBN: 978-0471850083

20. Germaine, J.T, and Ladd, C.C. (1988). Triaxial Testing of Saturated Cohesive Soils. Advanced Triaxial Testing of Soil and Rock, ASTM STP 977, R.T. Donaghe, R.C. Chaney, and M.L. Silver, Eds., American Society for Testing & Materials, Philadelphia, 1988, ISBN: 9780803109834, pages 421-459.

21. Germaine, J.T, Santagata, M.C. and Ladd, C.C (1999). Initial stiffness of A0-normally consolidated clay measured in the triaxial apparatus. Pre-failure Deformation Characteristics of Geomaterials, Jamiolkowski, Lancellotta & Lo Presti (eds), Balkema, Rotterdam, ISBN: 9058090752.

22. Germaine, J.T, Sheahan, T.C. , et al (31 Jul 2003). American Society of Civil Engineers, ISBN: 978-0784406595.

23. Head, K.H (1982). Manual of Soil Laboratory Testing. Vol. 1, Pentech Press, London, Plymouth.

24. Head, K.H (1982). Manual of Soil Laboratory Testing. Vol. 2, Pentech Press, London, Plymouth.

25. Holtz, R.D, and Kovacs, W.D (8 March 1981).An Introduction into Geotechnical Engineering, Prentice Hall, ISBN: 978-0134843940.

26. Jumikis, A.R (Mar 1967). Introduction to Soil Mechanics, Van Nost Reinhold U.S, ISBN: 978-0442041984.

27. Lacasse, S, and Berre, T (1988). Triaxial Testing Methods for Soils. Advanced Triaxial Testing of Soil and Rock, ASTM STP 977, R.T. Donaghe, R.C. Chaney, and M.L. Silver, Eds., American Society for Testing & Materials, Philadelphia, 1988, ISBN: 9780803109834, pages 264-289.

28. Lambe, T.W. (1951). Soil Testing for Engineers, John Wiley and Sons, Inc., New York.

29. Lambe, T. W. and Whitman, R. V. (1969). Soil Mechanics, SI Version. John Wiley & Sons, New York, New York, pages 553.

30. Liu, C (12 May 2005). Soils and Foundations, Prentice Hall Singapore; 6th Edition, ISBN: 978-0131973084.

31. Liu, C and Evett, J.B (12 Jun 2008). Soil Properties: Testing, Measurement and Evaluation, Prentice Hall; 6th Edition, ISBN: 978-0136141235, pages 5-51, 337-351.

32. McCarthy, D.F. (2007). Essentials of Soil Mechanics and Foundations: Basic Geotechnics, Pearson, London.

33. Osman, A. S., and Bolton, M. D. (2004). Plasticity based method for predicting settlement of shallow foundations, Submitted to ASCE Journal of Geotechnical and Geoenvironmental Engineering, for review.

34. Prandtl, L. (1921). Uber Die Eindringungsfestigkeit (Harte) Plastischer Baustoffe Und Die Festigkeit Von Schneiden. Zeitschrift fur angewandte Mathematik und Mechanik. Vol. 1, No. 1. pages 15-20.

35. Saada, A.S. (1988). Hollow Cylinder Torsional Devices: Their Advantages and Limitations. Advanced Triaxial Testing of Soil and Rock, ASTM STP 977, R.T. Donaghe, R.C. Chaney, and M.L. Silver, Eds., American Society for Testing & Materials, Philadelphia, 1988, ISBN: 9780803109834, pages 766-795.

36. Sowers, G.F (1962). Foundation Engineering, McGraw Hill Book Company Inc SP (2004) SP – Geotechnical Report Gue & Partners Sdn Bhd

37. Sowers, G.F. and Gore, C.E. (1961). Large Scale Preconstruction Tests of Embankment Materials for an Earth-Rock fill Dam. 5th International Conference on Soil Mechanics and Foundation Engineering. pp. 717-720.

38. Skempton, A.W (1951). The bearing capacity of the clay, Proceedings of Building Research Congress, vol. 1, pages 180-189.

39. Skempton, A.W, Bjerrum, L (1957). Contribution to the settlement analysis of foundations on clay, Geotechnique,Vol.7, No. 4,pages 168-178.

40. Skempton, A.W (1 Jan 1984).Selected Papers on Soil Mechanics, Thomas Telford Ltd, ISBN: 978-0727702050

41. Tatsuoka, F (1988). Some Recent Developments in Triaxial Testing Systems for Cohesionless Soils. Advanced Triaxial Testing of Soil and Rock, ASTM STP 977, R.T. Donaghe, R.C. Chaney, and M.L. Silver, Eds., American Society for Testing & Materials, Philadelphia, 1988, ISBN: 9780803109834, pages 7-67.

42. Tatsuoka, F, Shibuya, S, and Kuwano, R (1 Jan 2001). Advanced Laboratory Stress-strain Testing of Geomaterials, Taylor & Francis; 1st Edition, ISBN: 978-9026518430, pages 20-47.

43. Terzaghi, K (15 Jan 1943).Theoretical Soil Mechanics, Wiley, ISBN-13: 978-0471853053

44. Vesić, A.S. (1973). Analysis of Ultimate Loads of Shallow Foundations. Journal of the Soil Mechanics and Foundations Division, ASCE. Vol. 99, No. SM1, pages 45-73.

45. Vickers, B (1983). Laboratory Work in Soil Mechanics, Granada, Second Edition, USA.

46. Wood, D.M (26 April 1991)Soil Behaviour and Critical State Soil Mechanics, Cambridge University Press, ISBN: 97805213378

47. BS 8004 (1986). British Standard code of practice gives recommendations for the design and construction of foundations for the normal range of buildings and engineering structures, page 27.

48. BS 1377 (29/06/1990).This British Standard has nine parts. Part 1 is general information relevant to the other Parts. Parts 2 to 8 explain methods of soil tests for civil engineering purposes where samples are taken for testing in a laboratory. Parts 1, 2, 4 (Clause 3.3, 3.4), 7 (Clause 8, 9), 8.

49. Soil Mechanics Design Manual 7.01. (1986). Naval Facilities Engineering Command, Alexandria, Virginia, pages 348.