Lateral pressure behavior of soft clay in construction of deep excavations

Through investigation of field observations of a diaphragm wall supported excavation, the significance of wall panel construction effects on change of in-situ stress state of adjacent so

LATERAL PRESSURE BEHAVIOR OF SOFT CLAY

IN CONSTRUCTION OF DEEP EXCAVATIONS

by

Nguyen Hoang Quan

Dissertation submitted to the Faculty of the Graduate School of Engineering, Osaka Sangyo University

in partial fulfillment of the requirements for the degree of

Doctor of Philosophy

in Environmental Development Engineering

January, 2005

ii

LATERAL PRESSURE BEHAVIOR OF SOFT CLAY

IN CONSTRUCTION OF DEEP EXCAVATIONS

Nguyen Hoang Quan (ABSTRACT)

This dissertation presents investigations on aspects in lateral pressure behaviour of soft clay

during the construction of deep excavations In these investigations, physical experiments, field

tests and field observations were investigated in close association with inclusive numerical

analyses The purpose of this research is to enhance engineering knowledge of particular

behaviour of soft clay in the progressing state of the construction process of deep excavations

Aspects of active pressure behavior, passive pressure behavior and anchorage behaviour of soft

clay are investigated The active pressure behaviour was studied upon investigations on

deformation and stability of four cases of experimental slurry trenches excavated in soft clay

ground Variations of earth pressures and pore water pressures are elucidated as key factors of

the deformation and stability mechanism of trench wall Effect of initial soil-water conditions and

time-dependent performance of slurry trenches are presented and discussed

Through investigation of field observations of a diaphragm wall supported excavation, the

significance of wall panel construction effects on change of in-situ stress state of adjacent soft

clay are verified Analysis procedures to model the wall installation effects and to represent these

effects on analyses of diaphragm wall excavations are introduced and discussed

Study on passive resistance of overconsolidated soft clay using physical experiments was

enhanced by an extensive numerical experiment aimed at providing comprehensive information

on passive pressure behaviour of soft clay in vertical stress-relief condition of deep excavations

Progressing rate-dependent passive behaviour of soil in different loading conditions is elucidated

The applicability of current static solution of passive pressure (Rankine’s theory) to the evaluation

passive resistance of soft clay in deep excavations is discussed

Anchorage behaviour of soil is introduced through investigations on field tests for performance

of a newly developed type of under-reamed anchors (splits anchors) Results of field tests on two

anchors installed in Osaka Pleistocene clay are presented and examined Particular aspects in

performance of splits anchors are discussed Associated numerical analyses were performed to

clarify the field observations and to identify factors that influence the anchors’ performance

Keywords: Active Pressure, Deep Excavation, Numerical Analysis, Passive Pressure, Soft Clay;

Under-reamed Anchors, Wall Installation Effect

ACKNOWLEDGEMENTS

I would like to express my sincere appreciation to Professor Tomio Tamano for his guidance, insight and encouragement throughout the course of this research I have learnt a lot from his thorough and insightful review of the studies I am also grateful to my committee members, Professors Takeshi Iida and Professor Yasunori Nakamura for their patient review of this manuscript

I would like to thank Dr Manasobu Kanaoka who took care, supported and helped me to survive and enjoy living and studying in Japan Dr Kanaoka provided me any facilities needed to perform this research I am also grateful for the encouragement and suggestion of Mr Ikuo Sano

Mr Satoshi Fukui, Osaka City Office, made available all field measurements of slurry trench experiments and diaphragm wall excavation Dr Hiroshi Mastuzawa, OYO Corporation, and Mr Susumu Mizutani, Pacific Consultants Co Ltd., provided experimental data and their suggestions for the study on passive resistance Mr Yukio Fuseya and Mr Wataru Tonosaki, Nittoc Construction Co., provided results of field tests of splits anchors and information concerning the technology, design and construction of this newly developed under-reamed anchor I am grateful

to these persons for the support of my studies

I received financial support from the Japanese Government in the form of SHO Scholarship during the time I was conducting this research I gratefully acknowledge this support

MONBUKAGAKU-I owe a great deal to my teachers and colleagues at Bridge and Highway Department, Hochiminh City Polytechnic University, who encouraged me to come to Japan and pursue a Ph.D I particularly thank Dr Le Van Nam, previous Dean of Civil Engineering Faculty, who recommended and helped me to go study abroad

I would like to express my sincere thanks to my mother, my sister and my family members, for their love and encouragement I thank them for enlightening the value of hard work, honesty and perseverance I am also grateful to my father-in-law and mother-in-law for their care and support

of my studies

Finally, I am so grateful and fortunate to get married to the most understanding, loving, wise and entertaining person that I know I express my thanks and my love to my wife Le Thi Xuan Trang for her support, friendship, love and encouragements that helped me going through this endeavour

iv

TABLE OF CONTENTS

Page

Abstract ii

Acknowledgements iii

Table of Contents iv

List of Tables vii

List of Figures ix

1 INTRODUCTION 1

1.1 Lateral Pressure Behavior of Soft Clay in Construction of Deep Excavations 1

1.2 Objectives and Scopes of Research 2

1.3 Organization 3

2 RESEARCH BACKGOURND AND LITERATURE REVIEW 4

2.1 Introduction 4

2.2 Previous Reviews 4

2.3 Deformation and Stability of Slurry Trench in Soft Clay 7

2.4 Effects of Wall Panel Construction on Lateral Stresses of Adjacent Soft Clay 10

2.5 Passive Resistance of Soft Clay in Vertical Stress-Relief Condition of Deep Excavations 16

2.6 Performance of Under-reamed Anchors Constructed in Structured Clay 20

2.7 Finite Element Modeling 23

2.8 Conclusions 33

References 34

3 DEFORMATION AND STABILITY OF SLURRY TRENCH EXCAVATED IN SOFT CLAY 38

3.1 Introduction 38

3.2 Four Case Studies of Slurry Trenches in Soft Clay 39

3.3 Deformation and Stability of Slurry Trench 42

3.4 Performance of Slurry Trench in Reclaimed Ground 65

3.5 Conclusions 77

References 78

4 EFFECTS OF WALL PANEL CONSTRUCTION ON LATERAL

STRESSES OF ADJACENT SOFT CLAY 79

4.1 Introduction 79

4.2 A Case Study of Diaphragm Wall Deep Excavation 79

4.3 Field Observations in the Construction of Wall Panels 86

4.4 Numerical Modeling of the Wall Installation Process 93

4.5 An Approach to Represent the Installation Effects in Diaphragm Wall Analysis 102

4.6 Conclusions 107

References 108

5 PASSIVE RESISTANCE OF SOFT CLAY IN VERTICAL STRESS-RELIEF CONDITION OF DEEP EXCAVATIONS 109

5.1 Introduction 109

5.2 Physical Experiment 109

5.3 Numerical Experiment 112

5.4 Analysis Results and Discussion 117

5.5 Effects of Simultaneously Progressing Processes 122

5.6 Conclusions 137

Notations 138

References 139

6 FIELD TESTS AND NUMERICAL EXPERIMENTS OF UNDER-REAMED ANCHORS EMBEDDED IN OSAKA MARINE CLAY 140

6.1 Introduction 140

6.2 Background 140

6.3 Field Load Tests 146

6.4 Numerical Experiments 159

6.5 Analysis Results and Discussion 163

6.6 Conclusions 173

Notations 174

References 175

vi

7 SUMMARY AND CONCLUSIONS 176

7.1 Introduction 176

7.2 Research Summary and Conclusions 176

7.3 Overall Conclusions on Lateral Pressure Behavior of Soft Clay in Construction of Deep Excavations 179

7.4 Recommendation for Further Research 179

References 180

Related Publications 181

Biography 182

LIST OF TABLES

Page

Table 3.1 Descriptions of four case studies of slurry trenches excavated in soft clay 40

Table 3.2 Physical and mechanical soil properties 43

Table 3.3 Experimental procedure of the slurry level-lowering test 45

Table 3.4 Input parameters of two sand layers (HS model) 50

Table 3.5 Primitive parameters derived from results of laboratory tests (SS model) 51

Table 3.6 Primitive and best-fitted parameters of the alluvial clay 52

Table 3.7 Primitive parameters and best-fitted parameters of the diluvial clay 54

Table 3.8 Analysis cases 55

Table 3.8 Input parameters of the two sand layers (HS model) 68

Table 3.9 Input parameters of the two clay layers (SS model) 68

Table 4.1 Numerical analysis phases 95

Table 4.2 Measured and computed variation of horizontal effective pressure (stress) 99

Table 4.3 Modeling procedure for the pressure-equivalent approach 103

Table 4.4 Post-installation horizontal effective pressure (stress) 105

Table 5.1 Soil parameters adopted in the primary analysis 114

Table 5.2 Compressive strength q of the overconsolidated soft clay 115 u Table 5.3 Variables adopted in the parametric study 123

Table 5.4 Mobilized passive resistance of clay during the progressing of excavation ( Rankine P P P P ) 135

Table 6.1 Reference values of strength reduction factor α of cohesive soils 144

Table 6.2 Properties of the alluvial clay (Amc) and diluvial clay (Tc12) 146

Table 6.3 Estimated ultimate pullout resistance of anchors 151

Table 6.4 Input parameters backfill and loose sand (Mohr-Coulomb soil model) 160 Table 6.5 Input parameters of the alluvial clay and diluvial clay (Hardening-Soil model) 160

viii

Table 6.6 Descriptions of performed analysis cases 162 Table 6.7 Load loss percentages of anchors embedded in clays 165 Table 6.8 Measured and computed contributions of resistance components 171

LIST OF FIGURES

Page

Figure 2.1 Construction sequence of a typical diaphragm wall 8

Figure 2.2 Lateral earth pressure development during wall concreting 13

Figure 2.3 Design envelopes of concrete pressure 13

Figure 2.4 Recommendations of ACI to evaluate maximum lateral pressures of fresh concrete for various temperatures 13

Figure 2.5 Variation of passive resistance on excavated-side of deep excavation estimated as function of wall displacement and decrease of vertical effective stress 19

Figure 2.6 Stress-strain behavior of the elastic perfectly plastic Mohr-Coulomb model 26

Figure 2.7 The Mohr-Coulomb yield surface in principal stress space for cohessionless soil 26

Figure 2.8 Yield surface of the Soft-Soil model in p’-q plane 28

Figure 2.9 Representation of total yield contour of the Soft-Soil model in principal stress space 28

Figure 2.10 Hyperbolic stress-strain relation in primary loading for a standard drained triaxial test of the Hardening-Soil model 30

Figure 2.11 Representation of total yield contour of the Hardening-Soil model in principal stress space for cohesionless soil 30

Figure 3.1 Reclamation site and locations of the three trial slurry trenches (Case B, C and D) 41

Figure 3.2 Plan and cross section of the three trial slurry trenches 41

Figure 3.3 Subsoil profile and soil properties at the experimental site 43

Figure 3.4 The experimental slurry trench and its instrumentations (Case A) 44

Figure 3.5 Observed movements of subsoil adjacent to the experimental trench 46

Figure 3.6 Observed subsoil movements (a) and responses of pore water pressure (b) against variations of slurry level (at G.L.-16.0m, 1.5m from trench wall) 46

x

Figure 3.8 Axisymmetric finite element model for analyses of the experimental slurry

trench 49

Figure 3.9 Calibration input parameters of the alluvial clay (Ac) 53

Figure 3.10 Calibration input parameters of the diluvial clay (Dc) 53

Figure 3.11 Computed trench wall displacement and movements of adjacent subsoil 56

Figure 3.12 Computed responses of pore water pressure against variations of slurry level (at G.L.-16.0m, 1.5m distant from trench wall) 56

Figure 3.13 Progressive yield of soft clay adjacent to trench wall 58

Figure 3.14 Correlation between variations of pore water pressures and wall displacements 59

Figure 3.15 Progressive displacements of trench wall during standing period (taken at G.L.-16.0m) 61

Figure 3.16 Variations of lateral pressure, pore water pressure and horizontal effective pressure during standing period (taken at slurry-soil interface, G.L.-16.0m) 61

Figure 3.17 Continuous softening of subsoil during standing period (soil elements at slurry-soil interface, G.L.-16.0m) 62

Figure 3.18 Progressive yield of soft clay during standing period (slurry level at G.L.-1.5m) 62

Figure 3.19 Progressive yield of soft clay during standing period (slurry level at G.L.-2.5m) 63

Figure 3.20 Progressive yield of soft clay during standing period (slurry level at G.L.-3.5m) 63

Figure 3.21 Progressive yield of soft clay during standing period (slurry level at G.L.-5.0m) 64

Figure 3.22 Consolidation effects on predictions of trench wall displacements (at G.L.-16.0m) 64

Figure 3.23 In-situ stress conditions at locations of the three trial trenches 66

Figure 3.24 Observed trench wall displacements in Case B: a failure instance 66

Figure 3.25 Observed trench wall displacements in Case C: a critical instance 67

Figure 3.26 Observed trench wall displacements in Case D: a stable instance 67

Figure 3.27 Axisymmetric finite element model for analyses of the trial trenches 69

Figure 3.28 Model settings and excess pore water pressures generated by the reclamations (Case B) 71

Figure 3.29 Model settings and excess pore water pressures generated by the reclamations (Case C) 71

Figure 3.30 Model settings and excess pore water pressures generated by the reclamations (Case D) 72

Figure 3.31 Computed trench wall displacements of the four analysis cases 75

Figure 3.32 Influence of the remaining excess pore water pressure intensity (due to reclamation) on displacement and stability of slurry trench walls 76

Figure 4.1 Cross section of the diaphragm wall supported basement excavation 80

Figure 4.2 Element wall panels of the diaphragm wall 82

Figure 4.3 Locations of measuring devices and measuring sections 82

Figure 4.4 Position of earth pressure cells and pore water pressure cells on wall panels of the diaphragm wall at the three measuring sections 84

Figure 4.5 Placement of earth pressure cell and water pressure cell on the reinforcement cage 85

Figure 4.6 Variations of lateral earth pressures at elevations during and after wall concreting 87

Figure 4.7 Readings of representative earth pressure cells embedded in soft clay layer 88

Figure 4.8 Readings of representative pore water pressure cells embedded in soft clay layer 88

Figure 4.9 Variations of lateral earth pressure during wall concreting process 90

Figure 4.10 Axisymmetric finite element model for wall installation analysis 94

Figure 4.11 Predictions of the primary analysis (entire wall construction process) 96

Figure 4.12 Predictions of the primary analysis for wall concreting and soil consolidation 97

Figure 4.13 Predictions of the reference analysis for wall concreting and soil consolidation 98

xii

Figure 4.14 Decrease of effective stresses due to wall installation 101

Figure 4.15 Influenced extent of wall-installation effect 101

Figure 4.16 Plain strain finite element model for diaphragm wall analysis 104

Figure 4.17 Lateral pressure profile and displacement profile at the wall-soil interface (at the final setting of fresh concrete) 104

Figure 4.18 Post-installation horizontal effective stress represented in diaphragm wall analysis 106

Figure 5.1 Schematic diagram of the retaining wall apparatus 111

Figure 5.2 Experimental procedure of the physical test 111

Figure 5.3 Plane strain finite element model 113

Figure 5.4 Measured versus computed surface settlement and heave 116

Figure 5.5 Calibrating input parameters for shear strength of the overconsolidated soft clay 116

Figure 5.6 Measured versus computed passive resistances (during wall-displacing test) 118

Figure 5.7 Measured versus computed passive resistances (in time domain) 118

Figure 5.8 Measured versus computed (total) horizontal earth pressures (during wall-displacing test) 120

Figure 5.9 Computed distribution of excess pore water pressure (at d =d MAX ) 120

Figure 5.10 Measured versus computed heaves of backfill surface 121

Figure 5.11 Contours of induced maximum shear strain (at d =d MAX ) 121

Figure 5.12 Modeling procedures implemented in analysis groups 123

Figure 5.13 Computed responses of soil during wall-displacing process (Group A) 125

Figure 5.14 Computed maximum and steady passive resistances (Group A) 126

Figure 5.15 Contribution fractions of P P′ and P to W MAX P P (Group A) 126

Figure 5.16 Computed responses of soil during wall-displacing process (Group Ｂ) 128

Figure 5.17 Computed maximum and steady passive resistances (Group B) 129

Figure 5.18 Computed responses of soil during wall-displacing process (Group C,

mm

d MAX =44 ) 131

Figure 5.19 Computed maximum and steady passive resistances (Group C, mm d MAX =44 ) 132

Figure 5.20 Computed responses of soil during wall-displacing process (Group C, mm d MAX =4.4 ) 133

Figure 5.21 Computed maximum and steady passive resistances (Group C, mm d MAX =4.4 ) 134

Figure 5.22 Disparity of computed passive resistances of soft clay 136

Figure 6.1 Collapsible splits bit for under-reaming enlarged body of splits anchors 142

Figure 6.2 Construction sequences of splits anchors 142

Photo 6.1 Excavated body of a splits anchor constructed in medium stiff clay (D−800, L=1.5m) 143

Figure 6.3 Estimation ultimate pullout resistance of vertical under-reamed anchor embedded in clays 145

Figure 6.4 Subsoil profile at the experimental site 147

Figure 6.5 Grain size distribution curves of the alluvial clay and diluvial clay 147

Figure 6.6 Section of the sheet-pile wall supported by tieback splits anchors 148

Figure 6.7 Two experimental vertical splits anchors 150

Figure 6.8 Load histories of the field tests on anchor No.1 150

Figure 6.9 Load-movement curves of the performance test (anchor No.1) 152

Figure 6.10 Variations of anchor load in the relaxation test (anchor No.1) 154

Figure 6.11 Variations of anchor load at specific temperatures (anchor No.1) 154

Figure 6.12 Load-movement curves of the creep test (anchor No.1) 155

Figure 6.13 Creep movement and creep rate (anchor No.1) 155

Figure 6.14 Load-movement curve of the one-cyclic pullout load test on anchor No.2 158

Figure 6.15 Axisymmetric finite element analysis model 161

xiv

Figure 6.16 Observed versus predicted deformation of the structured clay (Tc12) in

Constant Rate of Strain (CRS) consolidation test 161

Figure 6.17 Computed load relaxation performance of splits anchor (anchor No 1) 164

Figure 6.18 Load loss induced by dissipation of pore water pressure 164

Figure 6.19 Load relaxations at different soil permeabilities 166

Figure 6.20 Computed load-movement curves of creep test 167

Figure 6.21 Computed creep movement and creep rate 167

Figure 6.22 Development of total anchor load and the resistance components (anchor No.1; without debonding effect) 169

Figure 6.23 Development of total anchor load and the resistance components (anchor No.2; without debonding effect) 169

Figure 6.24 Development of soil-grout failure along the anchor shaft 170

Figure 6.25 Long-term variations of resistance components in creep test (Q 830= kN) 170

Figure 6.26 Pre-loading effect in creep test 172

Figure 6.27 Pre-loading effect in load-relaxation test 172

The performance of deep excavations, i.e basal stability, wall deformation and ground movement, is often more uncertain during the construction than upon its completion For stiff supporting systems such as secant bored pile walls or braced diaphragm walls, wall deformations and ground movements are usually reported to occur mostly during the excavation process and be well restrained when the walls are properly braced after the completion of excavation works For that reason, it is preferred to access the performance of deep excavation during its construction process Consequentially, it become necessary to investigate the behavior of surrounding subsoils, which is obviously an essential factor governing the performance, with great concerns on the progressing of the excavation

Lateral pressure behavior is the most important aspect in behavior of soil that influences the performance of deep excavations Active earth pressure acting on the retained-side (back-side)

of the wall is the principal external load of the wall and its supporting system On the contrary, the passive pressure of soil beneath the excavation formation level acting on the excavated-side of the wall is a part of the supporting system of the wall Anchorage behavior is another aspect in soil behavior that is involved in deep excavations that utilize tieback anchors to support the retaining wall It is needed to build a strong and consistent understanding of these aspects of soil behavior to predict and elucidate the performance of deep excavations

Soft clay is known as a problematic geotechnical material for its very low shear strength and high potential of deformation In considering the behavior of subsoils during the construction

of deep excavations, time dependency behavior due to consolidation effect, is another characteristic of the soft clay should be involved It is because in current engineering practice,

Chapter 1 - Introduction

the soft clay (cohesive soils, in general) usually considered impermeable, and only its behavior

in short-term undrained condition is taken in to account It becomes even more doubtful when available calculation methods, which were originally developed for drained non-cohesion soils, are implicitly adopted During the construction of deep excavations, behavior

of soft clay, due to but despite its low permeability, is progressive and highly time-dependent Neither undrained approaches nor drained approaches can properly capture the behavior of the soft clay in this condition

The nature of studies presented in this dissertation is to investigate aspects in lateral pressure behavior of soft clay involved within the progressing state in construction of deep excavations

1.2 OBJECTIVES AND SCOPE OF RESEARCH

The objectives of this research are to provide information in the field of lateral pressure behavior of soft clay in construction of deep excavations that includes: active pressure behavior of soft clay adjacent to slurry trench, effects of wall panel construction on lateral stresses of soft clay, passive pressure behavior of soft clay in stress-relief condition of deep excavation, and anchorage behavior of structured clay Specific objectives and the research methods to approach are as followings:

(a) Investigate active pressure behavior of soft clay adjacent to slurry trench

This objective was accomplished by investigating field observations of four experiment cases

of slurry trench excavated in soft clay Numerical analyses are used in association with the experiments that will provide supplemental information on behavior of soil Parametric studies are conducted to identify and verify factors that might influence the deformation and stability of slurry trench excavated in soft clay

(b) Investigate effects of wall panel construction on lateral stresses of soft clay

This objective was accomplished by investigating observed variations of earth pressure and pore water pressure during the construction of wall panels in an experimental case study of diaphragm wall excavation The observations are elucidated and compared with current suggestions concerning the effects and modeling of wall installation in analysis of diaphragm wall excavation Analysis procedure and modeling approach are developed to simulate the wall construction process and to represent post-installation stress state of subsoil in analysis of diaphragm wall excavations

Chapter 1 - Introduction

3

(c) Evaluate passive resistance of soft clay in stress-relief condition of deep excavations

This objective was accomplished by performing a comprehensive numerical experiment based

on physical experiments and on apprehensions of real behavior of soils on excavated side of deep excavations The numerical experiment is firstly calibrated and validated by measurements of the physical experiment Parametric study is then applied to enhance modeling capacity of the numerical experiment that aimed at exploring soil behavior in progressing conditions similar to those in actual excavations

(d) Examine performance of under-reamed anchors constructed in structured clay

This objective was accomplished by investigating experimental observations in field tests of the newly developed type of under-reamed anchors, named as splits anchor, that were constructed in structured clay Performance of the tested anchors is compared with that of straight anchors and with current information about anchorage behavior of under-reamed anchors constructed in (unstructured) clay Numerical analyses with parametric study are utilized to verify and improve the understanding

1.3 ORGANIZATION

The remaining chapters of this dissertation describe the studies performed on lateral pressure behavior of soft clay in construction of deep excavations Chapter 2 describes the research background and the literature review of studies on deep excavations with respect to lateral pressure behavior of soft clay Chapter 3 describes the study on active pressure behavior of soft clay through investigation on field observations and numerical predictions of deformation and stability of slurry trenches Chapter 4 presents field observations and numerical modeling

of wall panel construction of a case study on diaphragm wall excavation in soft clay Changes

of in-situ lateral pressures of the soft clay during the wall construction process are investigated Chapter 5 describes the investigation on results of physical and numerical experiments for passive resistance of overconsolidated soft clay in vertical stress-relief condition of deep excavations Chapter 6 gives results of field tests on a newly developed type of under-reamed anchors constructed in Osaka marine clay Predictions of associated numerical analyses are examined and discussed Chapter 7 summarizes results and conclusions of the studies, manifests overall conclusions on the nature of this research, and presents recommendations for further studies

CHAPTER 2

Research Background and Literature Review

2.1 Introduction

2.2 Previous Reviews

2.3 Deformation and Stability of Slurry Trench in Soft Clay

2.4 Effects of Wall Panel Construction on Lateral Stresses of Adjacent Soft Clay 2.5 Passive Resistance of Soft Clay in Vertical Stress-Relief Condition of Deep Excavations

2.6 Performance of Under-reamed Anchors Constructed in Structured Clay 2.7 Finite Element Modeling

2.8 Conclusions

References

Chapter 2 – Research Background & Literature Review

2.2 PREVIOUS REVIEWS

State-of-the-art papers and reviews of Peck (1969), Lambe (1970), Goldberg et al (1976), O’Rourke (1981), and Clough and O’Rourke (1990) are extensive summaries of studies published before 1990 Bentler (1998) added a comprehensive review of analytical and field performance studies on deep excavations that were published by 1998 The following text summarizes learned lessons described in these reviews that related to the nature and objectives of this research

Deep excavations in soft soils can cause large soil movements and surface settlements

Soil type and condition were manifested to play determinant roles in performance of deep excavations In Peck’s summary of field observed settlements adjacent to excavations, three zones of settlement profiles were separated based on soil conditions and workmanship Excavations in soft clay were notified to induce the most intensive surface settlements Goldberg

et al (1976) noted that whereas wall displacements for excavations in sand, gravel or very stiff clay are usually less than 0.4% of the excavation depth, excavations in soft soils induce wall displacements at about 1% of the excavation depth Clough and O’Rourke’s (1990) recommended settlement profiles for estimating the distribution of settlements adjacent to excavations that were dependent on soil types In soft to medium clays, deep excavations induced more intensive surface settlements within the wall vicinity (d H ≤ 1 0)

Chapter 2 – Research Background & Literature Review

Construction sequencing and progression of excavation activities can affect the performance of deep excavations

Because soils are non-linear and their behavior depends on the loading path, different construction sequencings induce very different loading conditions on the subsoil, consequentially, lead to different in the performance of deep excavations The previews also suggested that supports must be installed promptly after each excavation stage in order to minimize wall displacements During delay period, progressive increases of wall displacements might occur due

to time-dependent deformation of subsoil (consolidation and creep) and/or over-excavation (i.e delay of support installation while excavation continues) Both in engineering practice and in analysis of deep excavations, performance of the excavation must be considered with reference to the progression of excavation activities

Consolidation is an important factor on the performance of deep excavation in clay

Continuing settlements of buildings next to a braced excavation in soft clay were report in field performance study of Brassinga and Van Tol (1991) Finno and Harahap (1991) noted significant changes of pore water pressure acting on to wall associated with displacements of the sheet pile wall during the excavation process Analytical studies of Osaimi and Cough (1979), Yong et al (1989), Finno et al (1991) and Ou and Lai (1994) all emphasized the importance of consolidation

Wall construction can cause significant soil movements and stress changes

It was reported that stiff diaphragm walls help limit movements of soils adjacent to the excavations, especially in case of soft soils (Goldberg et al., 1976; Clough & O’Rourke, 1990) Nevertheless, the effects of wall construction must be concerned Clough and O’Rourke (1990) investigated settlements measured in six case studies of diaphragm wall construction, which revealed that the settlements could be as large as 0.12% of wall depth and could occur out to a distance of twice the wall depth from the edge of the wall Finno et al (1991), Schweiger and Freiseder (1994), and Ng et al (1995) noted that case histories suggested significant movements and stress changes could occur during wall construction, prior to the bulk excavation Settlements during wall construction were observed approximately equal to the settlements that occurred during excavation Two factors were identified to cause subsoil movements during wall construction: one is the difference between the initial horizontal soil stresses, the slurry pressure and the wet concrete pressure; the other factor is the soil stiffness In the other aspect, the post-installation stress state was noted could be very different from initial stress state before wall construction The stress changes must be reasonably evaluated for accurate simulation of excavation performance

Chapter 2 – Research Background & Literature Review

6

Soil characterization and site investigation are important in design of deep excavations

It was addressed that inaccurate assumptions of soil profile and soil properties in analysis and design could lead to failure of deep excavations (Swanson & Larson, 1990) White et al (1993) compared predictions of finite element analyses with field performance of a deep excavation in Boston and concluded that adequate soil characterization was a prerequisite to accurate finite element calculations

Large in-situ lateral soil stresses were considered to adversely affect deep excavations Peck (1969) notified that basal failure could occur in excavations in soils with very large initial lateral stresses Goldberg et al (1976) noted experiences of very large displacements of tieback walls in over-consolidated clays, and Clough and O’Rourke (1990) indicated that soil movements in the anchorage zone of tieback walls could occur in over-consolidated clays with high lateral stresses

Finite element analysis with advanced soil models is effective in study on deep excavations

Finite element (FE) analyses were utilized in most of analytical studies Parametric study is a useful and effective feature of FE analysis that helps to investigate many factors that might control the performance of deep excavations Besides, since the excavation problem is complicated, FE analyses of deep excavations must be performed with advance constitutive soil models that can capture many aspects of real soil behavior In case of cohesive subsoil (clay), soil-water coupling feature is needed in FE analyses to simulate undrained behavior of the clay under effects of consolidation process

Even though three-dimensional effects were recognized significant in reality and in analyses of deep excavations, numerical analyses were mostly performed in plane strain condition Only few analytical studies were performed using real three-dimensional model due to it complexity There are still rare three-dimensional finite element analysis programs that comprise prerequisite features like advance soil models and soil-water coupling to perform accurate analyses of deep excavations in clays

Chapter 2 – Research Background & Literature Review

2.3 DEFORMATION AND STABILITY OF SLURRY TRENCH IN SOFT CLAY 2.3.1 Background

Figure 2.1 describes the construction sequence of a typical diaphragm wall, also named as slurry wall The excavation of wall panel is stabilized by filling it with bentonite slurry to support the trench walls Because slurry pressure at a given depth is usually smaller than the corresponding initial lateral stress of adjacent subsoil, the trench walls deform inwards that induces lateral movement of the subsoil and settlement of ground surface

Increasing the slurry pressure, by either using heavier bentonite slurry or maintaining a higher slurry level, is an effective way to minimize trench wall displacements and subsoil movements and to improve the stability of the slurry trench However, there are limitations on the density of bentonite slurry (≤1.3g / ml) and the slurry level (usually lower than ground surface) Therefore, trench panels excavated in soft soil might produce large trench wall displacements and subsoil movements that damage adjacent facilities Poor subsoil conditions and workmanship even cause intensive deformation and collapse of trench walls

Field observations and analytical analyses emphasized a large portion, as much as 50%, of subsoil movements and ground settlements caused by deep excavations, might occur during the panel trenching and wall concreting processes of the diaphragm wall construction Since diaphragm wall excavations have been increasingly constructed in congested urban areas, it is become necessary to understand and predict appropriately the deformation and stability of slurry trenches, especially those excavated in soft clay

From another point of view, behavior of subsoil adjacent to slurry trenches is a primary and representative instance of active behavior of subsoil in deep excavations There are no structural members, as wall, struts, anchors, etc., and external loads, as dewatering, bracing, pre-stressing of anchors, etc., that might influence behaviors of the soils Varying the slurry pressure, which is the only external load within this problem, in a slurry level variation test provides an effective access

to investigate many aspects in active behavior of soils

2.3.2 Previous Studies

Stability of slurry trench was focused in quite early studies (Dibiagio, 1972) It was noted that even in soft clay, the excavation of slurry trench meet no difficulty (Dibiagio, 1972; Tamano et al., 1996) However, failure cases of slurry trench where the trench walls experienced intensive displacements and finally collapsed were also reported (Tamano et al., 1995)

Figure 2.1 Construction sequence of a typical diaphragm wall

8

Chapter 2 - Research Background & Literature Review

Chapter 2 – Research Background & Literature Review

Subsoil movements induced by the trenching of wall panels attracted many recent studies those concerned the effects of wall construction in diaphragm wall excavations Field performance studies of Clough and O’Rourke (1990), Finno et al (1991), Schweiger and Freiseder (1994), Ng

et al (1999), Poh et al (2001) all showed that significant subsoil movements could occur during the trenching process of wall panels It was notified that subsoil movements induced by the panel trenching could be about 50% of the total subsoil movements caused by the excavation

Numbers of analytical studies using finite element analyses have been performed in attempt to predict the trench wall displacements, movements of subsoil and settlement of ground surfaces (Ng & Yan, 1998, 1999; Gourvenec & Powrie, 1999; Ng, 1999) Due to the lack of fully instrumented experiments, these analytical studies were not properly verified by experimental observations and therefore failed to elucidate reasonably the nature of slurry trench’s performance There have been very scarce studies in literature that utilize the association between field observations and numerical analyses to investigate the problem inclusively (De Moor, 1994;

Ng et al., 1999) Besides, wall displacements and surface settlements have attracted much attention of researchers than the stress-strain behavior or the variation of earth pressures and pore water pressures, which can provide principal information about deformation and stability mechanisms of slurry trench

Studies of Osaimi et al (1979), Ng (1999) using soil-water coupled analyses with effective stress soil models exposed essential aspects of soil behavior that could not be captured with commonly used undrained analyses Observed variations of pore water pressure and horizontal effective stress in centrifuge model test of Powrie (Powrie et al., 1996) are among few accessible experimental data It was apprehended that the development and variation of pore water pressure would have a substantial role in the stability of the slurry trench walls

Tamano et al (1995, 1996) provided intensive investigations on field observations from experiments of slurry trenches excavated Undrained responses of pore water pressure of soft clay adjacent to slurry trenches against variations of the slurry pressure were elucidated as a determinant factor of the deformation of slurry trench walls excavated in saturated soft clay A linear correlation of decreases of pore water pressures with increases of trench wall displacements was established using the experimental measurements This correlation was supposed to depend

on undrained deformation characteristics of the saturated soft clay and hydrostatic pore water pressures found in the initial ground

Chapter 2 – Research Background & Literature Review

of soil stresses induced by this construction process of diaphragm walls are commonly referred as wall-installation effects

During the wall construction, lateral stresses of soils within the panel trench are subsequently replaced by hydrostatic lateral pressure of the bentonite slurry (in panel trenching) and lateral pressure of the fresh concrete (in wall concreting) These changes of boundary condition at wall-soil interfaces of the panel trench cause movements of adjacent subsoil and changes of soil stresses In general subsoil condition, the panel trenching process will causes inwards movements

of adjacent subsoil, settlements of ground surface and decreases of soil stresses; and on the contrary, the wall concreting process will cause reverse effects with outwards and upwards movements of subsoil and increases of soil stresses Because the setting of wall concrete is associated with reductions of its volume and lateral pressure, the process likely induces further inwards movements of subsoil and decreases of stresses Results of the consolidation of soils are the recoveries of soil stresses (and pore water pressures) towards their initial condition

Subsoil movement and surface settlement are explicit and visible aspects of the wall-installation effects These quantities were intensively examined in most studies on wall-installation effects since they could directly cause damages to nearby buildings and facilities On the contrary, changes of soil stresses due to the wall construction is not that easily ascertainable Since behavior of soils is non-linear and highly stress-strain dependent, it is obvious that these changes will influence performance of the excavation, i.e forces and moments in supporting structures, and stresses, movements and deformation of the surrounding subsoil

2.4.2 Lesson Learned From Previous Studies

Issues addressed in published studies that concerned the wall-installation effects are summarized

in the following text

Chapter 2 – Research Background & Literature Review

Subsoil movements and ground surface settlements due to wall construction

Case histories that reported significant subsoil movements and ground surface settlements during the wall construction process have been increasing recently, which urged a great need of experimental and analytical studies on the wall-installation effects Symons and Carder (1993), Ng

et al (1998, 1999), Poh and Wong (1998), Poh et al (2001) provided field measurements during the construction of wall panels of diaphragm walls; most of these measurements were of subsoil movements and ground surface settlements It was notified that maximum inwards horizontal subsoil movements and surface settlements occurred during the panel trenching process A portion of these movements and settlements recovered during the wall concreting, then in couple

of hours after the wall concreting completion, probably by the final setting of wall concrete, some adversely inwards horizontal subsoil movements and surface settlements were observed

Based on the above apprehensions, it is inferable that when the maximum subsoil movements and ground surface settlements are of concerned as wall-installation effects, measurements and/or predictions of these quantities during the process of panel trenching are substantial and quite adequate Further investigations on the successive processes of wall concreting, concrete setting and soil consolidation are needed for appropriate estimation of changes of soil stresses

Changes of soil stresses and pore water pressures due to wall construction

Symons and Carder (1993), Lings et al (1994), Tamano et al (1996) and Ng et al (1998) provided valuable field measurements of changes of subsoil lateral pressures and pore water pressures during the wall construction process Field measurements reported by Tamano et al (1996) were made in ground with thick soft clay deposit; the others were made in over-consolidated stiff clay

It was notified in these field-performance studies that:

- At the soil-wall interface, lateral earth pressures substantially decreased to the hydrostatic bentonite pressure during the panel trenching and then increased to the lateral pressure of fresh concrete (hereafter, concrete pressure) during the wall concreting Pore water pressures underwent analogous variations during these processes

- Dissipation of excess pore water pressures at the soil-wall interfaces was rapid and the pore water pressures soon recovered to initial values within few days after the wall concreting

- After the wall concreting, the lateral pressures decreased significantly Some increases occurred during the standing period prior to the bulk excavation, but the pressures hardly recovered to their initial values found prior to the wall construction Consequently, post-installation earth pressure coefficients were smaller than those found in initial ground

Chapter 2 – Research Background & Literature Review

Lateral pressure of fresh concrete (concrete pressure)

In the field of formwork engineering, studies on lateral pressure of fresh concrete were made extensively The freshly placed concrete, particularly during vibration, behaves much like a liquid, exerting outward equivalent hydrostatic pressure on the confining boundary However, experiments show that most of the time the concrete pressure is less than the equivalent hydrostatic pressure During the course of fresh concrete placement, the concrete pressure at a given level rises to a maximum value during the concrete filling, and then remains constant or slightly reduces during the successive concrete filling As the fresh concrete starts solidifying, the lateral pressure exerting on formwork gradually reduces and becomes zero when the concrete completely set (Amziane et al., 2002)

In estimation of lateral pressure of fresh concrete placed in air, a bi-linear envelope of the maximum concrete pressures is usually assumed, as shown in Figure 2.3a At less than a certain depth below the surface of the concrete, denoted as critical depthh , the maximum concrete crit

pressures are hydrostatic, equaling the full fluid pressure of fresh concrete: P c =γ c×h At depths greater than this critical depth, the maximum concrete pressure has a limiting value

crit

c h

Pmax =γ × that is approximately constant with increasing depth The limiting value of concrete pressure depends on various factors like concrete mixture, rate of concrete placing, temperatures of fresh concrete, etc Figure 2.4 shows recommendations provided by American Concrete Institute (ACI) to evaluate the limiting concrete pressures for various concreting temperature

Figure 2.4 Recommendations of ACI to evaluate maximum lateral pressures of fresh concrete for

0 400 800 1200 1600 2000 2400 2800 3200

Figure 2.2 Lateral earth pressure development during wall concreting (after Lings et al., 1994)

Figure 2.3 Design envelopes of concrete pressure

Chapter 2 - Research Background & Literature Review

Concrete pressure Concrete pressure

Typical envelope of concrete pressure on formwork Slurry pressureDesign envelope

of concrete pressure Design envelopeof concrete pressure Envelope of pressure

Chapter 2 – Research Background & Literature Review

σ = + − , for z>h crit

, where σ is the lateral pressure and h zis the depth below the finished concrete level (Fig 2.3b) Lings et al (1994) calibrated the formulas with field measurements of maximum concrete pressures in three case histories of diaphragm-wall excavations in overconsolidated clay, and suggested an approximate critical height of one third of the trench panel depth (h crit=H 3) The authors also noted that this approximation was primitive and more detailed field measurements were needed to improve the estimation of the critical depth Even though no further attempts were made to validate the suggestion, it has been widely utilized in recent analytical studies (Ng & Yan, 1998, 1999; Gourvenec & Powrie, 1999; Gourvenec et al., 2002; Ng & Lei, 2003)

Numerical modeling of wall-installation effects

Attempts in numerical modeling the wall construction process gained essential achievements (Gun et al., 1993; De Moor, 1994; Ng & Yan, 1998, 1999; Gourvenec & Powrie, 1999; Gourvenec et al., 2002; Ng & Lei, 2003) Lessons learned from these studies are as the following:

- The modeling of the construction process of wall panels should be performed with full three-dimensional analyses Three-dimensional effects are substantial when the panel aspect ratio (depth/length)H L≥3 Horizontal arching effect leads to non-uniform horizontal stress distribution behind the wall panel, the stress being smallest at the centre but increasing in magnitude towards the edges

- Modelling the wall construction process as a plane strain event will lead to significant overestimations of both the magnitude and extent of the lateral stress reductions and the induced ground movements Axisymmetric analyses underestimate the stress changes but in general produce the soil response closer to the real response than predictions of

Chapter 2 – Research Background & Literature Review

plane strain analyses In principle, the stress changes calculated in axisymmetric analyses could be used in subsequent plane strain analysis of the diaphragm-wall excavation

- An approach to model the wall-installation effects of three-dimensional wall panels is to impose the maximum horizontal displacement profile calculated at the soil-wall interface

in the three-dimensional analysis of a single panel in a plane strain analysis of the diaphragm-wall excavation starting with the wall already in place However, this modelling procedure seems to underestimate the changes of soil stresses

- All of the significant changes of soil stresses and movements of subsoil in front of a given panel occur during the construction of that panel and the effects of the construction of subsequent panels on these quantities are insubstantial

2.4.3 Needs of Information on Changes of Soil Stresses Due to Wall Installation

In current assessment to changes of soil stresses as wall-installation effects, there are unascertained issues that include the magnitude and variation of the fresh concrete pressures and

a numerical procedure to model the entire wall construction process

Magnitude and variation of concrete pressures:

The lateral pressure of fresh concrete is a determinant factor that particularly governs the changes

of soil stresses, but the current knowledge and calculation of magnitude and variation of the pressures remain very primitive While the assumption of a bi-linear envelope of the maximum concrete pressure (Lings et al., 1994) seems reasonable, the estimation of the critical depth h crit

needs more refinement Since no further validations were made for the “one-third” suggestion of the critical height, it would be more appropriate to follow the recommendations provided in the CIRIA Report 108 to estimate the critical height

Moreover, variations of the concrete pressure after it reaches the maximum value must be considered because lateral stresses of adjacent soils copy all these variations It appears a deficiency that only the maximum concrete pressures are accounted in current consideration and modeling of the wall construction process

A wall-construction modeling procedure

A reasonable modeling procedure for the entire wall construction process will concern the variations of concrete pressures during the filling and setting of wall concrete Soil consolidation will be simulated during the standing period of the wall

Chapter 2 – Research Background & Literature Review

a passive zone beneath the excavation depth Therefore, this passive resistance of soils might influence wall displacements, subsoil movements and the distribution of supporting forces and moments in the support structures

During the excavation of each excavation stage, soils on excavated-side of the retaining wall experience a transient passive condition conjointly produced by the two progressive processes of vertical-stress relief (unloading) and inwards wall displacing The situation becomes even more complicated in the case of clays in which the consolidation (swelling) of the clays will accompany with these two processes This particular condition is not concerned in current practice of estimation passive resistance of clays in deep excavations The calculation of the passive resistance is roughly based on the conventional Rankine’s passive earth pressure theory, which was originally developed for backfills of non-cohesion soils and assumed implicitly a steady condition of soil stresses and pore water pressures

For non-cohesion soils, the Rankine’s formula is naturally expressed in terms of effective stresses:

P P

sin 1

Chapter 2 – Research Background & Literature Review

The Rankine’s passive pressure formula is still widely used in engineering practice not because it can accurately evaluate the passive earth pressure mobilized in soils, but because that few cases of the passive earth pressure in soft clays have caused troubles (Tanaka, 1994) In most cases of deep excavations in ground with soft clay layer, the walls are designed to key into a deep firm stratum where a sufficient passive resistance can be mobilized When the firm stratum is quite deep, soil stabilization is commonly conducted for soft clay in passive zone at the excavation base

to increase the soil’s passive pressure and resistance against base failure (Tanaka, 1993; Ou et al., 1996; Uchiyama et al., 1999; Xie et al., 1999) However, in cases without the soil stabilization, passive resistance of the soft clay can be a substantial factor that strongly influences the wall displacements and subsoil movements (Tanaka, 1994)

2.5.2 Previous Studies

There were quite few published studies concerning passive earth pressures of soft clay in deep excavations (Ichihara et al., 1977; Sugimoto, 1985; Tanaka, 1994; Tamano et al., 1996; Hashash & Whittle, 2002) In these studies, passive earth pressures calculated by the Rankine’s formulas were widely used as reference values However, these studies all notified substantial deficiencies of the Rankine’s solution, and attempts were made to elucidate the deficiencies and to improve the calculation of the passive earth pressures

Ichihara et al (1977) and Sugimoto (1985) conducted physical model tests to investigate the development of passive resistance of overconsolidated soft clay against rigid walls during the wall-displacing period The Rankine’s solution was shown slightly smaller than experimental results, which was elucidated due to the disregard of soil-wall adhesion The solution of Sokolovski with soil-wall adhesion c a =(12~2 3)s u was considered reasonably close to the experimental results It was identified in these experimental studies that modes of the wall displacing, i.e translation or rotation, could influence the passive resistance as well as deformation and failure of the soils

It is worth to notice that in the experimental tests conducted by Ichihara et al (1977) and Sugimoto (1985), a similar experimental procedure was utilized for the tests A backfill of remolded soft clay was loaded, unloaded and left to swell completely to arrive at an overconsolidated condition and achieve significant shear strength; then a rigid wall was displaced shortly towards the backfill while passive resistance acting against the wall was being monitored Evidently, this simple experimental procedure could not simulate the complicated loading condition of clay in excavated side of deep excavations where the passive resistance of soils is

Chapter 2 – Research Background & Literature Review

18

wall displacing These experimental studies should be acknowledged for contributing comprehension on characteristics of passive behavior of overconsolidated clays, rather than for reasonably simulating the responses of soft clay during the actual excavation process

Tanaka (1994) compared field measurements of lateral earth pressure in front of the sheet piles with the Rankine’s passive earth pressures calculated with undrained shear strength s It was u

found that the calculated results with s derived from u q ( u s u =q u 2 , q : unconfined u

compression strength) were significantly smaller than measured pressures The calculation with undrained shear strength obtained from triaxial extension tests s was suggested to give values ue

well approximate the field measurements However, it was realized and elucidated that many other important factors should be considered such as soil-wall adhesion, stress-strain relation, shearing rate, soil swelling, confining conditions and so on

Tamano et al (1996) suggested an empirical correlation based on field observations from a case study of diaphragm wall excavation in soft clay As indicated in Figure 2.5, the variation of passive resistance of soft clay on excavated-side was considered induced separately by the decrease of vertical effective stress ∆σ z′ and by the wall displacementδ : ∆P P =K h δ+K0∆σ z′, where K was denoted as deformation stiffness of subsoil and h K was lateral stress coefficient 0

Good elucidation of field measurements was demonstrated using this empirical correlation A remaining uncertainty in this correlation is that variation of pore water pressures was included implicitly in the factor K of which values and physical meanings are unclear h

In a recent analytical study of Hashash and Whittle (2002), the Rankine’s solution was employed

as a reference in evaluating predictions of finite element analyses Similar to the suggestion of other studies (Fourie & Potts, 1991; Tanaka, 1994), values of the undrained shear strength

u

s used for calculating the Rankine’s passive earth pressures were derived from undrained triaxial

extension tests

Figure 2.5 Variations of effective earth pressure on excavated-side of deep excavation estimated as function of wall displacement and decrease of vertical effective stress (after Tamano et al., 1996)

K h x δ

K 0 x σ' z

0 10 20 30 40 50 60

Wall displacement (mm)

Effective earth pressure

K h x δ

K 0 x σ' z

0 10 20 30 40 50 60 70 80

0 10 20 30 40 50 60 70 80

Wall displacement (mm)

Effective earth pressure

K h x δ

K 0 x σ' z

(c) At G.L.-16.5m (d) At G.L.-18.0m (a) At G.L.-13.5m (b) At G.L.-14.5m Chapter 2 - Research Background & Literature Review

Chapter 2 – Research Background & Literature Review

As for the anchor ultimate pullout resistance, there are major concerned issues as the following:

- The construction of under-reamed anchors is unconventional and can affect the stress state condition and shear strength of adjacent subsoil In structured clay, the effect can

be much more substantial due to the soil’s high sensitivity Since the intact structured clay always has two distinct values of shear strength: peak value and residual value, it should be decided which value to be adopted in evaluation the anchor pullout resistance Besides, it is needed to estimate the strength reduction factor α for calculating shear strength of the soil-grout interface (interface of subsoil and the grouting anchor body) There is no knowledge in current engineering practice regarding these issues

- The structured clay reveals strong softening behavior in shearing, similarly to dense sand Since the asynchronism of ultimate pullout resistance of under-reamed anchors constructed in dense sand was noted substantial (Hsu & Liao, 1998), a similar phenomenon is anticipated in case of splits anchors embedded in structured clay

Other concerned issues are related to the time dependent performance of anchors constructed in clays In anchorage engineering, field tests remain requisite and provide the most reliable information for evaluating the capacity and performance of anchors However, even in “long-term” relaxation tests, experimental anchors are commonly loaded and monitored in relatively short period (several days) compared to the serving life-time of working anchors (several months

to years) Because short-term and long-term behaviors of clay are distinctively different, it probably becomes problematic when one deduces the performance of long-term working anchors directly from observations in short-term field tests

In brief, for splits anchors constructed in structured clay, it should be concerned both reamed-anchor involved issues and structured-clay involved issues These issues, including ultimate pullout resistance, suction force, relaxation and creep, and practical significance of pre-loading effects, are to be examined in this investigation

under-Chapter 2 – Research Background & Literature Review

2.6.1 Ultimate Pullout Resistance

Under-reamed anchors can provide large pullout resistance upon their huge peripheral shaft surface and substantial contribution of an end resistance that develops at the top of the anchor-enlarged body (Hsu & Liao, 1998; Kanaoka et al 2002; Liao & Hsu, 2003) Splits anchors were successfully tested for uplift capacity at about 800kN in soft to medium stiff clay ground (SPT ≈3; D 800= mm, L free =4.0m, L fix =1.5m) and more than 1300kN in medium dense sandy ground (SPT ≈9; D 800= mm, L free =8.0m, L fix =1.5m) (Suga & Tonosaki, 2001)

In engineering practice of estimating uplift capacity of under-reamed anchors, the discipline of superposition is usually employed where the ultimate pullout resistance (uplift capacity) is the sum of ultimate shaft resistance and ultimate end resistance (Caltrans Manuals, 1994; Stephenson, 2003) In studies on under-reamed anchor in sands, Hsu and Liao (1998) noted that the shaft resistance and the end resistance would not reach their ultimate magnitudes at the same movement of the anchor body, especially for deep anchors From another point of view, failure zone of the end resistance was known strongly affected by the stress-strain field created by movements of anchor cylindrical body (Ghaly, 1999); thus, the end resistance hardly can attain its ultimate magnitude at practically allowable movement of the anchor body Moreover, softening behavior of the subsoil due to dilatancy (dense sand) or debonding effect (structured clay) is another factor that invalidates the application of the superposition discipline

Accordingly, the actual ultimate pullout resistance of under-reamed anchors is commonly smaller than the sum of ultimate magnitudes of the resistance components The ultimate pullout resistance of splits anchors embedded in structured clay, which exhibits a strong softening behavior due to debonding effect, should be elucidated with these concerns

2.6.2 Suction Force

A suction force is probably generated at the bottom of plate anchors and under-reamed anchors embedded in clayey ground when the anchors are pulled (Baba et al., 1994; Das et al., 1994; Merrifield et al., 2003) The suction force is dissolving with time and does not develop if the pullout load increases gradually It has been suggested to disregard practically this force considering that the force is temporary and meaningless for the anchors’ long-term resistance

In field tests of anchors embedded in clays, since the pullout load is applied sufficiently fast, a consequential suction force surely develops and is supposed to have essential influences on the observed performance of the anchors The splits anchor has a considerably larger cross section

Chapter 2 – Research Background & Literature Review

22

that magnifies the suction force’s contribution fraction and influences Hence, it is necessary to comprehend this phenomenon in order to interpret appropriately the field test results

2.6.3 Relaxation and Creep

Relaxation and creep are two particular features of anchors While the relaxation test is performed at design load level and aimed at evaluating the load loss in long-term performance of the anchor, the short-term creep test is usually performed up to the anchor’s ultimate load and creep movement is concerned as a failure criterion of the anchor’s pullout resistance The two phenomena would have a close correlation, as they are two aspects in the time-dependent performance of anchors (Kim, 2003)

Both load loss and creep movements have been known induced by the creeps of the steel tendons, tendon-grout bond and mostly by the dragging at the grout-soil interface (Briaud et al., 1998) When the anchors are embedded in clays, the consolidation process becomes a decisive factor The contributions of the end resistance and the suction force are significant in case of splits anchors; hence, the activity of the consolidation effect is more substantial, or rather determinant This aspect has not been addressed in studies of straight shaft anchors

2.6.4 Preloading Effect

Pre-loading effect was comprehended in field tests of straight shaft anchors embedded in clays where the anchor’s creep rate reduced significantly in a reloading test conducted hours after the first load test (Briaud et al., 1998) The mechanism of this pre-loading effect was not clarified and

it remained unknown whether the effect is temporary or permanent If it is permanent, the preloading effect will has a practical significance because an anchor overloaded in a pre-loading stage (pre-stressing) will perform less susceptibility to creep (and load relaxation) while carrying the long-term working load The improvement is especially significant for splits anchors, which are supposed more susceptible to creep and relaxation than normal shaft anchors

Chapter 2 – Research Background & Literature Review

2.7 FINITE ELEMENT MODELING

In this study, numerical modeling was performed using Plaxis version 8.2, a soil-water coupled finite element program intended for two-dimensional plane strain analyses and axisymmetric analyses of deformation and stability in geotechnical engineering (Brinkgreve, 2002) The program provides modeling with advanced constitutive models for the simulation of the non-linear, time-dependent and anisotropic behavior of soils Plaxis also equipped with features to deal with various geotechnical aspects like interfaces, anchors, initial stress generation, ground water generation, staged construction, etc

In the following text, basic theories and formulations utilized in the Plaxis program are summarized Soil models used in the analyses including Mohr-Coulomb model, Soft Soil model, and Hardening Soil model are briefly described, and the selection of soil model is discussed Special modeling features that were utilized in the analyses in this research, like initial stress generation and interfaces, are described

2.7.1 Basic Theories and Formulations

Deformation theory, ground water flow theory and consolidation theory are basic theories on which a soil-water coupling finite element program, like Plaxis, is based

(a) Deformation theory

The static equilibrium of a continuum can be formulated as:

0

=+ p

The equation relates the spatial derivatives of the six stress components, assembled in stress

zx yz xy zz yy

assembled in force vector p L is the transpose of a differential operator defined as: T

z x y

z y

x

L T

00

0

00

0

00

0

(eq 2.2)

The kinematics relation is formulated as: ε =L u (eq 2.3)

Chapter 2 – Research Background & Literature Review

24

The equation expresses the six strain components, assembled in strain vector ε as the spatial derivatives of the three displacement components, assembled in displacement vector u using the previously defined differential operator defined L The link between the two equations is formed

by a constitutive relation representing the material behavior; a general relation is in form:

ε

σ& =M & (eq 2.4), where M represents the material stiffness matrix

The three equations (2.1), (2.3) and (2.4) are basic equations of continuum deformation

(b) Ground water flow theory

Flow in a porous medium can be described by Darcy’s law Considering flow in a verticalx−y−plane, the following equations apply:

x k

γ

where yis the vertical position, pis the stress in the pore fluid (negative for pressure) and γ is w

the unit weight of the pore fluid For steady flow, the continuity condition applies:

0

=

∂

∂+

∂

∂

y

q x

σ

m= 1 1 1 0 0 0 (eq 2.9)