Characterization of lumpy fill in land reclamation

CHARACTERIZATION OF LUMPY FILL IN LAND RECLAMATION VIJAYAKUMAR ALAPAKAM NATIONAL UNIVERSITY OF SINGAPORE 2006 CHARACTERIZATION OF LUMPY FILL IN LAND RECLAMATION VIJAYAKUMAR ALAPAKAM B.Tech (SVU), M.S.(IIT Madras) A THESIS SUBMITTED FOR THE DEGREE OF MASTER OF ENGINEERING DEPARTMENT OF CIVIL ENGINEERING NATIONAL UNIVERSITY OF SINGAPORE 2006 ACKNOWLEDGEMENTS I would like to express my deepest gratitude and heartfelt thanks to Associate Professor Tan Thiam Soon for his invaluable advice and patient guidance throughout the entire course of this research. Despite his busy schedule as the Dean of Admissions, NUS. He tried to meet me with the best time he could find. Not only did he give me a lot of help technically, but also his advice to my personal and career development. Every discussion with him was very motivating and he brought out the best in me. I would also like to thank to Dr.R.G.Robinson, former research fellow in National university of Singapore, that brought a success to this research is highly appreciated. I would also like to thanks to Dr.Karthikeyan.M and Dr.Yang Li-Ang for their continuous encouragement and support throughout the entire course of this research. I would like to express my sincere thanks to the staff of the Geotechnical Engineering Group for their kind help and co-ordination. I have to acknowledge the National University of Singapore for the award of research scholarship from January 2003 to December 2005. Thanks are also due to all my friends who have helped me in one way or another during this period. Last but not least, I am deeply grateful to my parents and my wife Manjula for their constant encouragement and moral support. Vijaya Kumar.A i TABLE OF CONTENTS Page Acknowledgements i Table of Contents ii Summary v Nomenclature vii List of Figures ix List of Tables xv CHAPTER 1 Introduction 1.1 Background 1 1.2 Land reclamation in Singapore using Lumpy fill 2 1.3 Objectives of this research 4 1.4 Organization of the report 5 CHAPTER 2 Literature Review 2.1 Introduction 9 2.2 Experimental studies on Lumpy clay fill in land reclamation 10 2.2.1 Void Index and Intrinsic Compression Line 17 2.2.2 Possible structural mechanisms 19 2.3 Case Histories of Land Reclamation Using Lumpy Fill 20 2.4 Numerical modeling 21 2.5 Concluding Remarks 24 ii CHAPTER 3 Experimental Investigation 3.1 Introduction 34 3.2 Composition and basic soil properties 35 3.3 Swelling test on clay lumps 36 3.4 Small scale One-Dimensional Compression Test 37 3.4.1 Experimental Setup 37 3.4.2 Preparation of clay lump specimens 38 3.5 3.4.2.1 Preparing of undisturbed clay lumps 38 3.4.2.2 Preparing of remoulded clay lumps 38 3.4.3 Experimental Procedure and Instrumentation 39 3.4.4 Unconsolidated Undrained (UU) Triaxial test and Vane shear test. 41 Large scale One-Dimensional Compression Test 42 CHAPTER 4 Results and Discussion - I 4.1 Introduction 50 4.2 Softening of Clay Lumps 51 4.3 Experimental studies on small scale size lumps 53 4.3.1 Void Index 53 4.3.2 Lump Size and Shape 57 4.3.3 Degree of Swelling of Clay Lumps 60 4.3.4 Initial packing of lumps 64 4.3.5 Pre-consolidation pressure of lumps 67 4.3.6 Lumpy Fill in Water and Clay slurry 69 4.3.7 Pore pressure response 72 4.3.8 Shear strength profiles of the lumpy fill 74 4.3.9 Secondary compression 75 4.4 Experimental studies on large scale size lumps 77 4.4.1 Time-settlement curves 77 iii 4. 5 4.4.2 Consolidation characteristics 78 4.4.3 Shear strength profile 80 Summary 81 CHAPTER 5 Results and Discussion – II 5.1 Introduction 117 5.2 Fabric and hydraulic conductivity 118 5.3 Possible structural mechanisms 121 5.3.1 123 Lump size and Degree of swelling of lumps 5.4 Shear resistance 126 5.5 Laboratory versus field behaviour 127 5.6 Summary 129 CHAPTER 6 Summary and Conclusions 6.1 Summary of Work 144 6.2 Conclusions 145 REFERENCES 148 APPENDIX 152 iv SUMMARY The disposal of large volumes of clay lumps obtained from underground construction activities and seabed dredging is a problem in a highly built-up environment such as Singapore. The use of such clay lumps for reclamation of land is an environment- friendly proposition. Also, the use of these lumps directly as reclamation fill provides a sustainable and cost effective solution to the disposal problem while recycling these materials into one with economic value. A major field study is in progress to evaluate the properties of such fill, as part of the Pulau Tekong Reclamation project, off the eastern shore of Singapore. However the behaviour of such fill is very complicated and is affected by the size, the degree of softening these lumps have been subjected to, the way these lumps have been arranged and the fluid that is filling the voids in between the lumps. A number of series of tests have been conducted to understand better the parametric influence of these various parameters. The lump sizes used range from very small lumps of size of 12.5mm onwards to lumps that are up to 200mm and which require specially design large testing tanks. These results are discussed in this report, with a particular emphasis on understanding what is the key factors affecting the overall behaviour. A lumpy fill is characterised by a dual void systems, referred to as inter-lump voids and intra-lump voids. It is clear that one of the most important parameters is the system overall void ratio, where the lumps themselves are considered like solid, in relation to the inter-lump voids. A new concept void index is introduced for study of the consolidation behaviour of lumpy fill in land reclamation. Lumps with varying sizes but similar system overall void ratio are shown to produce fairly similar behaviour. Another v important parameter is the degree of softening. The consolidation behaviour of lumpy fill is strongly influenced by the degree of swelling. Also the pre-consolidation pressures of lump have significant effects on the compressibility behaviour of lumpy fill. Pore water measurement in the inter and intra lumps voids and shear strength measurements under different loading stages are also used to evaluate the closing of inter and intra lumps voids present the lumpy fill system. Olsen (1962) cluster model is also used to study the changes in void ratio and permeability in inter and intra lump voids. Also, the possible stages of consolidation behaviour in lumpy fill were discussed. Finally, confirm the phenomena observed in the laboratory, the few laboratory results were compared with field studies. Key words: Land reclamation, lumpy fill, clay lump, consolidation, void index, interlump voids, intra-lump voids, shear strength, secondary compression, pore pressure vi NOMENCLATURE σ'v = Consolidation Pressure σvo = Vertical overburden pressure Cc = Compression index C*c = Intrinsic compression index Cα = Coefficient of secondary compression Cs = Swelling index Cαs = Secondary swelling index eo = In-situ void ratio e*30 = Void ratio on ICL for σv' = 30kPa e*100 = Void ratio on ICL for σv' = 100 kPa e*800 = Void ratio on ICL for σv' = 800 kPa e*1000 = Void ratio on ICL for σv' = 1000 kPa eer = Inter-lump void ratio era = Intra-lump void ratio ec = Cluster void ratio eL = Void ratio at liquid limit eT = Total void ratio ep = Inter-cluster void ratios es = System void ratio e50 = Void ratio of homogeneous lump with Us=50% Iv = Void Index Cu & S u = Undrained shear strength eiv = Inter lump-void ratio k = Coefficient of permeability P'c = Pre-consolidation pressure t = Time t50 = Time for a degree of swelling of 50% vii Viv = Volume of inter-lump voids Vl = Total volume of lump Su = Undrained shear strength w = Water content at any time t after submerging in water wi = Initial water content of the lump wf = Water content of the swollen clay lump qc = Cone tip resistance qCM = Estimated flow rate for a cluster model qCM = Flow rate predicted by the Kozeny-Carman equation qt = Corrected cone tip resistance N = Total no of cluster / no of lumps Nkt = Cone factor Us = Degree of swelling Abbreviations ADU = Autonomous data acquisition unit ICL = Intrinsic Compression Line ICC = Intrinsic compression curve ILV = Inter-lump voids LVDT = Linear variable displacement transducer. OCR = Over consolidation ratio IML = Intermediate layer LMC = Lower marine clay NC = Normally Consolidated ND-CPT = Nuclear-Density cone penetration test OC = Over-consolidated RI = Radioisotope SSMC = Surface soft marine clay UMC = Upper marine clay viii LIST OF FIGURES Figure No. Title Page No. Fig.1.1. Original Land Profile versus exiting and future reclamation Profiles (After The Straits Times, 2000). 6 Fig.1.2. Fill materials used in Singapore land reclamation projects (after the Ministry of information and communication, 1989). 6 Fig. 1.3 Dredging of seabed using clamshell grab. 7 Fig.1.4. Dredged lumps placed in a barge. 7 Fig.1.5. Dumping of clay lumps by bottom-open barge. 8 Fig.1.6. Schematic profile of land reclamation fill with clay lumps. 8 Fig.2.1. Compressibility Curves in incremental tests (after Mendoza and Hartlen, 1985). 26 Fig.2.2 26 Pore pressure built up during the test (after Mendoza and Hartlen. 1985). Fig.2.3. Variation of surface settlement with time at different load steps (after Leung et al., 2001). 27 Fig. 2.4. Effect of lump shape on settlement of lumpy fill (after Leung et al., 2001). 28 Fig.2.5. Excess pore pressure isochrones for lumpy fill made up of irregular lumps (after Leung et al., 2001). 28 Fig.2.6 Singapore marine clay lump of 205mm diameter before and after three dimensional swelling (after Robinson et.al (2004)). 29 Fig.2.7 Variation of water after three dimensional swelling 205mm diameter Singapore marine clay lump (after Robinson et.al (2004)). Fig.2.8. The use of void index Iv to normalize intrinsic compression curve. 29 Fig.2.9 30 Cluster model for permeability prediction (after Olsen,1962). Fig. 2.10 Assumed relationship between the total, cluster, and inter cluster void ratios (after Olsen, 1962). 30 31 ix Fig.2.11. Hydraulic conductivity discrepancies according to the cluster model (after Olsen, 1962). Fig. 2.12 Shear strengths measured by vane test(x) and unconfined compression test (o) (after Hartlen and Ingers, 1981). 32 32 Fig.2.13 Spring - Box analogy (after Yang et al., 2002). 33 Fig.2.14 Pore pressure profile for self weight only, and self weight plus surcharge after two days (after Yang et al., 2002). 33 Fig.3.1. Typical clay lump of Singapore marine clay obtained at a depth 14 m below seabed. 45 Fig.3.2 Clay lumps undergoing swelling process. 45 Fig.3.3. Typical experimental set up of lumpy fill system. 46 Fig.3.4. Preparation of cubical lumps by using a wire saw. 47 Fig.3.5 25mm cubical lumps used in experiments. 47 Fig.3.6 Schematic view of rectangular Tank of length 1.4m, width of 1.0 m and height of 1.5m. 48 Fig.3.7 Photo view of rectangular Tank of length 1.4m, width of 1.0 m and height of 1.5m. 49 Fig.3.8 Photographic view of rectangular tank filled with 200 mm size cubical lumps 49 Fig.4.1 Comparison of degree of swelling for different size and initial water content of cubical clay lumps. 84 Fig.4.2 Comparison of degree of swelling for different pre-consolidation pressure of 25mm remolded cubical clay lumps. 84 Fig.4.3 State of clay lumps after swelling process. 85 Fig.4.4 Schematic profile of reclamation with dredged clay lumps and sand Surcharge. 86 Fig.4.5. The apparatus on One-dimensional consolidation tests of lumpy clay. 86 Fig.4.6 Typical e-log p and permeability curves for lumpy fill consolidation experiments. 87 Fig.4.7. One-dimensional compression curves for reconstituted soil samples 88 Fig.4.8. One-dimensional compression curves for reconstituted soil samples 88 x in terms of void index. Fig. 4.9. The use of void Index Iv to normalized compression curve of lumpy fill. 89 Fig.4.10 90 Typical dredged material obtained from lower marine clay. Fig. 4.11 Effect of size of lumps on e-log σ’v relationship. 90 Fig.4.12. Variation of voids index with consolidation pressure for different sizes of clay lumps. 91 Fig.4.13. 91 Permeability of lumpy fill made of 12.5mm, 25mm and 50mm cubical Lumps with 100% degree of swelling. Fig.4.14. Variation of consolidation pressure required to close the inter-lump voids of lumpy fill with size of the lump. 92 Fig.4.15. Effect of shape of clay lumps made up of 25 mm lumps with 100 % degree of swelling on Iv - log σ′v curves. 92 Fig.4.16. Permeability of lumpy fill made of 25mm for different shapes with 100% degree of swelling. 93 Fig.4.17. Variation of moisture content with distance from the centre of clay lump (Us = 50%). 93 Fig.4.18. Effect of degree of swelling, Us on e- log σ′v of lumpy fill made of 50mm cubical lumps. 94 Fig.4.19. View of the lumpy fill made of 50mm cubical lumps under consolidation 94 pressure of 50kPa with lumps of (a) Us=0%, (b) Us= 50%, (c) Us= 100%. Fig.4.20. Permeability of lumpy fill made of 50mm cubical lumps for different Degree of swelling. 95 Fig.4.21 Typical time-compression curve of lumpy fill made up of 50mm cubical lumps under consolidation pressure of 50 kPa. 95 Fig.4.22 Variation of voids index with consolidation pressure for different degree of swelling for lumpy fill made of 50mm size cubical lumps. 96 Fig.4.23. Variation of consolidation pressure required to close the inter-lump voids of lumpy fill with degree of swelling. 96 Fig.4.24. Photographs of lumpy fill made up of 50 mm cubical lumps with Us= 100% and eiv = 1.68 under different loading stages. 97 Fig.4.25 Typical time-compression curve of lumpy fill made up of 50 mm 98 xi cubical lumps with Us=100% for (a) eiv = 2.25, (b) eiv = 1.68 and (c) eiv=1.41. Fig.4.26. View of the lumpy fill made of 50mm cubical lumps under consolidation pressure of 50kPa with lumps of (a) eiv = 1.41 (b) eiv = 1.68, (c) eiv = 2.25. 99 Fig.4.27 Effect of initial inter-lump void ratio of lumpy fill made up of 50 mm cubical lumps with 100% degree of swelling on Iv - log σ′v curves. 99 Fig.4.28. Permeability of lumpy fill made of 50 mm cubical lumps with different initial inter-lump void ratio. 100 Fig.4.29. Variation of consolidation pressure required to close the inter-lump voids of lumpy fill made of 50mm cubical lumps with initial inter-lump void ratio. 100 Fig.4.30 Geological location of Reclamation site. 101 Fig.4.31 Geotechnical Characteristics of Seabed soil prior to reclamation (After Karthikeyan, 2005). 101 Fig.4.32. Effect of pre-consolidation pressure of clay lumps made up of 50 mm cubical lumps with 0% and 100% degree of swelling on Iv - log σ′v curves. 102 Fig.4.33. Permeability of lumpy fill made of 50mm cubical lumps with 0% and 100% degree of swelling for different pre-consolidation pressure 102 Fig.4.34. Variation of consolidation pressure required to close the inter-lump voids with degree of swelling for a pre-consolidation pressure of 100kPa and 350kPa. 103 Fig.4.35 Interpreted soil profile at Pulau Tekong (After Tan et.al). 103 Fig.4.36 Typical Cone resistance and pore pressure Profile of seabed. 104 Fig. 4.37 A view of the surface soft marine clay dredged from the seabed (SSMC). 104 Fig.4.38 Typical time-compression curves for inter-lump voids (ILV) filled with water / slurry. 105 Fig. 4.39 Effect of inter lump voids filled with slurry / water for the lumpy fill made up of 50 mm lumps on Iv - log σ′v curves. 106 Fig.4.40 Variation of permeability of lumpy fill made of 50mm cubical lumps with inter-lump voids filled with slurry/water. Pore pressure under pressure increments of (a) 25-50kPa; and 106 Fig.4.41 107 xii (b) 100-200kPa (inter-lump voids filled with water). Fig.4.42 Spring - Box analogy (after Yang et al., 2002). 108 Fig.4.43 Pore pressure under pressure increments of 25-50kPa and 100-200kPa (inter-lump voids filled with slurry). 108 Fig.4.44 Shear strength profiles obtained from cone penetration tests for lumpy fill made of 50mm cubical lumps with eiv=2.25. 109 Fig. 4.45 Secondary compression of lumpy fill. 110 Fig.4.46 Variation of surface settlement with time at different load steps. 111 Fig.4.47 Variation of normalized settlement with time at different load steps. 111 Fig.4.48 Fig.4.49 Closing of inter- lump voids. One dimensional compression curves for reconstituted soil sample under 50, 100 and 200kPa. 112 113 Fig.4.50 Comparison of one-dimensional compression curves for undisturbed 114 samples with ICL under surcharge pressure of 50kPa, 100kPa and 200kPa. Fig.4.51. The variation of (a) Water content, (b) pre-consolidation pressure and (c) Compression index with depth from bottom surface of the lumpy fill layer for surcharge pressure of 50kPa, 100kPa and 200kPa. 115 Fig.4.52 Variation of compression Index with void ratio. 116 Fig.4.53. Shear strength profiles obtained from cone penetration test for 100 kPa and 200kPa surcharge pressure. 116 Fig.5.1 Cluster model for permeability prediction (after Olsen, 1962). 131 Fig. 5.2 Assumed relationship between the total, cluster, and inter-cluster void ratios (after Olsen, 1962). 131 Fig.5.3 Assumed relationship between the total, inter and intra lump void ratios a) 12.5 mm b) 25 mm c) 50 mm. 132 Fig. 5.4 Permeability discrepancies according to the cluster model for different size of lumps. 133 Fig. 5.5 Permeability discrepancies according to the cluster model for Different degree of swelling. 133 Fig.5.6 Typical Permeability variation in the lumpy fill under consolidation pressure. Modified assumed relationship between the total, inter and intra 134 Fig.5.7 134 xiii lump void ratios. Fig.5.8 Assumed relationship between total, inter and intra lump void ratios. 135 Fig.5.9. Permeability of lumpy fill made of 12.5mm, 25mm and 50mm cubical Lumps with 100% degree of swelling. 136 Fig.5.10. Water content variation along the depth of lumpy fill for 100% swelling under different consolidation pressure for (a) 12.5mm (b) 25mm and (c) 50mm. 137 Fig.5.11. Variation of inter and intra lump void ratio with consolidation 138 pressure for lumpy fill made of 12.5mm, 25mm and 50 mm cubical lumps. Fig.5.12. Variation of inter and intra lump void ratio with consolidation pressure for lumpy fill made of 50 mm cubical lumps with 0%, 50% and 100% degree of swelling. 139 Fig.5.13. Triaxial shear strength for (a) Different size of lump, (b) Different degree of swelling and (c) Different pre-consolidation pressure lumps. 140 Fig.5.14. Vane shear strength for lumpy fill made up of 50 mm cubical lumps with 0% degree of swelling for (a) Pc =100 kPa and (b) Pc=200 kPa. 141 Fig.5.15 Generalization of compression paths for laboratory and field data. 141 Fig.5.16 (a) Typical cone resistance for Test area TA4. (b) Comparison of pre-consolidation pressure measured by the ND-CPT in TA4 with the laboratory experiment results using 200mm cubical lumps. 142 Fig.5.17. Typical cone resistance, wet density and void ratio profiles at location TA2-1-46 and TA4-1-18. 143 xiv LIST OF TABLES Figure No. Title Page No. Table.3.1 Physical properties of the soil used. 43 Table 3.2 Initial condition of the lumpy fill system for small scale consolidation experiments. 44 Table.4.1 Time required reaching t50 and t100. 85 xv CHAPTER 1 Introduction 1.1 Background The rapid economic development over the last three decades in Singapore has led to a continuous demand for land to be used for housing, transportation, commercial and industrial needs. But land is scarce in this country. One of the solutions to this problem is to create land from sea by land reclamation. Fig.1.1 shows the reclaimed land in Singapore from 1960’s to 1990’s. The city-state has grown from an area of 580 square kilometers to an estimated area of 662 square kilometers today, an increase of 17 percent (Lui and Tan, 2001). Land reclamation has become an integral part of Singapore’s development. In the beginning, hill-cut materials and sand deposits in the surrounding sea-bed have been used as fill materials. Fig.1.2 shows the fill material used in previous Singapore land reclamation projects (Ministry of Information and Communication, 1989). The two sources were limited, and virtually exhausted by the early 1980s (Ministry of communications and information, 1989). Since then until two years ago, dredged sand imported from neighboring countries is used as fill. The cost of reclaimed land per square meter has increased many folds during the last two decades partly due to the increasing cost of imported sand. Another important factor to note for future reclamation is that most of the coastal areas around Singapore comprising the local 1 continental shelf have been reclaimed. Future reclamation works therefore have to contend with depths greater than 5 meters even up to a depth of 10 to 20 meters. In this case, importing sand is not only likely to be more costly, but there is also the likelihood of inadequate supply for large reclamations. Therefore, the need for alternative fill materials for future reclamation projects in Singapore has become increasingly evident. In Singapore, large volumes of clay lumps are obtained from underground construction activities like deep basements for buildings, excavations and tunneling etc. Further, dredging is frequently done for the maintenance of navigation channels and port construction activities. In addition, large quantities of dredged clay are obtained from excavation done in the seabed for the construction of sand bunds and sand keys for land reclamation purposes. The disposal of the clay lumps is a constant challenge, and one ideal solution is to use it as an alternate fill for the reclamation (Hartlen and Ingers, 1981 and Karunaratne et al., 1991). Using such dredged and excavated materials for land reclamation is an attractive proposition for solving the environmental problem of finding suitable dumping ground for the disposal of these materials, as well as creating new land for land-scarce Singapore. 1.2 Land reclamation in Singapore using Lumpy fill In the 1500 hectares propose Pulau Tekong reclamation, clay lumps from both seabed dredging and from construction activities are used as fill materials. The dredging of in-situ clay from the seabed is carried out using large clamshell grab(Fig.1.3), of size from 10m3 to 18m3 and often produces clay lumps with volume more than 1m3 as shown in Fig.1.4 ( Leung et al. 2001, Robinson et al. 2003). The dredged materials are placed in 2 a barge. The lumps are transported and dumped at the reclamation site by opening the bottom of the barge carrying them (Fig. 1.5). Fig.1.6 shows a schematic section of the reclamation site with dredged clay lumps and sand surcharge. During initial stages of reclamation, there would be large voids between the big dredged clay lumps and these inter-lump voids (void between lumps) could partly filled with small lumps. Due to consolidation and compression from the sand surcharge, average density and void ratio of lumpy fills constructed using dredged clay lumps changes dramatically. If the size of the inter-lump voids reduces to size of intralump voids (voids within lump), then it is considered that the lumpy fill has been “homogenized”(Karthikeyan, 2005). Therefore, monitoring of the changes in void ratio and permeability is important to facilitate the assessment of the state of ground at any time for a realistic understanding of the physical processes involved in the consolidation behaviour. There are very few studies concerning lumpy fill until the works by the NUS group. Manivannan et al. (1998), Leung et al. (2001), Yang (2004) and Karthikeyan (2005). Based on centrifuge tests at 100g using stiff silty clay lumps of 1cm3 in volume, Manivannan et al. (1998) concluded that substantial closure of inter-lump voids occurs under a surcharge loading of 120 kPa. One-dimensional compression test results of Leung et al. (2001) indicated that most of the voids between the clay lumps closed up during the first stage of loading of 25 kPa and the fill underwent substantial settlement. It is still a challenge to model the consolidation of such lumpy fill clay. In reclamation using such compressible fill materials, self-weight consolidation is important because the site will usually take years to fill with dredged and excavated materials before a sand fill 3 is placed on top. Yang (2004) introduced solutions to the idealized problem where the soil is assumed to behave like a double-voids system formed by inter-lump and intralump voids. However, the mechanics of lumpy fills is still unclear. Key concerns include an understanding of the consolidation pressure required to close inter-lump voids, the equilibrium state of the lumpy fill under different consolidation pressures, swelling of clay lumps, packing of lumps and size of lumps. An understanding of the behaviour of lumpy fill and the influence of these factors at the ultimate state would provide significant opportunity to further optimize the design. 1.3 Objectives of this research In connection with the Pulau Tekong reclamation project using lumpy clay, a study of settlement and consolidation behaviour of such lumpy clay fill is required. The objectives of this work are i) To evaluate the significance of the effect of size and shape of lump, degree of swelling, pre-consolidation pressure of lump and various packing arrangement on the consolidation behaviour and the ultimate state of lumpy fill. ii) To study the compression characteristics of the lumpy clay using Burland (1990)’s intrinsic framework and to propose method of void index to explain the compressibility behaviour of the lumpy fill. iii) To study the structural and fabric changes during compression of lumpy fill by using voids ratio and permeability data. 4 iv) To evaluate the laboratory results together with field results in order to have a better understanding of how the phenomena observed in the laboratory can be used. 1.4 Organization of the report Chapter 2 presents a review of the literature, which covers the general introduction to the characterization of lumpy fills and case histories of land reclamation using lumpy fill. It also reviews some laboratory experiments on layered sand scheme, slurry fill and lumpy fill and numerical studies. Chapter 3 presents the experimental set up and associated instruments needed to carry the tests. Also the basic physical properties of soil used in the present study are discussed in this chapter. Chapter 4 presents the experimental studies on lumpy fill using small and large size lumps. The results from one dimensional consolidation test were analyzed by using new concept of void index. Chapter 5 presents the possible structural mechanisms in the lumpy fill during surcharge load by using Olsen model with different size and degree of swelling of lumps. Finally the a few laboratory results were compared with field results. Chapter 6 presents the summary and conclusions drawn from this study. In addition, some recommendations for further research works are highlighted. 5 Fig.1.1. Original Land Profile versus exiting and future reclamation Profiles (After The Straits Times, 2000) Fig.1.2. Fill materials used in Singapore land reclamation projects (after the Ministry of Information and communication, 1989) 6 Fig. 1.3 Dredging of seabed using clamshell grab Fig.1.4(b) Dredged lumps placed in a barge Fig.1.4. Dredged lumps placed in a barge 7 Fig.1.5. Dumping of clay lumps by bottom-open barge Mean sea level Sand surcharge Clay lumps Inter-lump voids Sea bed Fig.1.6. Schematic profile of land reclamation fill with clay lumps 8 CHAPTER 2 Literature Review 2.1 Introduction Land is continuously created through reclamation in Singapore to meet the continuous demand for more land to be used for housing, transportation, commercial and industrial needs. The cost of imported sand is extremely high and alternate materials in huge volumes are required for future land reclamations which will be in deeper water, some as deep as 25 m. On the other hand, large volumes of dredged clay are obtained from underground construction activities inland and from the seabed during the maintenance of navigation channels and port construction activities, which need to be disposed off. These materials comprising mainly of highly plastic marine clay in the form of lumps, if used for reclamation, can solve the environmental problem of finding suitable dumping grounds as well as creating new land. Conventional practice in land reclamation is to transport and deposit the fill materials as hydraulic fills, which consists mainly of clay slurry with some small lumps suspended within the slurry (Casagrande 1949, Whitman 1970). However, lumpy fill constructions using large clay lumps have been successfully adopted in a few land reclamation projects (Hartlen and Ingers 1981, Leung et al. 2001). The use of large clay lumps in a major way is now being implemented in reclamation at the eastern coast of 9 Singapore Island and the present research is an integrated part of that project. The literature review relating to the consolidation behaviour of land reclamation using lumpy clay fill is discussed in the following sections. 2.2 Experimental studies on lumpy clay fill in land reclamation Casagrande (1949) and Whitman (1970) reported the use of hydraulic fills to reclaim low-lying areas and for the construction of embankment. Whitman (1970) observed that fills consisting of soft clay are highly compressible, and the compressibility can be estimated satisfactorily from laboratory tests on representative samples. However, the rate of consolidation in the field is generally much greater than that indicated by laboratory tests, owing to the stratified nature of the fill. Except where the fill is composed of clay lumps, laboratory tests on undisturbed samples can be used to provide an estimate of the compressibility of the reclamation fill. On the other hand, laboratory tests will be of little help in estimating the rate of consolidation. Hydraulically placed lumpy fills usually consist of small clay balls with suspended material filling the inter-lump voids as reported by Hartlen and Ingers (1981). On the other hand, there is little slurry or semi-fluid clay between large clay lumps and the large inter-lump voids are often filled with fluid in barge-dumped fills. Therefore, their engineering properties are different. For example, hydraulically placed lumpy fills could be more easily turned into homogenous clay under self-weight consolidation than barge-dumped lumpy fill according to Hartlen and Ingers (1981). Mendoza and Hartlen (1985) performed laboratory tests to compare the compressibility characteristics of clay slurries with different initial water contents, and 10 mixtures of clay balls and slurry. The mixtures of slurry and clay balls were prepared to reproduce the mass of the Halmstad hydraulic fill (Hartlen and Ingers, 1981). 5 to 50 mm diameter clay balls were obtained by cutting rough cubes of different sizes of clay blocks, and then put with sea water inside a rotating drum. Incremental consolidation tests in Rowe cells 152 and 254 mm in diameter with specimen 300mm in diameter and height were carried out. Compressibility curves showed clearly that the mixtures of clay balls and slurry had a much smaller compressibility than the slurry alone (Fig2.1). Also, it was observed that the pore pressures in clay balls were slightly higher than in the slurry between the clay balls (Fig.2.2). As the mixtures were loaded and the clay balls were encapsulated by slurry, the increase of pore pressure within the clay balls and in the slurry was almost concurrent. As a result, Mendoza and Hartlen (1985) implied that the mixtures of clay balls and slurry might turn into normal clay finally. Wong (1997) and Leung et al. (2001) studied consolidation behavior of clay lumps in reclamation fill through laboratory tests to model barge-dumped fill. One dimensional compression test conducted on 50mm spherical clay lumps showed that the lumpy fill layer compresses rapidly under an initial loading of 25kPa(Fig.2.3a) At subsequent loading stages, relatively smaller surface settlement was recorded (Fig2.3(b)). It was observed that the rate of consolidation was considerably faster for the lumpy fill than for homogenous clay, but the difference decreased considerably after the void spaces between the clay lumps closed up with increase in loading pressure. It was also revealed that the clay lumps located near the surface experienced large movement, while the lumps near the bottom of the container experienced relatively little settlement but considerably more deformation. 11 Parametric studies were also carried out using centrifuge modeling technique to evaluate the effect of lump size and shape and original in-situ shear strength of clay on the performance of lumpy fills. From this study, it was observed that the lumpy fill made up of spherical lumps settles less than the irregular and cubical lumpy fill due to the denser packing of spherical lumps in the lumpy fill (Fig 2.4). As the size of the lumps increases, larger settlement is observed. Clearly this aspect from the one-dimensional consolidation test for different sizes of lumps needs to be checked for 1g condition due to possible scale effects in the centrifuge. Leung et al. (2001) also noted that lumpy fill consisted of two pore pressure dissipation patterns: “One is the relatively fast dissipation of pore pressure at the inter-lump voids, and the other inside the lumps where the pore pressure dissipates at a much slower rate”. Fig.2.5 shows that for lumpy fill made up of irregular lumps, the pore pressure iscrones inside the lumps are considerably higher than those at fixed elevations under a surcharge of 120kPa. According to them, this is due to the individual lump’s soil skeleton having a much greater stiffness than the fillers in the inter-lump voids and therefore more loads would be carried by the lumps. A lumpy fill consisting of large inter-lump voids is shown schematically in Fig1.6. Generally, the inter-lump voids are expected to be filled with water. However, there are occasions where the inter-lump voids are filled with clay slurry. This situation arises when stiff clay lumps are dumped directly on very soft marine clay. Also when the dredged soft marine clay is dumped over the stiff clay lumps, it will fill the inter-lump voids. Robinson et al. (2003) investigate the influence on consolidation behavior due to the presence of clay slurry in the inter-lump voids of a lumpy fill. Comparison was made with the results obtained on lumpy fill whose inter-lump voids are filled with water. 12 Consolidation tests conducted in the laboratory on lumpy fill made of cubical lumps of 25 mm suggest that the presence of slurry in the inter-lump voids has significant influence on the consolidation behaviour during the early stages of loading. The rate of settlement of the system filled with slurry is much smaller than that filled with water for pressure less than about 50 kPa. Under higher pressure, these differences become smaller. Consolidation pressure required to reduce the global void ratio of the fill to the void ratio of a fully swollen homogeneous lump is about 25 kPa, which may be considered as the pressure required to substantially close the inter-lump voids. Permeability measurements and visual observations of movements of lumps under loads substantiate this. However end state of the lumpy fill filled with slurry is not significantly different from that without slurry. Yang Jun wei (2004) carried a series of large one dimensional consolidation tests on lumpy fill to evaluate the effect softening, lump size, permeability of the inter lump voids and inter-lump voids ratio. He found that the rate of softening is slower for larger lumps and the final water content of lumps after fully softening is close to its liquid limit. In addition, a numerical solution to evaluate the softening degree of single lumps was proposed. He also compared the experimental results with those of finite elements analysis using ABAQUS. In order to study the consolidation characteristics of the sludge cakes, Nishimura et al.(2003) conducted large-scale consolidation tests with sludge cakes whose preconsolidation pressure was, Ps = 157, 314 and 627kPa and the thickness of sludge cake pieces was 5mm and 20mm. During the consolidation of pieces of the sludge cakes, a major settlement occurred soon after loading and the consolidation almost converged 13 after 5-10 minutes. The settlement of the sludge cake pieces was larger than that of the sludge cake paste. Firstly, the large inter-lump voids between the sludge caked will be closed, and then the intra-lump voids shrink due to the compression of sludge cake. The time until the convergence of the major settlement was much shorter for the consolidation of the sludge cake pieces than for the paste. Besides, Manivannan et al., 1998 and Leung et al., 2001 who studied the use of big clay lumps in Singapore, no major field study together with laboratory testing has yet been carried out to evaluate the ultimate state of lumpy fills. Kathikeyan et al. (2004) have carried out an extensive site investigation at a test site on the island of Punggol Timor in Singapore, which was reclaimed about 12 years ago using big dredged clay lumps. The radioisotope cone penetration test was employed to measure the in-situ density of the site. The results indicated that the initially large inter-lump voids have been reduced to the size of intra-lump voids. However, the layer formed from clay lumps is heterogeneous and exhibits variable engineering properties. After placing the clay lumps in the reclamation site, they would be exposed to water and as a result, subjected to different degrees of swelling depending on the amount of time they were left to stand in water. When the system is allowed to stand, self weight consolidation of the lumpy fill will take place even after the first load is added. The amount of initial settlement due to self weight consolidation is highly affected by degree of swelling. The rate of swelling is mainly dependent upon the suction present in the clay lump. When clay lumps with large initial suction are used, a very large surcharge may be needed for closing the inter-lump voids, if the clay lumps are prevented from swelling. Swelling will result in lower effective stress within the lumps and reduce the requirement 14 of large surcharge pressure needed for closing the inter-lump voids. Hence an understanding of the swelling of clay lumps is important when dealing with reclamation using clay lumps. Mesri et al. (1978) conducted a series of tests to study the rate of swelling of over consolidated clays subjected to unloading. From their analysis of deformations of reconstituted soil specimens with time over 400 load decrements, both in onedimensional and isotropic tests for four different shale compositions, it was found that Terzaghi’s theory of swelling correctly predicted the percent swell – log time response up to 60% primary swell. However, beyond 60%, the shape of the swell curve is a function of the magnitude of Cαs / Cs and the load decrement ∆σ ' / σ 'f , where Cαs is the secondary swelling index, Cs is the swelling index, ∆σ' is the load decrement and σ 'f is the effective stress at the end of a load decrement. Calabresi et al. (1983) studied the effect of swelling by unloading two heavily over consolidated natural clays at very low stress, both in one dimensional and isotropic test. The response of the swelling soils is examined with respect to their compressibility and failure characteristics. Finally, it is not yet clear how time influences the development of swelling in the performed tests, but the observed phenomena conclude that larger swelling times at low stress would not alter the results significantly. AlTabbaa (1995) reported the non-monotonic increase in pore water pressure during consolidation and swelling with radial drainage. The soil sample was placed in a Rowe’s stress cell, radial drainage was allowed and sample was restrained in the lateral direction. As the soil samples are laterally confined and drainage was allowed only in the radial 15 direction, the results are not applicable for clay lumps in reclamation projects. But it gives some idea about the swelling behaviour. Wong (1997) studying the effect of water softening on model clay specimen by using fall cone test. The test involved dropping a suspended cone from a permanent magnet onto the flat surface of the clay sample, which was specially cut to fit into a Perspex mould (155mm diameter, 73mm high) submerged in a water bath. From this, it was observed that the rate of softening decreases with depth. Also moisture content test with lumps of different size and shape was carried. Based on these test results, it was observed that the rate of softening decreases with increasing lump size and increases with increases of surface area being exposed to surrounding water. Also, Robinson et al. (2004) investigated the swelling of clay lumps with free lateral boundary with all round drainage. To examine this aspect a series of experiments was carried out in which reconstituted cylindrical clay lumps of 105, 205 and 400 mm in diameter with height approximately equal to the diameter were soaked under water. The dissipations of suction at the centre of these samples were measured. Fig. 2.6 shows the photographs of 205mm diameter Singapore marine clay lump before and after threedimensional swelling. It was found that the state of the clay lumps were stable at the end of the dissipation of suction, as can be seen in Fig2.6 The variation of water content with depth of the sample at the three different locations was also measured, as shown in Fig.2.7. This figure shows that the measured water content at the end of the test is not uniform along the diameter of the specimen. The water content at the edge of the clay lump is higher than at the centre, a fact of some practical relevance. During the deformation process of the clay lumps in a lumpy fill under sand surcharge, the edge of 16 the clay lumps experience very high contact stress and will cause partial disintegration of edges of the bigger clay lumps while the centre of these lumps remains close to its original consolidated state. In addition to the experimental study, finite element analyses were performed using linear and non-linear elastic soil models. The final stage of dissipation of pore water pressure was much faster in experimental compared to the FE results due to micro cracking of the clay lumps, when the suction gets reduced to small values, leading to an increase in the permeability. Laboratory and field investigation are in progress to evaluate the characteristics and consolidation behavoiur of such dredged lumpy clay (Karthikeyan et al., 2004). However, it is still a challenge to model the consolidation of such lumpy clay fill. In particular, one difficulty is to monitor the closing of inter lump voids. In this present study, the effect of size of lump, degree of swelling, pre-consolidation pressure of lumps and packing arrangement were studied using modified burland frame work. The burland intrinsic frame works were discussed in detail in the next page. 2.2.1 Void Index and Intrinsic Compression Line In the characterisation of soil structure in terms of compressibility, Burland(1990) referred to the intrinsic compression curve (ICC) obtained from one-dimensional compression test on clay reconstituted in the laboratory with water content between 1 to 1.5 times liquid limit without air drying or oven drying. Ideally the chemistry of the water should be similar to that of the pore water in the clay in its natural state. He also proposed the replacement of void ratio with a void index in the study of the 17 compressibility of natural soil to eliminate the effect of variation in soil type. He defined void index as follows IV = * * e − e100 e − e100 = * * e100 − e1000 C c* (2.1) where e*100 and e*1000 are the void ratios corresponding to σ’v=100 kPa and σ’v=1000 kPa respectively, on the ICC (Fig.2.8a). C*c is the intrinsic compression index which is also equivalent to the change in void ratio on the ICC for an increase in effective stress from 100kPa to 1000kPa. As indicated in Fig.2.8b, when e =e*100, Iv = 0 and e = e*1000, Iv = -1. Burland (1990) showed that ICC from soils with a wide range of liquid limits, when plotted in terms of Iv-logσ’v, can be represented with sufficient accuracy by the following equation: Iv = 2.45-1.285X +0.015X3 (2.2) Where X = logσ’v in kPa. Burland (1990) termed this unique line in Iv-logσ’v space as the Intrinsic Compression Line (ICL). The ICL can be easily obtained experimentally. Alternatively, Burland (1990) showed that e*100 and Cc could be reasonably related to the void ratio at the liquid limit, eL, as follows: e*100 = 0.109 + 0.679eL - 0.089eL2 + 0.016eL3 (2.3) Cc* = 0.256eL – 0.04 (2.4) Karthikeyan (2005) has already demonstrated that the ICL and void index from oedometer tests can be used in the study of the compression characteristics of the ultimate state of a reclaimed land. Therefore, this framework will be used to study the compression characteristics of the lumpy fill layer under different conditions. 18 2.2.2 Possible structural mechanisms A given soil that has a fabric with a high proportion of large pores is much more pervious than those with small pores. Because of this, the hydraulic conductivity of remolded undisturbed soft clays has been found to reduce by as much as a factor of 4(Mitchell, 1956). This can be explained by the breakdown of a flocculated open fabric and the closer of large pores. Olsen (1962) introduced a soil model with a fabric composed of small aggregates of clusters as shown in Fig.2.9, having intracluster void ratio, ec. The spaces between the aggregates comprise the intercluster void ratio ep and the total void ratio eT is equal to the sum of ec and ep. Fluid flow in such a system is dominated by flow through the inter cluster pores. Olsen (1962) made assumptions on the variation of ec with eT as shown in Fig.2.10 for the cluster model so as to be able to calculate the ratio of estimated flow rates to predicted flow rates in the cluster. The compressibility of individual cluster is small at high total void ratios, so compression is accompanied by reduction in the intercluster pore sizes, but with little changes in intra cluster void ratio. As a result, the actual hydraulic conductivity decreases more rapidly with decreasing void ratio during compression as shown in Fig.2.11 until the intercluster pore space is comparable to that of the intra pore space. Further decrease in porosity involves decease in both ec and eT. As the intercluster void ratio now decreases at a slower rate with decreasing porosity than that predicted by the Kozeny - Carman equation, it means that most of the inter lump voids were closed. 19 2. 3 Case Histories of Land Reclamation Using Lumpy Fill Casagrande (1949) performed the first detailed evaluation of the use of hydraulic dredged clay lumps as reclamation fill materials during the construction of Logan Airport in Boston. The size of dredged clay lumps varied from 20mm to 200mm, which were laid down in a matrix of semi-fluid clay. They found that compressibility of lumpy fill resulted from the plastic deformation of the clay lumps, while the rate of consolidation was determined by the characteristics of the semi fluid clay surrounding the clay lumps. Whitman (1970) reported the construction of a dredged fill island, developed as the Lacustre Marine terminal, in Lake Maracaibo. Stiff clay was dredged and pumped into a sheet pile enclosure until the hydraulic fill reached 2 m above lake level. The material at the dredged point consisted of rounded clay lumps varying in size from 25 to 150mm. A series of load tests using small steel tanks was performed during early stages of filling. Whitman (1970) made conclusions similar to Casagrande (1949) that the compressibility of a lumpy fill was primarily due to deformation of the clay lumps, while the rate of consolidation was determined by the slurry surrounding the clay lumps. Hartlen and Ingers (1981) reported the reclamation work to create an approximately 1.2 km2 industrial area in Halmsted Harbor in Sweden. The total thickness of the fill was 6.4m, of which 5.0m was placed under water. The bottom 3.0m of the fill consisted of barge-dumped clay containing 1.0m3 lumps, on top of which a 3.4m layer of clay balls with diameter of up to 300mm was placed hydraulically (Fig.2.12a). They concluded that the fill tends to become homogeneous as a result of consolidation under self-weight. However, the scattered values of shear strength determined by the vane tests 14 months after the completion of filling (Fig.2.12b) poses the problem of whether self- 20 weight alone is able to close all the voids and make the fill homogeneous. Another possibility is that the complex nature of the interaction of the lumps with water causes this heterogeneity. Yap (2001) reported the use of large clay lumps in the construction of Pasir Panjang container terminal in Singapore. The instrumentation results indicated that large inter-lump voids were present immediately after dumping. Upon application of surcharge, large settlement was observed and it was attributed to plastic compression of the clay lumps. Bo et al. (2001) reported the use of dredged material and excavated material in the form of lumpy clay as fill material. It was suggested that dredged lumpy clay could be dumped directly after a sand blanket of one or two meters was placed onto seabed. Comparing with in-situ clay, the characteristics of lumps were changed slightly, and vertical and lateral volumetric changes of clay lumps were noted during the reclamation. They also concluded that the lumpy fill could be treated by prefabricated vertical drain together with foundation soil. However, whether settlement came mainly from lumps themselves or inter-lump voids was not discussed in this paper. 2.4 Numerical modeling Although there are a few case histories and tests related to hydraulic and bargedumped fills, much works, especially theoretical work, still need to be done on this topic. Only a few published papers touched on the analytical and numerical model of lumpy fills (Manivannan et al., 2000, Nogami et al., 2001 and Yang et al., 2002). It is still a big challenge to model the consolidation of such lumpy clay. Double porosity consolidation model, first used in the petroleum industry to deal with flow in a fractured medium may 21 be used to deal with the consolidation of lumpy clay in land reclamation. Manivannan et al (2000) carried the consolidation settlement analysis using a 2dimensional finite element programme based on the Biot’s theory. The lumps are modeled by regular spatial arrangement of blocks representing the large discrete chunks of clay lumps while the inter-lump voids are modeled as very soft slurry. From this numerical model, it was observed that the most of the inter-lump voids are closed immediately after applying the load. Also, the rate of dissipation of pore pressure in the inter-lump voids is faster than that in the lumps themselves. The same observations were observed by Leung et al. (2001) in their experimental study. Nogami et al.(2001) presented a simple one-dimensional lumpy clay consolidation model based on the double-porosity concept. Two governing equations were derived with the assumption that total stress changed only with varying load. They realized the important role of inter-lump voids in accelerating the soil consolidation. Parametric studies were also conducted to show the importance of three non-dimensional parameters in controlling the consolidation behavoiur of lumpy clay. However, selfweight and non-linearity was not mentioned in the above two paper, and these two aspects are more critical in the consolidation of a lumpy fill. Yang et al. (2002) has introduced an analytical solution to an idealized problem where the soil behaves in a linear elastic fashion with constant permeability and subjected to self weight and surcharge. In reality, the behaviour is highly non-linear, and Yang et al.(2004) used the finite element (FE) method to solve the non-linear equations. The consolidation of this fill is highly complicated, and a simple box-spring (Fig.2.13) model analogy is used to provide a better understanding of the consolidation process. This 22 model is also adopted from the dual porosity model developed for fissured clay. The inter-lump system is made up of clay lumps and inter-lump voids. The deformation of the inter-lump voids results mainly from the reduction of inter-lump voids due to rearrangement and distortion of lumps. In practice, self weight and self weight with surcharge case is actually more important than the case of surcharge only, mainly because in the field, often a significant length of time is needed to complete the dumping of dredged and excavated materials before a sand fill is placed on top of this lumpy clay fill. In the Pulau Tekong reclamation project in Singapore, which commenced in November 2000, an area of over 400 ha will be opened for disposal of such materials for at least five years before sand filling starts. During this period the consolidation will be due mainly to self weight and the end state of this process will define the initial condition when surcharge is applied. Fig.2.14 show the pore pressure profile for self weight only, and self weight plus surcharge applied after 2 days. From this figure, it is clearly seen, most of the inter-lump voids, especially at the lower level, would have closed up before the surcharge load applied on the lumpy fill. As there is little self-weight at the top, the closing of the inter-lump voids at the upper level is likely to be less. In this case surcharge will have a more significant effect, as seen by the large pore pressure dissipation for the case of self-weight and surcharge. Still the effect of softening and preconsolidation pressure of the lumps on the consolidation behaviour of lumpy fill is not known, and these two parameters are likely to have a significant effect on the lumpy fill behaviour. 23 2.5 Concluding Remarks When dredged clay lumps are used for land reclamation, the lumpy fill consists of large initial inter-lump voids. Usually sand is used as surcharge to accelerate settlement. Due to consolidation and compression from the sand surcharge, the void ratio and strength profile of lumpy fill change continuously. Continuous monitoring of these changes in void ratio and strength profile is important to facilitate the assessment of the state of the ground at any time to provide a more realistic understanding of the physical processes involved in the consolidation behavior of lumpy fill. The literature review has shown that in the study of consolidation of a lumpy fill, there have been a number of field studies, accompanied by a more limited series of laboratory studies including one centrifuge study. What these studies have clearly shown is the presence of two systems of voids in a lumpy fill, most of the voids in between the clay lumps close up during first stage of loading. But, it is not clear at what pressure the inter-lump voids were closed. This is one of the main objectives in the present project. So far, the consolidation behavior of lumpy fill is analyzed by using void ratio – consolidation pressure concept. The compressibility and strength characteristics of reconstituted clays are used as a basic frame of reference for interpreting the corresponding characteristics of lumpy fill layer. The properties of natural clay differ from its intrinsic properties due to the influence of soil structure (fabric and bonding). Thus intrinsic properties provide a frame of reference for assessing the in-situ state of the clay lumps. In the present study, a new normalizing parameter called the Void index (Iv), modified from that first introduced by Burland (1990), was used to aid in describing the 24 compression characteristics and the ultimate state of a lumpy fill layer. From above review, it was observed that the structural mechanisms helping to change the soil state from dual porosity state (inter and intra lumps voids) to homogeneous soil state. Also there is no study of the different stages of compression in the lumpy fill. Hence, in the present study, attention is given to the different stages of consolidation curve of lumpy fills. 25 Fig.2.1. Compressibility Curves in incremental tests (after Mendoza and Hartlen, 1985). Fig.2.2 Pore pressure built up during the test (after Mendoza and Hartlen. 1985). 26 Fig.2.3. Variation of surface settlement with time at different load steps (after Leung et al., 2001). 27 Fig. 2.4. Effect of lump shape on settlement of lumpy fill (after Leung et al., 2001). Fig.2.5. Excess pore pressure isochrones for lumpy fill made up of irregular lumps (after Leung et al., 2001). 28 Fig.2.6 Singapore marine clay lump of 205mm diameter before and after three dimensional swelling(after Robinson et.al(2004)). Fig.2.7 variation of water after three dimensional swelling 205mm diameter singapore marine clay lump(after Robinson et.al(2004)). 29 Fig.2.8. The use of void index Iv to normalize intrinsic compression curve Fig.2.9 Cluster model for permeability prediction (after Olsen,1962) 30 Fig. 2.10 Assumed relationship between the total, cluster, and intercluster void ratios (after Olsen, 1962) 31 Fig.2.11. Hydraulic conductivity discrepancies according to the cluster model (after Olsen,1962) Fig. 2.12 Shear strengths measured by vane test(x) and unconfined compression test (o) (after Hartlen and Ingers, 1981). 32 Stress σ Piston Vessel “Water in intralump voids” “Soil skeleton” “Water in inter- “Lump structure” lump voids” Fig.2.13 Spring - Box analogy (after Yang et al., 2002). Fig.2.14 Pore pressure profile for self weight only, and self weight plus surcharge after two days (after Yang et al., 2002). 33 CHAPTER 3 Experimental Investigation 3.1 Introduction The use of stiff dredged clay as fill in a reclamation may lead to possible problems as the fill contains large voids between clay lumps. These voids will close up under surcharge and cause large settlement. More importantly, if these voids do not close completely, there is concern about the long term residual settlement. The magnitude of consolidation pressure required to close the inter-lump voids is one key parameter still not properly established. The equilibrium state of a lumpy fill, under different consolidation pressures, is also not known. Swelling of clay lumps and several other factors like packing of lumps, size of lumps, inter lump fluid, etc. are known to influence the behavior of the lumpy fill, and their impact on the ultimate state needs to be ascertained. For subsequent engineering design purpose, the engineering properties of the fill after stabilization also need to be established. An understanding of the behavior of lumpy fill and the influence of these factors allows for significant opportunity to optimize the design process. In order to address the above aspects, experiments were conducted in the present study on lumpy fills whose inter-lump voids are filled with water and slurry. Slurry is considered because from the experience in the field thus far, the voids are often filled with it. The effects of size of lumps and the arrangement of lumps in the fill were studied 34 under controlled conditions. The role of swelling of clay lumps, when placed under water, on the compressibility behaviour is also evaluated. Also, the effect of preconsolidation pressure on the behaviour of the lumpy fill was investigated by using clay lumps formed from reconstituting clay slurry to two different preconsolidation pressures, namely 100kPa and 350kPa. The closing of inter-lump voids is especially difficult to establish in the laboratory. In the present study this is evaluated by using a combination of visual observations, monitoring void ratio changes, using the void index concept and from permeability measurements. By using more tell-tale parameters, a better assessment can be made. Secondary compression of the lumpy fill was also studied. Also, the shear strength profiles are evaluated using a miniature cone penetrometer as this is another tell-tale sign of the presence of small voids still not closed up. The details of the properties of the soils used for the present work, the experimental set up, sample preparation and experimental procedure are given in the following sections. 3.2 Composition and basic soil properties The dredged soil used in the present study, was taken from the Pulau Tekong reclamation site at a depth of 14m below the seabed. A typical clay lump dredged from seabed at this depth is shown in Fig.3.1. Tests for measuring specific gravity, Atterberg limits, and initial water content(before testing) for Lower Marine Clay was carried out according to British Standard 1377(British Standard Institution, 1990) and the results are reported in Table.3.1. Based on the test results, the Lower Marine Clay can be classified 35 as clay of very high plasticity (CV) according to the British Soil Classification System (BS5930). 3.3 Swelling test on clay lumps In the field, clay lumps are subjected to different degree of swelling. When they are placed back in water during the land reclamation process. To examine this aspect a series of experiments was carried out in which both undisturbed and reconstituted samples were used. These clays were cut into cubical shape of different sizes using a thin wire cutter. After cutting to the required size, each lump was submerged individually in water as shown in Fig.3.2. The sample was submerged in water and allowed to swell. The weight of each specimen was measured at an interval of 5 minutes for a period of an hour initially. Subsequent readings were taken with increasing time intervals till the weight of a sample becomes stable. The degree of swelling, Us, is defined as follows:  w − wi   ×100  wf − w i  Us =  (3.1) Where, w = moisture content or weight of the specimens after immersing in water at any instant of time wi = initial moisture content or weight of the specimen wf = final moisture content or weight of the fully swollen specimen 36 3.4. Small scale One-Dimensional Compression Test 3.4.1 Experimental Setup A one-dimensional compression model test was conducted to evaluate the loadsettlement response and the movement of clay lumps. Fig 3.3 shows a photograph of the setup of a one-dimensional small scale compression test which consists of a loading frame and cylindrical Perspex container. The lever arm of the loading frame is supported by an adjustable screw thread which acts as a pivot with one end holding a fixed counterweight and other the applied dead weight. Containers with three different diameters were used depending on the size of lumps used in the test. In order to reduce the side wall friction, the inner wall was greased and covered with a thin sheet of polyethylene sheet. Khoo et al. (1994) reported that the substantial reduction in soil-wall friction could be achieved with this arrangement. A burette with stand is placed on the top of the loading frame, which is used for measuring the permeability of the lumpy fill under different loading pressure. The permeability of lumpy fill was measured at the end of each load with falling- head method by connecting a burette with the bottom. The base of the Perspex container was seated in such way the centre line of container and loading cap will coincide. By so doing, the applied loading through the loading plate would always remain vertical. Depending on size of lumps, containers of different sizes were used. The test details including the size and shape of lumps, number of lumps, height of fill, degree of swelling, diameter of containers and void ratios of the fill are given in Table.3.2 37 3.4.2 Preparation of clay lump specimens Dredging of seabed using large clamshell grab, of size from 10 m3 to 18 m3, can produce lumps of varying size and shape with volumes in excess of 1 m3(Fig.3.1). The lump was then cut into cubes at site and wrapped using plastic bags to prevent the loss of water content and transported back to the laboratory. In the present study, tests were carried out using both remolded and undisturbed samples. The detail procedure for preparation of the above two types of sample is given in below sections. 3.4.2.1 Preparing of undisturbed clay lumps The soil sample was first transported from the site and stored in plastic containers in a cool and shady place. Cling-film was used to wrap the big lumps and another layer of water retaining material was placed on the top to reduce the rate of water evaporation. Also, water was sprinkled onto the water retaining material regularly to maintain the moisture content of the undisturbed samples. A thin steel wire saw was then used to cut the undisturbed samples into the required size and shape as shown in Fig.3.4 and Fig 3.5. 3.4.2.2 Preparing of remoulded clay lumps Dredged soil material was found to differ in quality as they contain shells and other fragments. This is actually a true replication of site condition as there are hardly any natural soils which are truly homogenous in nature. Thus, our next study will involve the use of remoulded clay which will be uniform. To prepare the remolded lumps, clay fragments were mixed to a slurry with a water content of 1.5 times its liquid limit. A rectangular container of length 44.5cm by width 43.5cm was used to compress the slurry 38 into a clay cake. The required pre-consolidation pressure of 100 kPa was achieved by the use of hydraulic pressure. The consolidation settlement was monitored using LVDT (Linear Variable Displacement Transducer). The compression is terminated when consolidation is completed. The same procedure was adopted for making lumps with a pre-consolidation pressure of 350 kPa. The consolidated clay cake thus obtained was then cut into the required size using a thin steel wire saw similar to the case to undisturbed samples were prepared. 3.4.3 Experimental Procedure and Instrumentation The pore pressures developed at the centre of a lump (intra-lump) and in between the lumps (inter-lump) were measured by using miniature pore pressure transducers. The length and diameter of the pore pressure transducer are 11.4 mm and 6.4 mm, respectively. Prior to installing the pore pressure transducers, they were first placed in an ultra-sonic bath for two hours to remove all impurities that might exist inside the ceramic porous filter due to earlier usage. For the transducers to be embedded in the clay lumps, the pore pressure transducer was inserted in a pre-drilled hole in the lump which was subsequently sealed up with silicone seal. One pressure transducer is inserted inside the lump for measuring intra lump pressure and two pressure transducers are placed in between the lumps at a location of ¼ and ¾ of the initial height of fill for measuring the inter-lump pressure. Perspex containers were used for the one dimensional consolidation test to allow a visual observation of the deformation of the clay lumps. The inner side of the perspex container that was in contact with the clay lumps was first smeared with a thin layer of silicon grease to reduce the side friction of the walls. At the bottom of the 39 container, a thin layer of coarse sand was sandwiched between two pieces of geo-textiles. This will act as a drainage system during testing. The prepared clay lumps were then placed in the perspex container in a regular fashion. It will be followed by a geo-textile and a loading plate which serve as a top drainage system. The transducer cables were carefully retrieved through pre-drilled holes in the loading plate and connected to an Autonomous Data Acquisition Unit (ADU). The whole lumpy fill system was then immersed in water to allow the clay lumps to soften to the required degree of swelling, based on the time established in earlier experiments. Load was applied onto the loading plate by using appropriate weights and the settlement of the lumpy fill system was monitored by using the LVDT. After the sample reached a equilibrium state, increase the load as double with previous load and the clay lumps were the loaded in steps to a maximum of 400 kPa. The surcharge pressure was capped at 400 kPa due to the constraint of the machine. The final stage involved the unloading and swelling of the clay lumps. Falling head permeability tests were conducted at the end of each load step. Data was collected at the end of each loading stage and time-settlement graphs could be plotted for each surcharge pressure. In the case of test conducted with 50mm size cubical lumps with 100% degree of swelling, a miniature cone of 10mm diameter with a tip angle of 600 was used at a rate of penetration of 5mm/s to measure the shear strength profile of lumpy fill. The cone penetration test was conducted at the end of application of pressure of 50kPa, 100kPa, 200kPa and 360kPa. 40 3.4.4. Unconsolidated undrained (UU) Triaxial test and Vane shear test Unconsolidated undrained (UU) triaxial test were conducted on the soil sample retrieved at the end of the application of pressure using the open drive samplers. The test specimen is prepared by pushing a mould into the sample. A rubber membrane is placed around the specimen and a confining pressure is applied to the sample. During the axial loading of the specimen, the confining pressure is kept constant and drainage of pore water is not allowed. The test is strain controlled at the rate of 2.5mm per minute. The height to diameter ratio of the sample is about 2, and the test is performed to failure or constant reading, whichever occurs first. The test results are presented in the form of deviator stress versus axial strain. The undrained shear strength defined as half the ultimate deviator stress can be obtained from these curves. Vane shear testing is one of the most common in-situ methods for the estimation of the undrained shear strength of the soil. Vane shear test is a useful method of measuring the shear strength of clay. It is a relatively cheap and quick method. The test gives the undrained strength of the soil. The vane shear apparatus used consisted of a four-bladed cruciform vane of 12.7mm wide and 12.7mm long mounted on a rotatable shaft. A torque is applied to the vane via a device connected to the shaft. The soil sample is held directly beneath the vane. The vane is then lowered and pushed into the sample to the desired depth and the test is started. The torque is then applied by rotating a monotorised torsion head until there is slipping of the soil. The peak strength of the sample is recorded via the maximum torque applied and the angle of rotation of the vane at the instant of failure. The vane test is then repeated for more locations along the depth of the sample. 41 3.5 Large scale One-Dimensional Compression Test From the results of tests on small size lumps only, it would be difficult to understand the behaviour of such lumpy fill in the field. Laboratory modeling (small scale) of reclamation process using small clay lumps could provide useful insights into the mechanics but has to deal with scale effects and heterogeneity of the problem. This is due to the faster rate of softening of small clay lumps used in the laboratory experiments, and also contact stresses generated on small sized lumps, for a given surcharge are higher than in the field (Tan, 2002). Therefore, large scale one dimensional consolidation tests using 200mm cubical shape clay lumps in a tank of size 1.4 m length, 1.0 width and 1.5 heights were carried. A loading frame with 500KN capacity is used to apply a maximum consolidation pressure of up to 350 kPa. The maximum size of the lump is 200 mm which will weigh about 13 kg each. This size is fixed as the maximum size, as handling anything bigger will cause significant logistic problems. The height of the fill will be about seven layers of lumps. The total number of lumps used in this test is about 150 and the estimated weight of soil involved is about 2 tones. The experimental setup for the 200 mm size cubical shape lumps was similar to that of the smaller size lumps as explained in the previous sections. In this test, a hydraulic type of compressor was used to apply the very large surcharge load needed over the fill instead of dead weight. The clay lump specimens of 200mm size cubical shape were prepared using the same procedure as in the small scale test. The details of the experimental procedure adopted are discussed in the following paragraph. 42 The schematic and photo view of the rectangular tank are shown in Fig.3.6 and 3.7. One side of the tank is made of Perspex and this allows visual inspection of the deformation of the cubical lumps. At the bottom of the tank, there is a drainage system to facilitate the dissipation of pore water pressure upon loading. The drainage system consists of a thin layer of coarse sand sandwiched between two pieces of geo-textiles. The prepared cubical lumps are placed randomly in the rectangular tank for six layers to a height of 1.3 m. The tank was then filled with water and the lumps are allowed to swell for a one month period so that the lumps reaching fully swelling condition. A geo-textile and a loading plate will then be placed on top and this also serves as a top drainage system. Fig3.8 shows a photograph of 200mm cubical lumps being placed in a rectangular tank. The loads applied were 50, 100 and 200 kPa. The cone penetration test was conducted at the end of each stage of application of pressure at 50kPa, 100kPa and 200kPa to obtain the variation of shear strength of soil due to the consolidation process. Also, soil samples were collected by using sampler for conducting oedometer test and water content test for observing the variation of soil properties along the depth of fill. While the consolidation process is interesting and of some importance, in the present work, the more important focus is on the end state after each load application. 43 Table.3.1 Physical properties of the soil used Property Singapore marine clay Liquid Limit,% 77 Plastic Limit, % 36 Plasticity Index, % 41 % Sand 5 % Silt 55 % Clay 40 Specific gravity 2.67 44 Table 3.2 Initial condition of the lumpy fill system for small scale consolidation experiments Test No Size of lump(mm) Diameter of container(mm) Number of lumps Height of fill(mm) 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 50 50 50 12.5 25 50 50 50 50 50 50 50 50 25(speri) 25 50(water) 50(slurry) 200 235 235 235 100 150 235 235 235 235 235 235 235 235 150 150 235 235 1.0m×1.0m 60 60 60 200 90 60 60 60 60 60 60 60 60 90 90 60 60 130 300 300 270 85 165 280 266 275 284 272 272 265 265 165 170 280 281 930 Interlump void ratio (eiv) 1.69 1.75 1.83 1.56 1.75 1.83 1.41 1.68 2.25 1.84 1.85 1.71 1.74 1.54 1.73 1.83 1.83 1.20 Initial system Void ratio(es) 3.16 2.96 3.14 2.86 3.11 3.14 2.99 3.14 3.56 3.15 3.05 2.96 2.99 2.89 2.99 3.26 3.15 2.47 Degree of swelling (%) 0 50 100 100 100 100 100 100 100 0 0 100 100 100 100 100 Pre-consolidation pressure, kPa 200 200 200 200 200 200 200 200 200 100 350 100 350 200 200 200 200 200 44 1.50 m Fig.3.1. Typical clay lump of Singapore marine clay obtained at a depth 14 m below seabed. Fig.3.2 Clay lumps undergoing swelling process 45 LVDT Burette Loading frame Geotextile filter 200mm Clay lumps Sand drain Fig.3.3. Typical experimental set up of lumpy fill system 46 Fig.3.4. Preparation of cubical lumps by using a wire saw Fig.3.5 25mm cubical lumps used in experiments 47 Fig.3.6 Schematic view of rectangular Tank of length 1.4m, width of 1.0 m and height of 1.5m 48 Fig.3.7 Photo view of rectangular Tank of length 1.4m, width of 1.0 m and height of 1.5m Fig.3.8 Photographic view of rectangular tank filled with 200 mm size cubical lumps 49 CHAPTER 4 Results and Discussion - I 4.1 Introduction The experimental procedure and instrumentation for one-dimensional consolidation tests on lumpy fills were discussed in the previous chapter. In this chapter, the results of swelling test and one dimensional consolidation tests performed on small and large size clay lumps are discussed. In order to understand the effect of size, shape, softening, pre-consolidation pressure and initial inter – lump void ratio on the consolidation behavoiur of a lumpy fill, a series of laboratory tests using the clay lumps prepared from remoulded and undisturbed samples with different initial conditions were carried out. In the first part of this chapter, the effect of clay softening in water with time was examined. A study is carried out to investigate this effect and the results are presented to provide some insights of this softening phenomenon. In the second part of this chapter, the results of small scale one dimensional tests results are analyzed using different approaches with various parameters. In the final section of this chapter, a few tests were conducted on large size lumps (200mm cubical) to understand the real site behaviour and the results were presented. . 50 4.2 Softening of Clay Lumps As a result of the removal of overburden stress during dredging operation, suction equivalent to the mean overburden pressure would develop in the lumps. When placed back in water during the filling process, these lumps would swell due to the suction in the soil. Subsequently, the suction reduces with time and the water content increases resulting in a decrease in the strength of the clay lumps. (Robinson et al. 2004). However, as the lumps in the field could be very large and of different sizes in some cases, they will take a much longer time to swell fully due to the low permeability of clay. If the time between placing the lumps in water and the application of surcharge is short, the lumps will be in a partially swollen state. Before conducting the test, to study the effect of swelling on consolidation behaviour of a lumpy fill, a relationship between degree of swelling, Us and time (t) was established by submerging the lumps in the water and allowing them to swell. The weight of the lump was recorded at regular time intervals, till it is nearly constant in weight. The rate and extent of swelling was studied for undisturbed cubical samples with sizes of 25mm and 50mm, and reconstituted cubical lumps of 25mm with pre-consolidation pressure of 100kPa and 350kPa. Tests were conducted on clay lumps with initial moisture content of 35 % and 60%, respectively. These tests were conducted primarily to investigate the softening effect of water on miniature clay lumps of different sizes and pre-consolidation pressure. Generally the water content of lumps increased with time. The water content of a lump, regardless of size and pre-consolidation pressure, increases very rapidly at the beginning of the test and slows considerably after some time. Fig. 4.1 shows the time versus degree of swelling plots for different sizes and initial water contents of the lumps. From Fig. 4.1, 51 it only requires 20 minutes for the 25 mm lump with wi=50% to reach Us = 50%, but it takes 11 hours for 50 mm lumps to reach the same degree of swelling, which amount to reaching a moisture content of about 66% for both sizes. The initial rate of swelling for a 25mm lump is much faster than for a 50mm lump. The final moisture content for both lumps is about 74%. These results suggest that the water content of the clay lumps after it has fully softened approaches that of its liquid limit. From Fig.4.1, it was observed also that the rate of increase in moisture content was higher for the lumps with lower initial water content as compared to the lumps with higher initial water content. Fig. 4.2 shows the time versus degree of swelling plots for lumps of 25mm cubical shape consolidated different pre- consolidation pressure. In Fig. 4.2, the rate of swelling is more pronounced for the case with P′c = 350 kPa than that for P′c=100kPa, as expected due to the higher hydraulic gradient as a result of greater suction. Another point to note is the fact that P′c = 350kPa has a lower initial water content of 44% as compared to P′c=100kPa with a value of 50% but they all swell to about the same final water content. Upon submergence in water, for the case with P′c=350kPa, the lumps start to absorb water rapidly. The time required to reach 50% and 100% degree for swelling for different size, initial water content and pre-consolidation pressure are summarized in Table 4.1. Due to the high rate of absorption, swelling at the outer edge of a lump was relatively fast than the inner layer. The differential expansion of the clay would result in the generation of internal tensile stress; since expansion of the outer layer was being restrained by the inner layer. This process may cause the cracking and breaking up of the clay lump itself as shown in Fig. 4.3. 52 4.3 Experimental studies on small scale size lumps As shown schematically in Fig.4.4, a lumpy fill layer of about 8m thickness was constructed using big clay lumps. On top of this, 10m thick sand layer, which acts as a surcharge to accelerate the consolidation of the clay lumps, was placed. In the field NDCPT was extensively used to evaluate the in-situ state of the lumpy fill formed by the dredged clay lumps (Karthikeyan et.al, 2004). But for the laboratory studies, such tests are difficult to carry out and a new apparatus was developed to be able to be handle bigger size specimens and the larger magnitude of the settlement expected during consolidation (explained in detail in Chapter-3). Fig.4.5 shows the typical experimental set of one dimensional compression test apparatus for consolidation test of a lumpy fill as shown in Fig. 4.4. Fig.4.6 shows the typical e-log σ′v and k-log σ′v changes for a test made of 25mm cubical lumps with 185mm initial thickness of lumpy fill. From this figure it is very difficult to study how the inter and intra lump voids are closing and at what consolidation pressure the inter-lump voids reduced to the size of the intra lump voids. To overcome the above problem, multi parameters including comparison with intrinsic and intact compression curves, normalizing parameter (void index), visual observations and strength profile evaluated using a miniature cone penetrometer were used to estimate the closing of inter-lump voids. These methods of interpretation of the one dimensional consolidation results were discussed in the subsequent sections. 4.3.1 Void Index The compressibility and strength characteristics of reconstituted clays are used as a basic frame of reference for interpreting the corresponding characteristics of the intact 53 natural material. Burland (1990) introduced a new normalizing parameter called the void index to aid in correlating the compression characteristics of various undisturbed soil samples. The one-dimensional compression curves of reconstituted clay may be normalized by assigning fixed values to e*100 and e*1000 and the parameter void index, Iv is defined as * * e − e100 e − e100 IV = * = * e100 − e1000 C c* (4.1) where e is the void ratio, e*100 and e*1000 are the intrinsic void ratios corresponding to σ′v = 100kPa and 1000kPa. C*c is the intrinsic compression index which is also equivalent to the change in void ratio on the ICL for an increase in stress from 100kPa to 1000kPa (i.e. e*100 –e*1000). Burland (1990) showed that the intrinsic compression curve from soils with a wide range of liquid limits, when plotted in terms of void index, could be represented with sufficient accuracy by the following equation: Iv = 2.45-1.285X +0.015X3 (4.2) where X = logσ’v in kPa. Burland (1990) termed this unique line in Iv-logσ’v space as the Intrinsic Compression Line (ICL). The ICL can be easily obtained experimentally. Alternatively, Burland (1990) showed that e*100 and Cc could be reasonably related to the void ratio at the liquid limit, eL, as follows: e*100 = 0.109 + 0.679eL - 0.089eL2 + 0.016eL3 (4.3) Cc* = 0.256eL – 0.04 (4.4) Karthikeyan (2005) has already demonstrated that the ICL and void index from oedometer tests can be used in the study of the compression characteristics of the ultimate state of a reclaimed land. Therefore, this framework will be used to study the 54 compression characteristics of the lumpy fill layer for different initial conditions. The definition of void index in equation (4.1) is not directly applicable to the lumpy fill system, and has to be modified. The compression curves plotted in Fig. 4.7 represent the intrinsic compression curves for Singapore Lower Marine Clay obtained from Pulau Tekong. The quantities e*30 and e*800 are the intrinsic void ratios corresponding to σ′v=30kPa and 800kPa respectively. The in-situ void ratio of the given clay varied between 1.68 to 1.76, with an average of about 1.72. The average in-situ void ratio of the clay sample is also shown in Fig. 4.7. During the early stage of loading, the reconstituted clay sample experience large contact stress. The void ratio of intrinsic compression line is reduced due to the expulsion of water from the reconstituted soil. At pressure less than about 30kPa, the void ratio on the intrinsic compression line is more than that of the in-situ void ratio of the lumpy clay. It is assumed that the reconstituted clay reached the in-situ soil state when the void ratio of the intrinsic compression curve reaches the value of the in-situ void ratio of the lumpy clay. So, the e*30 value is selected in such a way that the consolidation pressure required to reach the void ratio on the average intrinsic compression line that is about the same as the in-situ void ratio of the given soil that is being tested and this pressure is taken as a reference pressure for selecting the intrinsic void ratio e*30. Similarly, the final consolidation pressure used in the one dimensional compression test on clay reconstituted in the laboratory with a water content of 1.5 times liquid limit is taken as the reference pressure for selecting the intrinsic void ratio e*800. The modified void index is defined as * e − e30 * * Void Index, = e30 − e800 = * e − e30 * *   e30 − e800  1.42  log10 30 − log10 800  55 Iv = * e − e30 1.42C c* (4.5) where e is the void ratio of ICL. The compression data in Fig. 4.7 are replotted in Fig. 4.8 by replacing void ratio, e with void index, Iv. When plotted in terms of Iv, the data fall on a unique ICL curve, as reported by Burland (1990). The ICL proposed by Burland (1990) as given in Eq.(4.2) is also plotted in Fig.4.8 and it can be seen that there is significant agreement with modified void index, with little scatter. This ICL, obtained from reconstituted samples, can be used to aid in correlating the compression characteristics of lumpy fill. Similarly, the e-log′σv curve for lumpy fill may be normalized by assigning fixed values to e*30 and e*800. The normalizing parameter chosen is defined as the void index Iv such that * e s −e30 * * Void Index, = e30 − e800 = Iv = * e s − e30 * *   e30 − e800   1.42 log 30 − log 800 10  10  * es − e30 1.42C c* (4.6) VT − VS , VT= Total volume and VS =Volume of solids. VS Thus the curve in Fig. 4.9 (a) may be transformed to the normalized curve in Fig. 4.9 (b) Where system void ratio, es = where the void index Iv, as defined by equation (4.6), is the ordinate. When es = e*30 ( ≅ eo), Iv = 0 and when es = e*800, Iv = -1. The void index may be thought of as a measure of the closing of the inter-lump voids present in the lumpy fill system. When Iv is greater than zero, this means that there is still inter lump voids present and when Iv is equal or less than zero, this means the inter-lump voids are closed. In other words, once load is 56 applied on the lumpy fill, the inter lump voids will start to close and after some time, the system void ratio, es , will approach the e*30 (equal to the in-situ void ratio), that is most of the inter-lump voids would be closed (i.e. void index value approaches zero). Subsequently, there is a further reduction in the void ratio of the lumpy fill due to the compression of the lump itself. The above normalizing parameter called the ‘void index’ is used to aid in correlating the compression characteristics of lumpy fill made of various size and shape of lumps, different degree of swelling, pre-consolidation pressure of lumps and packing arrangements. 4.3.2 Lump Size and Shape Typical state of the clay lumps obtained from lower marine clay immediately after dredging is shown in Fig.4.10. The dredging of these materials produces stiff clay lumps of varying size and shapes but in general, these lumps are usually quite big (1m3 to 8m3 in volume) as shown in Fig.4.10. At the same time, these dredged clay lumps when placed in the reclamation pond, will undergo three-dimensional swelling due to the high suction developed when dredged and the subsequent availability of the seawater all around the lumps (Robinson et al. 2004). So, to investigate the effect of size of clay lumps on consolidation behaviour of lumpy fill, one dimensional consolidation tests were carried out on lumpy fill made up of 12.5, 25 and 50 mm cubical clay lumps with interlumps voids filled with water. All the lumps were fully swollen before apply the load. The e-logσ′v compression curves obtained from one-dimensional consolidation of lumpy fill made of 12.5mm, 25mm and 57 50mm size cubical lumps are shown in Fig.4.11. The lumpy fills were prepared to approximately the same initial inter-lump void ratio (eer=1.75 ± 0.15). Inter lump void ratio, eer = VT − VL , using this equation changing total volume and no of lumps VL (Volume of lumps) to get same initial inter lumps voids for different size of lumps (by trial). What is very significant in this result is that the compression curves are practically the same in all three cases. This suggests that for the size of the lumps evaluated here, the size has little influence on the consolidation behaviour for the lumps pre swell to 100 % before apply load. The compressions data in Fig. 4.11 are replotted in terms of void index and compared with Intrinsic Compression Line (ICL) are shown in Fig. 4.12. The two mechanical properties (e*30 and C*c) were determined from one-dimensional compression tests on reconstituted soil samples. These parameters were used to compute void index using Eq.(4.6). The changing parameter in the void index term is the system void ratio, es. This figure shows that, at the initial state of loading stage the intact compression curves lies further away from ICL but approaches the ICL as loading increases. During early stage of loading, the lump may deform and fill the inter-lump voids without significant volume change and the volume reduction of the lumpy fill may come mainly from the closing of inter-lump voids. The inter-lump voids may be considered closed when the system void ratio is equal to in-situ void ratio or e30 (i.e. Iv approach to zero value). When Iv is greater than zero, this means that there is still inter-lump voids present and when Iv is equal or less than zero, this means the inter-lump voids are closed. The lumpy fill reaches this state for all three sizes of lumps under a consolidation pressure of about 55 to 65 kPa. Also permeability under different consolidation pressure 58 was measured before the next load increment and the results are shown in Fig.4.13. From this figure, it was observed that the change in permeability is insignificant at initial stage of loading perhaps a reflection that the permeability tests cannot measure as accurately large permeability or that the load is too small to affect the overall system. After a few load increment the permeability begins to decrease suggesting that the closing of the inter lump voids begins to affect permeability. Towards the end of loading again there is little change in permeability in spite of significant increase in loading, suggesting that most of the inter-lump voids are closed already. The small change in permeability at this stage is probably due to reduction in intra lump voids itself. The intersection of two tangents was drawn in Fig.4.13 and provides another indicator of the pressure at which the most of the inter lump voids were closed due to the surcharge load applied on the lumpy fill. The results deduced from the void index concept are compared with those obtained from the permeability measurements. The variation of consolidation pressure required to close the inter-lump voids for different sizes of the lumps estimated from the void-index approach and permeability measurements are shown in Fig. 4.14. The estimated consolidation pressures obtained from the permeability– logσ′ curve for closing of inter-lump voids agree well with the results from Iv – log σ′v curve. From this figure, it can also be observed that almost all the points lying within a narrow band and is clearly indicating that there is not much difference in the surcharge pressure needed to close the inter lumps voids for the above three sizes of lumps. This suggests that for the size of the lumps evaluated here, the size has no influence on the consolidation behaviour for the fully swollen case. The dredged lumps deposited on to the seabed are of different shapes and the 59 shape of lumps is closely related to the initial packing density of the lump arrangement. For the present study, the effect of shape will be investigated using cubical and spherical lumps. Among these two shapes of lumps, using the spherical lumps gives the densest arrangement compared to the cubical shape lumps. The Iv – log σ′v and k- log σ′v plots of lumpy fills made of 25mm cubical and spherical lumps are shown in Fig. 4.15 and Fig. 4.16. It is noted that the consolidation pressure required to close the inter-lump voids for cubical and spherical shape lumps are 45 and 35 kPa respectively. The results illustrate that shape of the lumps has a small influence on the consolidation characteristics of the fill, though there is also a small difference in the initial system void ratio. The difference in initial system voids ratio is due to the difficulty in making the spherical lumps due to irregular in shapes. For further studies, cubical shape of lumps were used to investigate the influence of the other factors on consolidation behaviour of the lumpy fill, since it is more readily prepared from dredged soil slumps. 4.3.3 Degree of Swelling of Clay Lumps The removal of overburden stress during the dredging operations led to a suction equivalent to the mean overburden pressure developing in the clay lumps. When these clay lumps are dumped in seawater during the land reclamation process, they will absorb water because of the suction present in the soil. Subsequently, the suction reduces with time and the water content increases resulting in a decrease in the strength of the clay lumps (Robinson et al. 2004). The variation of water content with the distance from the centre of clay lumps is shown in Fig.4.17. This figure shows that the measured water content for the partially softened ( Us=50%) clay lumps is not uniform along the width of 60 the lumps. The water content at the edge of lump is higher than that at the centre, a fact of some practical relevance. During the consolidation process of the clay lumps in a lumpy fill under sand surcharge, the edge of the clay lumps experience very high contact stress and will cause partial disintegration of edges of the bigger clay lumps while the centre of these lumps remains close to its original consolidated state. However, the effect of these variations of moisture content along the depth of lump and degree of swelling of clay lumps on consolidation behaviour of lumpy fill is not established to date. The influence of the above factors on consolidation characteristics was studied and discussed below. For study the effect of swelling on consolidation behaviour of lumpy fill, here assumed that the lumps were do not softened further upon the loading. The e-log σv′ plots of the lumpy fills subjected to different degrees of swelling are shown in Fig.4.18. The effect of swelling is clearly evident. Under small consolidation pressures, the void ratio of the fill made of lumps with Us=0% is considerably higher than those corresponding to the lumps subjected to swelling. When the clay lumps are subjected to swelling, the water content increases with a consequent reduction in shear strength and stiffness. Therefore the lumps cannot withstand the high contact stresses and slip occurs at the contact points to form a stable configuration leading to reduction in inter-lump voids. In contrast, the lumps which were not subjected to swelling are much stronger and are able to withstand a much higher pressure without plastic deformation and inter-lump slippage. This is clearly seen in Fig. 4.19a, b and c, which shows the photographs of the lumpy fills with different degrees of swelling subjected to a loading of 50kPa. Under the same consolidation pressure of 50 kPa, the inter-lump voids of the 61 fully swollen lumps are visibly closed but large inter-lump voids can be seen for the fill in which the lumps were not allowed to swell. The in-situ void ratio (e0) of the lumps used for making the lumpy fill is 1.59 and in its fully swollen state, the void ratio (es) of an individual 50 mm lump is 1.83. The void ratio (e50) corresponding to Us=50% is 1.75. These values are marked also in Fig. 4.18. During the early stages of loading, the lumps experience large contact stresses. It is likely that the individual lumps cannot consolidate significantly until the mean pressure on the individual lumps exceeds the mean preconsolidation pressure of the single lump. Therefore, the lumps may deform plastically and fill the inter-lump voids without significant volume change of the lumps themselves and the volume reduction of the lumpy fill may come mainly from the closing of inter-lump voids. Therefore, when Us=0%, the inter-lump voids may be considered closed when the global void ratio is equal to the in-situ void ratio of 1.59. consolidation pressure of about 120 kPa. The lumpy fill reaches this state under a When fully swollen lumps are used, the pressure required to reduce the global void ratio to 1.802 is about 40kpa, which is considerably less than that required for the case Us=0%. Permeability measurements on the lumpy fill under a consolidation pressure of 50 kPa shows a sudden reduction in permeability as can be seen in Fig. 4.20 suggesting that the inter-lump voids are significantly closed. The e-log σv′ of the partially swollen lumps lie in between those of Us=0% and 100% but is more close to that of Us=100%. The e-log σv′ curve reaches e50 at a pressure of about 60 kPa where the inter-lump voids are substantially closed. A typical time-settlement curves obtained under 50 kPa loading pressure in tests of 50 mm cubical lumps with 0%, 50% and 100% degree of swelling are shown in Fig.4.21. The 62 settlement values are normalized with the final settlement for different degree swelling to facilitate comparison. From this figure it was observed that the rate of normalized settlement for fully softened lumps is much slower than that with lesser swelling. This is due to the fact the most of the inter-lump voids were closed before this loading pressure was applied and the shape of the curve is S-shaped curves and it indicates the soil is becoming homogenized. The compression curves obtained from LVDT, as shown in Fig.4.18, are plotted in terms of Iv and compared with the ICL and shown in Fig.4.22. The inter–lump voids may be considered closed when the void index, Iv reaches zero. The lumpy fill reaches this state under a consolidation pressure of about 90kPa for 0% degree of swelling. When the fully swollen lumps are used, the pressure required to reduce the voids index to zero is about 50kPa, which is considerably less. The Iv – log σ′v of the partially swollen (Us= 50%) lumps lie in between those of Us=0% and 100% but is closer to that of Us=100%. The estimated consolidation pressures obtained from the void index curves for closing of inter-lump voids agree well with the results from permeability measurements. Also, the variation of consolidation pressure required to close the interlump voids for lumps subjected to different degree of swelling obtained from Iv- log σ′v, k– log σ′v and e– log σ′v results are shown in Fig. 4.23. From this figure, it is clearly seen that there is a significant effect of degree of swelling of the clay lumps on the consolidation behaviour of the lumpy fill. The above results show that the degree of swelling has a direct influence on the pressure required to close the inter-lump voids. Lumps with low degree of swelling, Us will swell faster and hence more than lumps with high degree of swelling, Us. When the 63 lumps are allowed to swell fully, the effective stress within the lumps is reduced due to the loss of suction with a consequent reduction in the strength and stiffness. With weaker and more deformable lumps, the pressure required to close the inter-lump voids will be smaller, consistent with the test. Therefore it is important to consider the influence of swelling of lumps, when clay lumps are used. 4.3.4 Initial packing of lumps In section 4.3.2, the effect of size and shape were discussed and it was concluded that the size of the lumps has no direct influence on the consolidation behaviour of the lumpy fill. But the size and shape of the lumps will influence the arrangement and initial inter lump voids present during the dumping of dredged lump at the reclamation site by opening the bottom of the barge carrying them. A lumpy fill system can be simply classified into a macro system of clay lumps with inter-lumps voids as well as a micro system of individual clay lumps with intra-lumps voids. Thus, at low surcharge pressure, the compressibility behaviour of a lumpy fill system significantly depends on the volume of inter-lumps voids present. Thus, it is necessary to examine the effect of the volume of inter lump voids present in a fill on the consolidation behaviour of the lumpy fill. At the early stages of loading, the lumpy fill contains large inter-lump voids. The lumps are in contact with each other at a few contact points. With application of consolidation pressure, the contact stress will be much higher than the applied average vertical stress. The evaluation of contact stresses is very complicated and therefore the experimental results were analyzed using the average vertical stress (σ′v). Also, the lumpy fill consists of two systems of voids; one is the intra lump voids (void within the 64 lumps, the usual void in a soil) and the other is the inter-lump voids. The volume of inter-lump voids is known only before the application of the consolidation pressure, but subsequently, the compression process is the result of the combined effect of the volume changes in the two systems of voids and it is very difficult to separate them. As only the surface settlement can be measured, only the global void ratio can be calculated directly. This is used to analyze the data reported. Depending on the packing arrangement, the fill will have different initial void ratios (system), and for understanding the effect of packing arrangement on consolidation behaviour of lumpy fill, 50mm cubical lumps subjected to zero degree of swelling with three different inter-lump voids of 1.41,1.68 and 2.25 were used. Fig. 4.24a shows the state of a lumpy fill immediately after placing the lumps in water in case of 50 mm cubical lumps with inter-lump void ratio, eer of 1.68. At this stage there are large interlump voids are visible. As the surcharge pressure was increased, the size of the interlump voids decreased as seen in Fig, 4.24b to 24d. When the pressure reached 50 kPa, most of the inter-lump voids will be closed and the fill begins to look more like a homogenized clay. Leung et al. (2001) also observed that under a consolidation pressure of 50kPa, the outer part of the lumps squeezed into the inter-lump voids and effectively closed them up. A typical time-settlement curves obtained under different step-loadings in tests of 50mm cubical lumps for eer values of 2.25, 1.68 and 1.41 with fully softened lump are shown in Figs. 4.25a to c. The settlement values are normalized with the final settlement under each loading to facilitate comparison. Under small surcharge loadings, the rate of settlement is faster due to the presence of inter lump voids, which permits practically free 65 drainage of water during the consolidation process. As the consolidation pressure is increased, the rate of settlement decreases markedly. For consolidation pressure less than 50kPa, the shape of the time-compression curve in the logarithmic plot is concave, in marked contrast to those observed for homogenous clays, following the classic Terzaghi’s model. Beyond a consolidation pressure of 50kPa, the rate of settlement is much slower due to the closure of the inter-lump voids and the shape of the curve started changing. Typical S-shaped curves are obtained in the logarithmic plot for pressure beyond 100kPa. Also the photographs were taken under 50kPa load for all three cases of initial voids and these are shown in Fig.4.26 (a) to (c). From this visual observation revealed that for densely packed lumps (for eer=1.41), a small pressure is needed to close inter-lump voids compare to a more loosely packed lumps. In order to confirm the above observation, the e-log σ′v results were analyzed using normalized parameter void index. Fig. 4.27 shows the variation of void index with logarithmic pressure for three different initial inter-lump void ratios. In this series of tests with 50mm lumps, the influence of the initial packing arrangement of the fill is prominent in the early stages of loading. The consolidation pressure required to bring the void ratio of the fill to the in-situ void ratio for 50mm cubical lumps with eer values of 1.41, 1.68 and 2.25 are 38 kPa, 46 kPa and 60 kPa respectively. Also the variation of permeability with consolidation pressure is shown in Fig. 4.28. From this plot, it is noted that there is drastic reduction of permeability value for eer values of 1.41, 1.68 and 2.25 under a consolidation pressure of 42kPa, 53kPa and 62kPa respectively. Fig 4.29 shows the variation of consolidation pressure required to close the inter-lump voids for the above three different initial inter lump void cases obtained from void index and 66 permeability measurements. This figure illustrates that there is some agreement in the calculation of the pressure required to close inter lump voids from above two methods. The above results suggest that initial packing, with inter-lump voids ratio in the range of 1.41 to 2.25, for the cubical lumps of 50mm size, has a significant influence on the compressibility of the lumpy fill. 4.3.5 Pre-consolidation pressure of lumps In the present land reclamation works in Singapore, the dredged clay materials are taken from different depths below the seabed and thus were subjected to different preconsolidation pressures. A detailed investigation as carried out by Karthikeyan(2005) who found that there was some scatter in the pre-consolidation pressures and OCR of the seabed soil prior to carrying out dredging activities at the site is located close to Changi in the northeastern part of Singapore Island, as shown in Fig.4.30. The variations of preconsolidation pressure and OCR with depth are shown in Fig 4.31(a) and (b). As variation in pre-consolidation pressure is expected, that impose of this aspect on the characteristic of a lumpy fill needs to be evaluated. Clay with a higher pre-consolidation pressure is stiffer due to the lower moisture content. One concern is whether clay lumps with higher pre-consolidation pressure will require a greater loading pressure to close the inter-lump voids. The effect of pre-consolidation pressure on the consolidation behaviour of the lumpy fill was investigated by reconstituting clay slurry to two different preconsolidation pressures namely 100 kPa and 350 kPa. Figs. 4.32 shows the Iv – log σ′v curves of the lumpy fill made of 25mm cubical lumps with the two different pre-consoldiation pressure and subjected to 0% and 100% 67 swelling. From these curves, it was observed that for 0% degree of swelling, the lumpy fill made of lumps subjected to 350kPa pre-consolidation pressure take a considerably larger load to close the inter-lump voids than the case with lumps subjected to 100kPa. The consolidation pressure required for P′c=100kPa and 300kPa to close the inter-lump voids are 70 kPa and 100 kPa respectively. For P′c=100kPa (Us=0%), the intrinsic strength of the individual lump was much lower than P′c =350kPa (Us=0%), therefore it experienced a larger settlement per loading pressure at the initial stage. Individual clay lumps of lower preconsoldiation pressure may not have sufficient strength to withstand the stress applied and deform plastically by shearing or breaking up upon application of the load. For P′c=350kPa (Us=0%), the individual clay lumps had been preconsolidated to a much higher pressure resulting in a higher intrinsic strength. It was observed from the Iv – log σ′v plot that at a pressure of 70 kPa, most of the inter lump voids are closed for clay lump made of 100 kPa with 0% degree of swelling. But in the case of lump with pre-consolidation pressure of 350kPa, a 95kPa surcharge is required to close the interlump voids. Similarly for the case of Us =100%, the consolidation pressure required for closing the inter-lump voids for 50mm cube with 100kPa and 350kPa is 50kPa and 35kPa respectively. From Fig.4.32, it was observed that the effect of swelling is more predominant in the case where the lumps were subjected to a much higher preconsolidation pressure, due mainly to the high suction of the P′c=350kPa. The permeability results shown in Fig. 4.33 show the same pattern to the above Iv– log σ′v results. Also, the pressures required for closing the inter-lump voids obtained from void index concept and permeability measurements with different pre-consolidation pressure 68 and degree of swelling of a lump are shown in Fig. 4.34. From above results, it is noted that the pre-consolidation pressures of individual clay lumps play a significant role in the estimation of pressure required to close the inter-lump voids present in the lumpy fill. 4.3.6 Lumpy Fill in Water and Clay slurry Extensive site investigation of the seabed surrounding Pulau Tekong, reclamation site, showed that the seabed in general consists of four layers namely; surface soft marine clay (SSMC) of recent deposit, upper marine clay (UMC), intermediate layer (IML) and lower marine clay (LMC) followed by the residual soil or weathered granite. A detailed characterization of Singapore marine clay at this site is also given in Tan et al. (2002). A typical geological profile of the seabed soils at the reclamation site is shown in Fig.4.35. The SSMC is very soft with liquidity index higher than 1.0 and when dredged, exists in the form of slurry. The UMC is generally of soft to medium consistency. The IML is highly heterogeneous and often consists of sand and silt sized particles and LMC is generally stiff. After dredging, UMC and LMC exist in the form of lumps while the IMC often do not form lumps. Generally, the inter-lump voids are expected to be filled with water. After a series of ND-CPTs tests were carried by Karthikeyan, M (2004) and he confirm that the voids have not only been filled with sea-water, but filled with clay slurry. This situation arises when stiff clay lumps are dumped directly into the very soft marine clay (SSMC) as in the case of Pulau Tekong. A typical set of measured cone tip resistance and pore pressure response at the test site is shown in Fig.4.36. From this figure it was observed that cone resistance of this very soft clay is very much less than that of other layer of 69 seabed soil. Due to this low value, slurry will form during dredging instead of lumps. These surface soft marine clay dredged materials are then dumped into the reclamation site and as a result, the inter lump voids of the lumpy fill system are filled with slurry. A view of the surface soft marine clay and upper marine clay dredged from the seabed and placed in a barge before dumping in sea is shown in Fig.4.37. So, it is also necessary to investigate the effect of slurry present in the inter-lump voids on the consolidation behaviour of lumpy fill. To investigate the effect of water/slurry present in the inter-lump voids on the compressibility behaviour of the lumpy fill system, one – dimensional consolidation tests were conducted on lumpy fill made of 50 mm cubical lumps. In this way, any difference in the time-settlement, Iv – log σ′v and permeability response will be due solely to the materials in the inter – lump voids. Typical time-compression curves under different step loadings for 50mm cubical lumps with inter-lump voids filled with water or slurry(w=150%) are shown in Fig. 4.38. From this figure it was observed that where the inter-lump voids (ILV) are filled with water, the rate of settlement under small surcharge pressure is very fast compared with cases with the ILV filled with slurry. This is expected because of the fact that when the inter-lump voids are filled with water, it permits free flow of water during the consolidation process. The shape of the curve is also different from what is expected for homogeneous clay. The presence of the slurry in between the voids retards the flow of water. Consequently, the rate of settlement is slower. When the water content of the slurry is higher, the permeability is higher leading to faster rate of consolidation (Shirley (2003)). Also, typical S-shaped curves are obtained when the inter-lump voids are filled with slurry similar to that for usual soft clay. Under higher consolidation pressures, the 70 difference in rate of settlement decreases for the two cases due to the closing of most of the large inter-lump voids and the S-shaped curve is obtained even for the case where the inter lump voids are filled with water. The void index – consolidation pressure relationship obtained for 50mm cubical lumpy fill system with the inter-lump voids filled with water or slurry are plotted in Fig. 4.39. It is observed that the system with slurry of water content 150% reaches to zero void index earlier than the system with water. This could be because of the slurry present in the inter lump voids reduces the difference in permeability between the inter-lump and intra-lump system. For a lumpy fill system with inter-lump voids filled with slurry, the inter-lump voids are already partially filled with fine suspended materials and while more time may be needed by the system to dissipate the excess pore water pressure because it is much less permeable, it will close the voids at a lower effective pressure. Under consolidation pressure less than about 25 kPa, the permeability of the lumpy fill system with inter-lump voids filled with water is very much higher than when filled with slurry (Fig. 4.40). This is expected as the large inter-lump voids permit practically free flow of water. Beyond 25 kPa, drastic reduction in permeability was observed. As expected, the permeability of the system filled with slurry is significantly smaller compared to the system filled with water even under small consolidation pressure as the flow of slurry through the inter-lump voids will face greater resistance or drag due to the viscosity of the slurry. This indicates that the consolidation pressure required to close the inter lump voids is less for a system filled with slurry than water. However, the influence of slurry on the end state of lumpy fill is not very significant when compared to that with water, for consolidation pressure beyond about 100kPa. 71 4.3.7 Pore pressure response The discussions in the first three paragraphs of this section will applicable only if the inter lump voids filled with water only not slurry. For evaluate the closing of inter and intra-lump voids are to examine the variation of pore water pressure in the inter-lump and intra-lump voids. The variation of pore water pressure in the inter and intra lump voids for a lumpy fill made of 50mm cubical lumps with inter-lump voids filled with water are shown in Fig. 4.41 for load increment from 25 kPa to 50kPa and 100kPa to 200kPa. The variation of pore water pressure with time for the load increment from 25 kPa to 50 kPa is shown in Fig.4.41a. For homogeneous clay it is expected that immediately after the application of load the pore pressure increases to a value equal to the applied pressure increment. The pore pressure in between the lumps (inter-lump) increases after the application of the load but reaches a maximum value much less than the applied pressure increment and dissipates initially at a faster rate due to the presence of inter-lump voids which permit rapid flow of water. On the other hand, the pore pressure inside the lump reaches a maximum value higher than the applied pressure increment, behaviour akin to the Cryer-Mandel’s effect studied by Robinson et al. (2003). The pore pressures during the loading stage from 100 kPa to 200 kPa are shown in Fig.4.41b. Under this pressure increment, the pore pressure in the inter-lump voids and inside the lumps are close to each other and the maximum pressure is approximately equal to the applied pressure increment, very similar to the case of a homogeneous clay. This is due to the fact that the inter-lump voids, by now, are practically closed and the fill behaves more like a homogeneous clay. 72 The process of increase of pore water pressure inside the lump to a value more than the applied pressure increment can be explained using the spring-box analogy proposed by Yang et al. (2002). It consists of a perforated piston, supported by columns of perforated boxes of spring inter-connected by another set of springs (Fig.4.42). The springs connecting the perforated boxes represent the structure formed by the clay lumps, and the spring inside a box represents the soil skeleton within a lump as in the classical Terzaghi’s analogy. As in Terzaghi’s analogy, the spaces are filled with water. The perforation in each box of springs allows exchange of fluid between the inside and the outside of the box, wherever there is a pressure difference. When a stress, σ is applied on the piston, just as in the original Terzaghi model, the box-spring system cannot move initially and the water in the inter-lump voids and intra-lump voids has to carry the full increase of load. During the consolidation process the pore pressure in the inter-lump voids dissipates at a much faster rate compared to that in the intra-lump. During the process, the load carried by the inter-lump water will be transferred to the lump structure and consequently to the lump. This additional load on the lump will be initially carried by the pore water inside the lump leading to a pressure more than the applied stress. Since the slurry decreased the permeability of the inter lump system dramatically, the whole process of consolidation was not as fast as that in the test where inter-lump voids are filled with water initially as discussed in the above paragraphs. The variation of pore water pressure in the inter-lump and intra-lump voids with inter-lump voids filled with slurry is shown in Fig. 4.43. From this figure it was noted that the different in pore pressure between inter-lump and intra lump voids are less than if inter-lump voids are 73 filled with water (Fig.4.41). As explained earlier, in the case of the lumpy fill system with inter-lump voids filled with slurry, the inter-lump voids are partially filled with fine suspended materials and thus more time is needed by the system to dissipate the excess pore water pressure. From the above results, the difference in pressure between inter and intra lumps voids during loading is less significant in the case where inter-lump voids are filled with slurry, compared to that filled with water. 4.3.8 Shear strength profiles of the lumpy fill The above results suggest that for a lumpy fill made up of undisturbed clay lumps with 100 % swelling, the surcharge pressure required for the substantial closing of interlump voids is about 40-50 kPa but a pressure more than the calculated value is needed to bring down the void ratio equal to the in-situ void ratio. So, it is important to evaluate the strength profiles of the lumpy fill upon stabilization by surcharge loading of different magnitude. The undrained shear strength was estimated from CPT results from the following equation suggested by Lunne et al. (1997), su = qt − σ ' v N kt (4.7) where qt is the corrected cone resistance, su is the undrained strength Nkt is the cone factor and σ′v is the effective vertical consolidation pressure. The cone factor used in the present investigation was established by comparing the vane shear test conducted on the same soil sample with the CPT results and it showed that the range of values of the cone factor, Nkt for the lumpy fill layer was between 8 to 10. Studies carried out by Dobbie 74 (1988) in Singapore showed that Nkt ranged from 9 to 12. Also, Karthikeyan et al.(2004) obtained the Nkt values for lumpy fill layer in the range between 8 to 12. Therefore for the present study, the undrained shear strength was estimated by using a cone factor of 9. The shear strength profiles obtained using cone penetration tests under surcharge pressures of 50 kPa, 100 kPa, 200 kPa and 400 kPa are shown in Fig. 4.44a to d. A line representing the estimated undrained shear strength of normally consolidated Singapore marine clay using Su = 0.23σ′v as suggested by Tanaka et al. (2001) and for over consolidated soil, Su= 0.23σ′v(OCR)0.75 are also shown in Fig. 4.44. It is interesting to note that at consolidation pressures of 50 kPa and 100 kPa, the shear strength values from the CPT are lying in between the normally consolidated state and overconsolidated state. This is consistent with the findings of Karthikeyan et al. (2004) who reported the presence of zones of normally consolidated and overconsolidated states in the field, based on an extensive site investigation on a land reclaimed using dredged clay lumps. When the consolidation pressure is equal to the pre-consolidation pressure (Fig. 4.44c), the strength profile is almost uniform. The comparison reveals that most parts of the layer have strength similar to that of clay in a normally consolidated state. This suggests that to make the fill uniform, consolidation pressure equal to or greater than the preconsolidation pressure (Fig. 4.44d) of the lump should be applied to the fill. 4.3.9 Secondary compression In the lumpy fill system, the presence of any unclosed inter-lump voids may show large secondary compression due to creep at the lump contact points. Therefore, the secondary compression of the lumpy fill needs to be examined. The rate of secondary 75 compression is controlled by the rate at which the structure can deform. The relationship between void ratio and the logarithm of time during secondary compression is usually linear for the most soils over the time ranges of interest after completion of primary consolidation. A key parameter used for quantitative analysis is the coefficient of secondary compression Cα which is the slope of the void ratio-logarithmic time plot in the secondary compression phase. The ratio of the secondary compression index (Cα ) to the compression index (Cc), i.e., Cα /Cc is approximately constant for many normally consolidated clay over the normal range of engineering stress (Mesri and Godlewski, 1977; Mesri and Castro,1987). The coefficients of secondary compression obtained from the lumpy fill experiments are shown in Fig.4.45a. The values for the homogeneous clay in the undisturbed state and reconstituted state are also shown in the figure. The Cα values for the undisturbed sample is less than 1% in the overconsolidated range (σ′v < 200 kPa) and is about 2.4% in the normally consolidated state. For the sample reconstituted from slurry, Cα lies in the range 1.8 to 2.4%. The values of Cα for the lumpy fill are less than 2.4% which is within the range of values obtained for the homogeneous clay. This suggests that it is likely that all the inter-lump voids are closed. Mesri and Godlewski (1977) made a detailed study on the relationship between Cα and the compression index (Cc). They proposed the (Cα/Cc) concept and found that a majority of inorganic clays (Cα/Cc) lie in the range 0.04 ± 0.01. The values of (Cα/Cc) obtained for the homogeneous clay and the lumpy fill are plotted against the consolidation pressure in Fig. 4.45b. The range given by Mesri and Godlewski (1977) is also shown in the figure. The values obtained for the lumpy fill lie within the range of 76 other homogeneous soils. These results suggest that secondary compression is not a serious issue for the lumpy fill system, for the size evaluated here. 4.4 Experimental studies on large scale size lumps From the results of small size lumps ranging between 10 to 50mm, it is difficult to understand fully the behaviour of lumpy fill in land reclamation. So, in order to narrow this difficulty, consolidation tests using 200mm cubical shape clay lumps in a tank of size 1.4 m length, 1.0m width and 1.5m height were carried out. Cubical shape was selected, as it is easy to prepare lumps with this shape from large clay lumps. The weight and volume of each lump was measure to determine the wet density. The average wet density of the clay lumps used in this study is about 16.2 KN/m3. So far, because of the scale and the associated difficulties in arranging the logistics, only the case with fully swollen lumps (100%) was completed. 4.4.1 Time-settlement curves Typical time-settlement curves obtained under loading of 50kPa, 100kPa and 200kPa are shown in Fig. 4.46. Under small surcharge loadings, the rate of settlement is faster due to the presence of large inter-lump voids which permit practically free drainage of water during the consolidation process. Beyond a consolidation pressure of 50kPa, the rate of settlement slows down tremendously, due to the closing of the inter-lump voids, and the shape of the curve starts to change. The settlement data in Fig.4.46 is normalized with maximum value of the settlement for corresponding loading pressure to easy to compare with those of small size lump experiments results and it was shown in the 77 Fig.4.47. The same S-shaped compression curve as in the laboratory with small size lumps (12.5mm to 50mm) were observed at higher loading pressure. Fig. 4.48a shows the state of a lumpy fill immediately after placing the lumps in water. Large inter-lump voids are clearly visible. When the surcharge pressure reached 50kPa, the inter-lump voids are visibly closed as seen in Fig. 4.48b and the fill begins to look more like a homogenized clay. The same observations are observed in 50mm size cubical lumpy fill experiments. Thus the principal phenomena observed with tests in the laboratory with small size lumps are confirmed on a larger scale tests (with 200mm lumps). 4.4.2 Consolidation characteristics As already discussed in the section 4.3.1, the intrinsic frameworks proposed by Burland (1990) can be used to capture the compressibility characteristics of the lumpy fill layers. Figure 4.49 shows the e-log σ’v compression curves obtained from Oedometer tests on reconstituted soil specimens, which are sampled from the 200mm lumpy fill experiments at the end of applying 50, 100 and 200kPa surcharge pressure. Oedometer compression curves for samples obtained after 50kPa, 100kPa and 200kPa are plotted in terms of void index, Iv and are compared with the ICL in Fig. 4.50. When plotted in terms of Iv, the data fall on a unique ICL curve, as reported by Burland (1990). This ICL, obtained from reconstituted samples, can be used to identify normally consolidated and overconsolidated zones. Fig. 4.50a shows that most of the compression curves of intact samples lie below or close to the ICL. This means that the in-situ state of the sample is over consolidated under pressure of 50kPa. Similarly under pressure of 200kPa, most of the oedometer 78 compression curves of intact samples lie above the ICL, and some of these curves lie close the ICL as seen in Fig. 4.50c. From above results it was noted that, the strength and deformation characteristics of the lumpy fill layer was found to be variable with the normally consolidated and overconsolidated zones. The overconsolidated zones could be part of the original lumps, and the normally consolidated zones are believed to be from disintegrated material in the initial inter-lump voids. The data clearly indicate that under lower surcharge loading pressure the lumpy fill layer has both normally consolidated and overconsolidated zones, where as at higher surcharge load the most of the lumpy fill layers are in normally consolidated zones. Fig. 4.51a shows the variation of water content at the end of 50kPa, 100kPa and 200kPa with depth of lumpy fill. The water content of lumpy fill layer after 50kPa pressure varies from 50 to 70%, indicating that the lumpy fill layer is normally consolidated state. Similarly, after the consolidation pressure reached to the 200kPa, the water reduced to 25% to 40%, indicating that lumpy fill layer is overconsolidated to some extent and this is consistent with the findings of Karthikeyan et al. (2004). The variation of pre-consolidation pressure with depth at the end of 50kPa, 100kPa and 200kPa surcharge pressure are shown in Fig. 4.51b. The pre-consolidation pressures were determined from oedometer test results using the casagrande procedure (Casagrande 1936). The higher value of pre-consolidation pressure could means the soil sample was taken from inside the lumps, and a lower value of pre-consolidation pressure may indicate the soil was taken from the soil occupying the initial inter-lump voids. A significant degree of scattering is observed at the end 50kPa pressure and it is due to the presence of inter-lump voids in the lumpy fill layer. But at the end of applying 200kPa 79 pressure, the variation of pre-consolidation pressure with depth is almost uniform. It means that the lump fill layer is reached to the homogenious state. The variation of the compression index with the depth for 50kPa, 100 kPa and 200kPa are shown in Fig. 4.51c. The compression index varies from 0.1 to 0.3. The compression index, Cc is plotted versus the void ratio, e for the lumpy fill layer at the end of 50kPa, 100kPa and 200kPa in Fig.4.52. It is observed that at a given void ratio, Cc is probably higher for higher surcharge load pressure than lower pressure. 4.4.3 Shear strength profile It is important to evaluate the strength profiles of the fill upon stabilization by surcharge loading of different magnitude. A miniature cone of 10mm diameter with a tip angle of 600 was used at a penetration of 5mm/s in the present study to measure strength profile of the lump fill. The shear strength profiles obtained by using cone penetration tests at the end of surcharge pressures 100kPa and 200kPa are shown in Fig. 4.53. Under consolidation pressure of 100kPa, the shear strength is not uniform with depth. A line representing the shear strength of normally consolidated Singapore marine clay, su = 0.23σ′v suggested by Tanaka et al. (2001), is also shown in Fig. 4.53. The comparison reveals that certain parts of the layer have strength similar to that of clay in the normally consolidated state and also there are zones with relatively higher strengths. When the consolidation pressure applied is equal to the preconsolidation pressure (Fig.4.53b), the strength profile is almost uniform. This suggests that to make the fill uniform, consolidation pressure equal to or greater than the preconsolidation pressure of the lump 80 should be applied to the fill. The same type observation obtained from lumpy fill made of 50 mm cubical lumps with 100% degree of swelling. 4. 5 Summary An experimental study has been conducted to investigate the performance of reclamation fill made up of dredged stiff clay lumps. A small scale one-dimensional consolidation test was carried out to examine the effect of size, shape, degree of swelling etc. on compressibility behavoiur of lumpy fill. Also a large scale test using 200mm size cubical lumps was conducted to study the effect of size. From the series of small and large scale one-dimensional consolidation tests on lumpy clay fills, the following findings are established. Tests on how the clay lumps would soften were conducted to provide a consistent platform for comparison between samples in different tests. The results discussed earlier in this chapter showed that the degree of swelling has a direct influence on the surcharge pressure required to close the inter-lump voids. When the lumps were allowed to swell fully, the effective stress within the lumps was reduced due to the loss of suction with a consequent reduction in the stiffness. With more compressible lumps, the pressure required to close the inter-lump voids will be smaller as shown in all the tests. So, the surcharge pressure required to close the inter-lump voids of fully swollen lumps is considerably smaller than that required for lumps that were subjected to partial swelling. Therefore it is important to consider the influence of swelling of lumps, when clay lumps are used in land reclamation. 81 To investigate the effect of size of lump on compressibility behaviour of lumpy fill, one dimensional consolidation tests were carried out on different sizes of clay lumps with the same initial packing. The results suggests that there is no significant effect of size on the compressibility behaviour of the lumpy fill system within the range of lump sizes investigated, when the lumps were subjected to 100% degree of swelling. The influence of initial packing arrangements of the lumpy fill system plays a significant role in defining the compressibility behaviour, especially in the earlier loading stages. The lumpy fill system with a higher initial void ratio will experience a greater reduction in void ratio when loaded. As expected, lumps with higher initial inter lump voids ratio, eer will need a higher surcharge pressure to close the inter-lump voids than lumpy fill with a lower initial inter-lump void ratio. The effect of pre-consolidation pressure is different with lumps subjected to different degree of swelling. In the case of Us=0%, the lumpy fill made of clay lumps with 350kPa pre-consolidation pressure requires a higher surcharge pressure to close the inter-lump voids, compared to the case with lumps subjected to 100kPa. But in case of Us=100%, the lumpy fill made of clay lumps with 100kPa pre-consolidation pressure requires a higher surcharge pressure to close the inter-lump voids than those of the 350kPa. It is noted that lumpy fill made up of spherical lumps experienced less settlement compared to cubical lumps due to its denser packing. It is also observed that the surcharge pressure required to close all inter lump voids is much less in the case of spherical lumps than for cubical lumps. This is attributed to spherical lumps having a 82 more stable open structure that effectively formed a densely packed structure within the lumpy fill. From Iv – log σ′v curves it was noted that, the lumpy fill system with inter lump voids containing water and slurry of 151% that the rate of settlement of the system filled with slurry is much smaller than that filled with water for pressure less than 50kPa. Under high pressure, the difference is small. The presence of slurry reduces the rate of settlement of the system under small surcharge pressures. Pore water pressure measurements in inter and intra lump voids for fully swollen lumpy fill indicate that, the consolidation pressure required to reduce the permeability of the fill to the order of magnitude of uniform clay is about 100 kPa. However, evaluation of shear strength using the cone penetration test indicates that the fill is highly heterogeneous under a pressure of 100 kPa. The shear strength of the fill lies in between that corresponding to the normally consolidated state and over-consolidated state, similar to that observed in one of the field tests on a reclaimed land using dredged clay lumps. When the consolidation pressure is equal to or greater than the preconsolidation pressure of the lumps, the strength profile is uniform. The coefficient of secondary compression of the lumpy fill is comparable to the homogeneous clay indicating that inter-lump voids were effectively closed and secondary compression is no more significant than for the undisturbed clay. . 83 Degree of swelling, % 120 100 80 60 40 50mm Wi =50% 50mm, Wi=60% 20 25mm,Wi=50% 25mm,Wi=60% 0 0 50 100 150 200 250 300 350 400 Time, Hours Fig.4.1 Comparison of degree of swelling for different size and initial water content of cubical clay lumps 120 Degree of Swelling, Us 100 80 60 40 P =100kPa 20 P =200kPa P =300 kPa 0 0 20 40 60 80 100 120 140 160 180 200 Time, Hours . Fig.4.2 Comparison of degree of swelling for different pre-consolidation pressure of 25mm remolded cubical clay lumps 84 Table.4.1 Time required reaching t50 and t100 Test No 1 Size of lumps, mm 25 Initial water content,% 50 Pre-consolidation pressure, kPa UD t50( min) 20 t100( min) 7380 2 25 60 UD 216 11700 3 50 50 UD 660 14700 4 50 60 UD 1020 18720 5 25 50 100 840 10680 6 25 46 200 900 9660 7 25 44 350 336 8820 ** t50= time at 50% degree of softening, t100= time at 100% degree of softening Fig.4.3 State of clay lumps after swelling process 85 Fig.4.4 Schematic profile of reclamation with dredged clay lumps and sand Surcharge Inter-lump voids filled with water Fig.4.5. The apparatus on One-dimensional consolidation tests of lumpy clay 86 3.5 Void ratio, e 3 2.5 2 1.5 (a) 1 1 10 100 1000 Consolidation pressure, kPa 1.E-04 Permeability, m/s 1.E-05 1.E-06 1.E-07 1.E-08 1.E-09 (b) 1.E-10 1 10 100 1000 Consolidation pressure, kPa Fig.4.6 Typical e-log p and permeability curves for lumpy fill consolidation experiments. 87 3.0 Void ratio, e 2.5 e*30 2.0 eo =1.72 30 1.5 1.0 0.5 0.0 1 10 100 1000 10000 Consolidation pressure, kPa Void Index, Iv Fig.4.7. One-dimensional compression curves for reconstituted soil samples. 2.5 2.0 1.5 1.0 0.5 0.0 -0.5 -1.0 -1.5 -2.0 ICL Proposed by Burland (1990) 1 10 100 1000 Consolidation Pressure, kPa Fig.4.8. One-dimensional compression curves for reconstituted soil samples in terms of void index. 88 e e30 es e800 (a) log σ′ v, kPa IV = Iv * * e s −e30 e s − e30 = * * e30 − e800 1.42C c* 0 Iv -1 (b) log σ′ v, kPa Fig. 4.9. The use of void Index Iv to normalized compression curve of lumpy fill. 89 Clay lumps Fig.4.10 Typical dredged material obtained from lower marine clay 3.5 S 3 / es / eer / era Void ratio, e 12.5 mm/2.86/1.56/1.30 25mm/3.11/1.75/1.36 2.5 50 mm/3.14/1.83/1.31 2 1.5 1 1 10 100 Consolidation pressure, kPa 1000 Fig. 4.11 Effect of size of lumps on e-log σ’v relationship (Us=100%) 90 2.0 S / es / eer / era Void index, I v 1.5 12.5mm/2.86/1.56/1.30 1.0 25mm/3.11/1.75/1.36 0.5 50mm/3.14/1.83/1.31 0.0 -0.5 -1.0 ICL ICL -1.5 S= Size of lump, es= System void ratio, eer= inter-lump void ratio, era=intra-lump void ratio -2.0 1 10 100 1000 Consolidation pressure, kPa Fig.4.12. Variation of voids index with consolidation pressure for different sizes of clay lumps(initial degree of swelling = 100%) 1.0E-04 S / es / eer / era Permeability, k(m/s) 1.0E-05 12.5mm/2.86/1.56/1.30 25mm/3.11/1.75/1.36 1.0E-06 50mm/3.14/1.83/1.31 1.0E-07 1.0E-08 1.0E-09 1.0E-10 S= Size of lump, es= System void ratio, eer= inter-lump void ratio, era=intra-lump void ratio 1.0E-11 1 10 Consolidation pressure, 100 ’ σ s' v v 1000 (kPa) Fig.4.13. Permeability of lumpy fill made of 12.5mm, 25mm and 50mm cubical Lumps with 100% initial degree of swelling. 91 Consolidation pressure,kPa 100 80 60 40 I v - log σ′v e- log σ′v 20 0 0 10 20 30 Lump size, mm 40 50 60 Fig.4.14. Variation of consolidation pressure required to close the inter-lump voids of lumpy fill with size of the lump. 1.5 S / es / eer / era 25mm spheri/2.89/1.54/1.35 Void Index, Iv 1.0 25mm cube/2.99/1.73/1.26 0.5 0.0 -0.5 S= Size of lump, es= System void ratio, eer= inter-lump void ratio, era=intra-lump void ratio -1.0 1 10 100 1000 Consolidation pressure,kPa Fig. 4.15. Effect of shape of clay lumps made up of 25 mm lumps with 100 % degree of swelling on Iv - log σ′v curves. 92 1.0E-04 S / es / eer / era Permeability, k(m/s) 1.0E-05 25mm speri/2.89/1.54/1.35 25mm cube/2.99/1.73/1.26 1.0E-06 1.0E-07 1.0E-08 1.0E-09 1.0E-10 S= Size of lump, es= System void ratio, eer= inter-lump void ratio, era=intra-lump void ratio 1.0E-11 1 10 100 Consolidation Fig.4.16. σv’ v s' pressure,s' 1000 (kPa) Permeability of lumpy fill made of 25mm for different shapes with 100% degree of swelling 90 50mm, UD Clay Lump 50mm, Pc'=100kPa Moisture Content, % 85 50mm, Pc'=350kPa 80 Pc'=100kPa 0 75 70 UD 65 Pc' =350kPa 60 -30 -20 -10 0 10 20 30 Distance from Centre of lump, mm Fig.4.17. Variation of moisture content with distance from the centre of clay lump (Us = 50%) 93 3.5 S / Us / es / eer / era 50mm/0%/3.16/1.69/1.47 Void ratio, e 3.0 50mm/50%/2.96/1.75/1.21 50mm/100%/3.14/1.83/1.31 Undisturbed 2.5 e50% =1.75 e100% =1.83 2.0 1.5 e0% =1.59 1.0 S= Size of lump, Us= Degree of swelling, es= System void ratio, eer= inter-lump void ratio, era=intra-lump void ratio 0.5 1 10 100 1000 Consolidation pressure, kPa Fig.4.18. Effect of degree of swelling, Us on e- log σ′v of lumpy fill made of 50mm cubical lumps (a) (b) (c) Fig.4.19. View of the lumpy fill made of 50mm cubical lumps under consolidation pressure of 50kPa with lumps of (a) Us=0%, (b) Us= 50%, (c) Us= 100% 94 1.0E-04 S / Us / es / eer / era 50mm/0%/3.16/1.69/1.47 1.0E-06 50mm/50%/2.96/1.75/1.21 50mm/100%/3.14/1.83/1.31 1.0E-07 1.0E-08 1.0E-09 1.0E-10 S= Size of lump, Us= Degree of swelling, es = System void ratio, eer= inter-lump void ratio, era=intra-lump void ratio 1.0E-11 1 10 100 1000 Consolidation pressure, kPa Fig.4.20. Permeability of lumpy fill made of 50mm cubical lumps for different degree of swelling. Time, s 1 10 100 1000 10000 100000 1000000 0.0 Normalized settelment Permeability,k( m/s) 1.0E-05 0.2 Us= 100% 0.4 Us=50% 0.6 0.8 Us= 0% 1.0 Fig.4.21 Typical time-compression curve of lumpy fill made up of 50mm cubical lumps under consolidation pressure of 50 kPa. 95 2.5 2.0 S / Us / es / eer / era 50mm/0%/3.16/1.69/1.47 50mm/50%/2.96/1.75/1.21 50mm/100%/3.14/1.83/1.31 ICL 1.5 Void Index, Iv 1.0 0.5 eer= 0 0.0 -0.5 S= Size of lump, Us= Degree of swelling, es = System void ratio, eer= inter-lump void ratio, era=intra-lump void ratio -1.0 -1.5 1 10 100 1000 Consolidation pressure, kPa Fig.4.22. Variation of voids index with consolidation pressure for different degree of swelling for lumpy fill made of 50mm size cubical lumps 140 Consolidation pressure, kPa 120 k - log σ′v Iv - log σ′v 100 e - log σ′v 80 60 40 20 0 0 20 40 60 80 100 120 Degree of swelling, Us (%) Fig.4.23. Variation of consolidation pressure required to close the inter-lump voids of lumpy fill with degree of swelling 96 (a) 0 kPa (c) 25kPa (b) 12.5kPa (d) 50kPa Fig.4.24. Photographs of lumpy fill made up of 50 mm cubical lumps with Us= 100% and eiv = 1.68 under different loading stages. 97 Time, s 1 10 100 1000 10000 100000 1000000 Normalised settlement 0 0.2 100-200 kPa 0.4 25 - 50 kPa 0.6 50-100 kPa 0.8 (a) 1 Time, s 1 10 100 1000 10000 100000 1000000 Normalized settlement 0 0.2 50-100 k Pa 0.4 100-200 kPa 16-27 kPa 0.6 27-50 k Pa 0.8 (b) 1 Time, s 1 10 100 1000 10000 100000 1000000 0.0 Normalised settlement 50-100 kPa 0.2 27-50 kPa 0.4 0.6 12 -27 kPa 0.8 1.0 6 -12 kPa (c) Fig.4.25 Typical time-compression curve of lumpy fill made up of 50 mm cubical lumps with Us=100% for (a) eiv = 2.25, (b) eiv = 1.68 and (c) eiv=1.41 98 Fig.4.26. View of the lumpy fill made of 50mm cubical lumps under consolidation pressure of 50kPa with lumps of (a) eiv = 1.41 (b) eiv = 1.68, (c) eiv = 2.25 2.0 1.5 S / es / eer / era Void Index, I v 50mm /2.99/1.41/1.58 50mm/3.14/1.68/1.46 1.0 50mm/3.56/2.25/1.31 0.5 eer = 0 0.0 -0.5 S= Size of lump, es= System void ratio, eer= inter-lump void ratio, era = intra-lump void ratio -1.0 1 10 100 1000 Consolidation pressure, kPa Fig.4.27 Effect of initial inter-lump void ratio of lumpy fill made up of 50 mm cubical lumps with 100% degree of swelling on Iv - log σ′v curves. 99 1.0E-04 S / es / eer / era Permeability, k(m/s) 1.0E-05 50mm/2.99/1.41/1.58 50mm/3.14/1.68/1.46 1.0E-06 50mm/3.56/2.25/1.31 1.0E-07 1.0E-08 1.0E-09 1.0E-10 S= Size of lump, es= System void ratio, eer= inter-lump void ratio, era=intra-lump void ratio 1.0E-11 1.00 10.00 100.00 1000.00 s' v(kPa) Consolidation pressure,s' Fig.4.28. Permeability of lumpy fill made of 50 mm cubical lumps with different initial inter-lump void ratio. Consolidation pressure, kPa 100 80 60 40 I v - log σ′v 20 0 1.00 k - log σ′v 1.25 1.50 1.75 2.00 2.25 2.50 Initial Inter-lump void ratio Fig.4.29. Variation of consolidation pressure required to close the inter-lump voids of lumpy fill made of 50mm cubical lumps with initial inter-lump void ratio. 100 Fig.4.30 Geological location of Reclamation site. Fig.4.31 Geotechnical Characteristics of Seabed soil prior to reclamation (After Karthikeyan, 2005). 101 4.0 S / P′c / Us / es / eer / era Void Index, Iv 3.0 50mm/350kPa/0%/3.05/1.85/1.20 50mm/350kpa/100%/2.99/1.74/1.25 2.0 50mm/100kPa/0%/3.15/1.84/1.31 50mm/100kPa/100%/2.96/1.71/1.25 1.0 eer= 0 0.0 -1.0 S= Size of lump, P′c= Pre-consolidation pressure, Us= Degree of swelling, es = System void ratio, eer= inter-lump void ratio, era=intra-lump void ratio -2.0 1 10 100 1000 Consolidation pressure, kPa Fig.4.32. Effect of pre-consolidation pressure of clay lumps made up of 50 mm cubical lumps with 0% and 100% degree of swelling on Iv - log σ′v curves. 1.0E-04 S / P′c / Us / es / eer / era Permeability,k(m/s) 1.0E-05 50mm/350kPa/0%/3.05/1.85/1.20 50mm/350kPa/100%/2.99/1.74/1.25 1.0E-06 50mm/100kPa/0%/3.15/1.84/1.31 50mm/100kPa/100%/2.96/1.71/1.25 1.0E-07 1.0E-08 1.0E-09 S= Size of lump, P′c= Pre-consolidation pressure, Us= Degree of swelling, es= System void ratio, eer= inter-lump void ratio, era=intra-lump void ratio 1.0E-10 1.0E-11 1 10 100 1000 Consolidation pressure, s ' v(kPa) Fig.4.33. Permeability of lumpy fill made of 50mm cubical lumps with 0% and 100% degree of swelling for different pre-consolidation pressure. 102 Consolidation pressure, kPa 120 350 kPa (I v - log σ′v ) 100 100 kPa (I v - log σ′v ) 350 kPa(k- log σ′v ) 100 kPa(k- log σ′v ) 80 60 40 20 0 0 20 40 60 80 100 120 Degree of swelling, Us (%) Fig.4.34. Variation of consolidation pressure required to close the inter-lump voids with degree of swelling for a pre-consolidation pressure of 100kPa and 350kPa. Fig.4.35 Interpreted soil profile at Pulau Tekong (After Tan et.al). 103 C orrected cone resistance, q t (MPa) 0 4 8 12 0 Surface soft marine clay 5 Upper marine clay Depth below seabed (m) 10 Intermediate layer 15 Pore pressure 20 Lower marine clay Cone resistance 25 30 Weathered rock 35 0 0.4 0.8 1.2 Pore pressure, u2 (MPa) 1.6 2 Fig.4.36 Typical Cone resistance and pore pressure Profile of seabed. Very soft surface marine clay Fig. 4.37 A view of the surface soft marine clay dredged from the seabed (SSMC). 104 Time, s 1 10 100 1000 10000 100000 1000000 0 ILV with slurry (w=150%) Settlement, mm 5 10 ILV with water 15 12 - 25 kPa 20 Time, s 1 10 100 1000 10000 100000 1000000 0 Settlement, mm ILV with slurry (w=150%) 4 ILV with water 8 50-100 kPa 12 Time, s 1 10 100 1000 10000 100000 1000000 Settlement, mm 0 ILV with slurry (w=150%) 4 8 ILV with water 200-300 kPa 12 Fig.4.38 Typical time-compression curves for inter-lump voids (ILV) filled with water / slurry. 105 2.5 Void Index, I v 2.0 50mm/water/0%/3.26/1.83/1.43 1.5 50mm/slurry/0%/3.15/1.83/1.32 1.0 0.5 eer=0 0.0 -0.5 -1.0 1 10 100 1000 Consolidation pressure, kPa Fig. 4.39 Effect of inter lump voids filled with slurry / water for the lumpy fill made up of 50 mm lumps on Iv - log σ′v curves. 1.0E-04 Permeability, k(m/s) 1.0E-05 50mm/water/0%/3.26/1.83/1.43 1.0E-06 50mm/slurry/0%/3.15/1.83/1.32 1.0E-07 1.0E-08 1.0E-09 1.0E-10 1.0E-11 1 10 100 1000 s' v ( kPa) Consolidation pressure,s' Fig.4.40 Variation of permeability of lumpy fill made of 50mm cubical lumps with interlump voids filled with slurry/water. 106 30 Pressure increment=25 kPa Pore pressure, kPa 25 Inside the lump 20 15 In between the lumps 10 5 (a) 0 1 10 100 1000 10000 100000 Time, s 120 Pressure increment=100 kPa Pore pressure, kPa 100 80 In between the lumps Inside the lump 60 40 20 (b) 0 1 10 100 1000 10000 100000 1000000 1000000 0 Time, s Fig.4.41 Pore pressure under pressure increments of (a) 25-50kPa; and (b) 100-200kPa (inter-lump voids filled with water) 107 Piston Stress σ Vessel “Water in intralump voids” “Soil skeleton” “Water in inter- “Lump structure” lump voids” Fig.4.42 Spring - Box analogy (after Yang et al., 2002). 120 Pressure increment=100 kPa Pore pressure, kPa 100 80 60 40 Inside the lump Pressure increment=25 kPa 20 In between the lumps 0 1 10 100 1000 10000 100000 1000000 10000000 Time, s Fig.4.43 Pore pressure under pressure increments of 25-50kPa and 100-200kPa (inter-lump voids filled with slurry) 108 s u , kPa 0 10 20 30 s u , kPa 40 50 0 0 0 20 20 10 20 30 40 50 80 100 su =0.23 σv ' (OCR) 0.75 40 Depth, mm Depth, mm 40 60 80 100 su =0.23 σv ' 120 s u=0.23 σ v'(O CR) 0.75 60 80 s u=0.23 σ v' 100 120 (a) σv'=50 kPa 140 (b) σv'=100 kPa 140 s u , kPa 10 20 30 40 50 0 0 0 20 20 40 40 60 s u=0.23 σ v' 80 100 Depth, mm Depth, mm 0 s u , kPa 20 40 60 60 80 100 s u=0.23 σ v' 120 140 (c) σv'=200 kPa 120 (d) σv'=360 kPa 140 Fig.4.44 Shear strength profiles obtained from cone penetration tests for lumpy 109 fill made of 50mm cubical lumps with eiv=2.25 4.0 Undisturbed ICL 12.5mm cube/2.86/1.56/1.30 25mm cube/3.11/1.75/1.36 Ca (%) 3.0 50mm cube/3.14/1.31/1.83 2.0 1.0 0.0 10 100 s' v (kPa) Consolidation pressure,s' 1000 (a) 0.07 Undisturbed ICL Range for homogeneous clay (Mesri and Godlewski, 1977) 12.5mm cube/2.86/1.56/1.30 0.06 25mm cube/3.11/1.75/1.36 50mm cube/3.14/1.31/1.83 (Ca/Cc ) = 0.05 (Ca/Cc) 0.05 0.04 0.03 (Ca/Cc ) = 0.03 0.02 0.01 10 100 1000 s' v (kPa) Consolidation pressure,s' (b) Fig. 4.45 Secondary compression of lumpy fill 110 Time, hr's 0.001 0.01 0.1 1 10 100 1000 0 ≈ 1.5m 40 60 80 50kPa 100 100 kPa 200 kPa 120 Fig.4.46 Variation of surface settlement with time at different load steps Time, s 0.001 0.01 0.1 1 10 100 1000 0.0 Normalized settlement . Settlement, mm 20 0.2 200 kPa 100 kPa 0.4 0.6 0.8 50 kPa 1.0 Fig.4.47 Variation of normalized settlement with time at different load steps 111 a) Before test (b) End of 50kPa loading Fig.4.48 Closing of inter- lump voids 112 2.2 2.0 Void ratio, e 1.8 53.5-56.5 cm 44.7-47.2 cm 38.9-41.4 cm 32.2-34.8 cm 26.7-29.3 cm 20.9-23.4 cm 14.9-17.9 cm 9-11.9 cm 3.5-6 cm 1.6 1.4 1.2 1.0 (a) 50kPa 0.8 0.6 1 2.2 2.0 Void ratio, e 1.8 10 100 1000 10000 3-6 cm 8-11cm 13.2-15.8cm 18.5-20.5cm 23.5-22.5cm 28-31cm 33-36cm 44-47cm 51.5-54.5cm 1.6 1.4 1.2 1.0 0.8 0.6 (b) 100kPa 0.4 1 10 2.4 2.2 Void ratio, e 2.0 1.8 100 1000 10000 0-3cm 7.5-10 cm 12-14.5 cm 16.5-19 cm 22-24.5 cm 26.5-29 cm 35.5-38 cm 40-42.5 cm 49-51.5 cm 1.6 1.4 1.2 1.0 0.8 (c) 200kPa 0.6 0.4 1 10 100 1000 10000 Consoldiation pressure,kPa Fig.4.49 One dimensional compression curves for reconstituted soil sample under 50, 100 and 200kPa 113 1.5 1.0 ICL 3.5-6.0cm 9.0-11.9cm 14.9-17.9cm 20.9-23.4cm 26.7-29.3cm 32.2-34.8cm 38.9-41.4cm 44.7-47.2cm 53.5-56.5cm Void Index, Iv 0.5 0.0 OC NC -0.5 -1.0 -1.5 (a) 50 kPa -2.0 1 10 100 1000 10000 1.5 1.0 ICL 3.0-6.0cm 8.0-11.0cm 13.2-15.8cm 18.5-20.5cm 23.5-25.5cm 28.0-31.0 33.0-36.0cm 44.0-47.0cm 51.5-54.5cm Void Index, Iv 0.5 0.0 NC -0.5 OC -1.0 -1.5 (b) 100 kPa -2.0 1 10 100 1000 10000 1.5 1.0 ICL 0.0-3.0cm 7.5-10.0cm 12.0-14.5cm 16.5-19.0cm 22.0-24.5cm 26.5-29.0cm 35.5-38.0cm 40.0-42.5cm 49.0-51.5cm Void Index, Iv 0.5 0.0 NC -0.5 OC -1.0 -1.5 (c) 200 kPa -2.0 1 10 100 1000 10000 Consolidation pressure,kPa Fig.4.50 Comparison of one-dimensional compression curves for undisturbed samples with ICL under surcharge pressure of 50kPa, 100kPa and 200kPa 114 Water Content,% 20 40 60 0 0 10 Depth, cm Depth,cm 20 30 40 50 60 Compression Index,Cc Preconsolidation Pressure, kPa 80 100 200 0 300 0 0 10 10 20 20 30 Depth, cm 0 40 50 50 kPa 50kPa 100kPa 200 kPa (a) 0.3 0.4 30 40 60 200kPa 70 70 0.2 50 60 100 kPa 0.1 70 (b) 50 kPa 100 kPa 200 kPa (c) Fig.4.51. The variation of (a) Water content, (b) pre-consolidation pressure and (c) Compression index with depth from bottom surface of the lumpy fill layer for surcharge pressure of 50kPa, 100kPa and 200kPa 115 Compression Index, Cc 0.5 50 kPa 100 kPa 200 kPa 0.4 0.3 y = 0.3533x - 0.224 0.2 y = 0.2216x - 0.1081 y = 0.1802x - 0.0426 0.1 0.0 0.5 0.7 0.9 1.1 1.3 1.5 1.7 1.9 Void ratio,e Fig.4.52 Variation of compression Index with void ratio s u , kPa s u , kPa 0 10 20 30 40 50 0 60 0 20 30 40 50 60 100 100 200 Depth, mm 200 D epth, mm 10 0 300 Su= 0.23 σ′v 400 Su= 0.23 σ′v (OCR)0.75 300 Su= 0.23 σ′v 400 500 500 (a) σ′v =100kPa (b) σ′v =200kPa 600 600 Fig.4.53. Shear strength profiles obtained from cone penetration test for 100 kPa and 200kPa surcharge pressure 116 CHAPTER 5 Results and Discussion - II 5.1 Introduction The effects of soil structure on the mechanical behaviours of natural soils have received considerable attention since last decade. Several frameworks have been proposed to characterize the structure in natural soils. In the previous chapter, an intrinsic framework that was proposed by Burland (1990) was used to characterize the compression behaviour of the lumpy fill for different initial conditions. In this chapter, another possible structural mechanism to describe the transition from a lumpy fill to homogenized state will be discussed. Lambe and Whitman (1965) and Mitchell (1993) used the term “soil structure” to describe the effects of fabric (arrangement of soil particles) and inter-particle bonds (those interparticle forces which are not of a purely frictional nature). Burland (1990) used the term ‘intrinsic’ to describe the properties of soil which have been reconstituted at water content of 1 to 1.5 times liquid limit and then consolidated under one- dimensional state. He used these intrinsic properties as a reference against which natural soils are compared, and the difference in behaviour is attributed to the structure. Feda (1998) studied the transition of clay from granular to a saturated cohesive material due to overburden and an increase in water content. This transition is modelled by laboratory experiments considering the effect of stress, water and time. Due to its high porosity, the materials that have a metastable structure collapse. 117 They studied the collapsible behaviour of fragmentary clay through rearrangement, collapse and crushing of fragments. So far, no study relating to the structural behaviour of the lumpy fill during the consolidation period was carried. In this chapter, the stages of consolidation behaviour of the lumpy fill are discussed, and from this, a possible mechanism is proposed. 5.2 Fabric and hydraulic conductivity For a given soil, a fabric with a high proportion of large pores is much more pervious than those with small pores. For example, remolding undisturbed soft clays has been found to reduce the hydraulic conductivity by as much as a factor of 4, with an average of about 2 (Mitchell, 1956). This is because of the breaking down of a flocculated open fabric and the destruction of larger pores. In the previous chapter, it was observed that for homogenous soil, the change in permeability with consolidation pressure is very small as compared with the change in a lumpy fill system. Also it was observed that in a lumpy fill system, there are three possible stages of consolidation behaviour based on the changing trend in the system permeabilility. At the initial stage, the permeability is very large but the change in permeability values with applied pressure is small. After this stage, there is a range with sharp drop in permeability and this point is very distinct. Finally, the permeability is of the order of the remoulded homogeneous clay and there is no significant change in permeability again. These three stages are likely to be associated with large plastic deformation of lumps followed be a stage with both plastic deformation and consolidation, and finally when the inter-lump voids are as small as the voids in the lump, there is the stage for usual soil consolidation alone. 118 These three stages of fabric changes are important when considering the consolidation behaviour of lumpy fill system moving towards a homogeneous state. The ideas behind this is discussed in the next section. Olsen (1962) proposed a cluster model for permeability prediction of fine grained soils with existence of unequal pore sizes. A typical soil with fabric composed of small clusters as shown schematically in Fig.5.1, having an intracluster void ratio ec. The spaces between the aggregates comprise the intercluster voids and are responsible for the intercluster void ratios, ep. The total void ratio, eT is equal to the sum of ec and ep. The size of cluster that formed depends on the mineralogical and pore fluid composition. In the present study, the size of lumps formed depends on the stiffness and water content of the soil. The cluster model developed by Olsen (1962) accounts for discrepancies between the predicted and measured variation of flow rates and the equation for the ratio of estimated flow rate for a cluster model, qCM to the flow rate predicted by the KozenyCarman equation, qKC is given in Equation 5.1. 3 q CM q KC  ec  1 −  eT  = N 2/3  [1 − ec ]4 / 3 (5.1) where N is the total number of clusters / number of lumps. As the flow rate is predicted using the Kozeny-Carman equation, the above equation is more suitable for cohesionless soil. But our present problem deals with cohesive soil only. The above equation was adopted to observe the flow behaviour of lumpy fill under surcharge loading by changing era for ec and es for eT. era is the intra lump voids is the lumpy fill, where ec is the cluster 119 voids ration in the Olsen(1962) cluster model. Similarly, es is the system void ratio and is equal to sum of the inter and intra lump voids of the lumpy fill, where as eT is the total void ratio of the cluster model and is equal to the sum of cluster and inter cluster void ratio. The modified equation for the lumpy fill system is given in Equation 5.2. 3 q CM = N 2/3 q KC  era  1 −  es   [1 − era ]4 / 3 (5.2) Application of the above equation will need some idea n how era and es changes with accompanying compression of the lumpy fill. Olsen(1962), made assumption for changing es and era with loading pressure and is shown in Fig.5.2, and these assumptions were based on consideration of the relative compressibilities of individual clusters and cluster assemblages. The compressibility of individual lumps is small at initial loading stages, so compression is accompanied by reduction in the inter-lump void ratio, but with litter changes in the intra-lump void ratio. As a result, the actual permeability decreases more rapidly with decreasing void ratio during compression than predicted by the Kozeny- Carman equation until the interlump pore space is comparable to the intra lump pore space. At this stage, the lumps themselves begin to compress as at this point, most of the inter lump voids are already closed. Further decrease in the overall system void ratio involves decrease in both intra and inter lump void ratios. But the decrease in the interlump void ratio is insignificant compared to that of decrease in intra lump void ratios. The above Olsen(1962) cluster model was modified to study the lumpy fill consolidation behaviour with different size of lumps and degree of swelling. 120 A line parallel to the variation of system void ratio at a perpendicular distance of 0.43 used by Olsen(1962). The line touch the intra lump void ratio, from this point onwards the inter lump void ratio is reduced with loading pressure. Fig. 5.3 shows the assumed relationship between the total, inter and intra lump void ratio for lumpy fill made of using 12.5mm, 25mm and 50mm cubical lumps. Using Equation 5.2 and Figure 5.3, the ratio of flow rate in inter lump voids to the flow rate in intra lump void was calculated and shown in Fig.5.4. From this figure, it was observed that, there is a rapid reduction in the ratio of estimated to predicted flow rates at high porosity values until the inter lump pore space is comparable to that in a system of closely packed lumps, when the lumps themselves begin to compress. Similarly for the lumpy fill made of 50mm cubical lumps with 0%, 50% and 100% degree of swelling, the results are shown in Fig.5.5. 5.3 Possible structural mechanisms In the previous chapter, the pressure required to close the inter lump voids for different size of lumps, degree of swelling, packing arrangements etc. were discussed. The mechanism leading to this behavoiur of the lumpy fill is very important in the study of how the inter and intra lump voids were closed during consolidation process. The Olsen (1962) cluster model shown in Fig.5.1 considered only two stages of consolidation. A typical variation of permeability with consolidation pressure for the lumpy fill system is shown in Fig. 5.6. From permeability measurement of the lumpy fill during the consolidation, it was noted that there is possibility of one more stage and in the present study the Olsen model is modified in to three stage of consolidation. A typical assumed 121 relationship between total, inter, and intra lump void ratios are shown in Fig.5.7. In this figure it also noted the stages of consolidation in terms of voids ratios and these stages were discussed in the below paragraphs. It is mentioned previously that there are three likely stages of consolidation behaviour in the lumpy fill system based on the trend of changing permeability. These structural mechanisms causing the transition of the lumpy fill into a homogeneous soil behaviour are a largely plastic deformation stage, a combined plastic deformation and consolidation stage and a mainly consolidation stage alone. • Stage -1 (Plastic deformation): In this beginning stage the lumps plastically deform without rupture of the lumps and thus the volume of each lump remains essentially unchanged, that is the lumps deformed in an undrained manner plastically . At this high porosity stage, compression results mainly from decreases in the size of inter lump voids. This means that the compressibility of individual lumps is small. Thus, at high void ratios, compression develop at the expense of inter lump voids. The intra lump voids are assumed to be a constant value in this largely plasticity deformation, i.e. in between points A and B. During this stage of consolidation, softening of the clay lumps plays an important role to facilitate the compression of the lumpy fill by plastic deformation of the lumps themselves. • Stage-2 (Plastic deformation and consolidation): In the second stage, the plastic deformation cum consolidation stage (between point B and C), there is a change in the mechanical behaviour as the soil transform from one structural configuration to another one. It is visualized by a stepwise compression and supported by the stepwise permeability measurements. In this stage the reduction in the system void ratio is coming 122 from reduction in inter lump voids and a little from consolidation of the lump. By the third stage, most of the inter lump voids were reduced to an extent that they are comparable to the voids in the lumps themselves. Beyond this point, the decrease in void ratio involves mainly compression of the lumps themselves only. • Stage-3 (Consolidation Alone): After the point C and up to the point D, the change in the system void ratio is mainly coming from the compression of lumps themselves. For the first two stages of consolidation, the shape of the time-compression curve in the logarithmic plot is concave, in marked contrast to those observed for homogenous clays, following the classic Terzaghi’s model. During the third stage of lumpy fill consolidation, the rate of settlement is much slower due to the closure of the inter-lump voids and the shape of the curve started changing. Typical S-shaped curves are obtained in the logarithmic plot in the third stage of consolidation. In the following sections, the above concept will be examined for different cases involves different sizes and degree of swelling of lumps to understand the changes in the structural mechanisms with the consolidation pressure for a lumpy fill. 5.3.1 Lump size and Degree of swelling of lumps In order to study the behaviour of lumps in a lumpy fill during consolidation pressure, a series of laboratory experiments was undertaken to model the fragmentary nature of the lumpy fill using cubical lumps with 12.5mm, 25mm and 50mm size. To make it easy to understand the variation of inter and intra lump voids present in the lumpy fill, assumptions similar to those made in Olsen’s (1962) cluster model are also adopted here. 123 The assumed relationships between the total, inter and intra lump void ratios for lumpy fill made of 12.5mm, 25mm and 50mm cubical lumps are shown in the Fig.5.8. In this figure, the intra lump void ratio in between the points A and B (plastic deformation stage) are assumed be constant. This means that the change in the total void ratios is due to the lumps deforming into inter lump voids. The permeabilities measured in the above three tests are shown in Fig. 5.9. In this figure, it was observed that, in the plastic deformation stage (in between A and B), the change in permeability values is insignificant. It means that the closing of voids are insignificant in this stage of consolidation. The permeability observation supports the assumption made in Fig.5.8 for the plastic deformation stage of consolidation in the lumpy fill. After point ‘B’ (after plasticity deformation stage), both inter and intra voids are changing. In order to find the point C (starting point of the consolidation alone stage), a total of 9 tests using lumps of the above sizes with similar initial conditions for 50kpa, 100 and 200kPa loading pressure were carried. At the end of the loading pressure, water content along the depth of fill was measured by using the standard method and the results are shown in Fig.5.10 From this figure, it is clearly shown that the water content along the depth of fill is reduced with increasing pressure. It is mainly due to the closing of inter lump voids present in the lumpy fill. From permeability measurements, the consolidation pressure at which the third stage started was selected and to this consolidation pressure, the water content was calculated from Fig.5.10 by interpolation method. The void ratio relating to this water content is assumed to be the intra lump void ratio itself for this consolidation pressure. After this point ‘C’ (for consolidation alone stage), the decrease in total void ratio involve mainly decrease in intra lump voids. 124 The variation of intra lump void ratio with total void ratio is shown in Fig.5.8. From the measured value of the total void ratio and assumed variation of intra lump voids, the variation of inter lump voids with consolidation pressure can be calculated using equation 5.3 ( details see in Appendix). es – era = eer (1+era) (5.3) The variation of inter and intra lump void ratios with consolidation pressure for three different size of lump under fully swollen are shown in Fig.5.11. From this figure, it was noted that the variation of intra lump voids with consolidation pressure was close to the oedometer results. Also typical S-shaped curves are obtained, very similar to that for soft homogenous clay from oedometer test. But the variation of the inter lump is distinctly different from that in intra lump void ratio and in homogenous clay. This supports the assumption made for intra lump void ratio variation in the lumpy fill is reasonable good. Similarly, the variation of inter and intra lump voids of a lumpy fill made of 50 mm cubical lumps with three different degree of swelling i.e. zero(0%), partially(50%) and fully swelled(100%) lumps are shown in Fig.5.12. From this figure, it was noted that in the case of zero and partially swelling lumps, the variation of intra lumps voids and the results obtained from oedometer is not close to each other. However, in the case of fully swollen lumps the variation of intra lumps with consolidation pressure matches with those from Oedometer test on homogenous clay. For all the three degree of swelling, the shape of the intra lump void ratio curve follows the shape of the homogenous clay. 125 5.4. Shear resistance In the field, the surcharge load (dumping of sand over dredged clay lumps) is applied in two to three stages. So it is required to study the behaviour of lumpy under the above same stages of loading instead of incremental load. In previous chapter, the consolidation test was carried according to standard loading increment method. In this section, the results of the one stage loading to the required loading pressure instead of gradually increasing pressure are discussed. The shear strength of the lumpy fill were measured using both triaxial and vane shear test methods and the results are now discussed. Fig.5.13 shows the undrained shear strengths determined from the unconfined compression triaxial tests the samples retrieved from the lumpy fill made of 25mm and 50mm cubical lumps with different degree of swelling and pre-consolidation pressure of the lumps. From this figure, it was observed that many of them display a generally bilinear behaviour with the change occurring at a consolidation pressure equal to the preconsolidation pressure of the clay lump. An important factor governing the shear behaviour is the degree of swelling and this is shown in Fig. 5.13(b). For partially and fully swollen lumps the shear strength increases still bilinearly with normal stress but that difference is not so apparent, where as in the case of zero degree of swelling, the variation of shear strength is distinctly bilinear. The size of the lumps, especially for lumps subjected to zero degree of swelling, is of little importance. It can also be observed from Fig. 5.13 that the shear strength varies linearly with normal stress for smaller preconsolidation pressure. 126 Fig.5.14 shows the shear strength profiles obtained using a laboratory vane shear apparatus for the 50 mm cubical lumps with pre-consolidation pressure of 100kPa and 200kPa. It is observed that the shear strength increases with applied loading pressure. The variability in the shear strength along the depth of fill especially for Pc=200kPa is a result of the high degree of variability of a lumpy fill. This shows that even though the lumpy fill appears to be uniform from visual observations, the fill is highly heterogeneous in terms of strength. 5.5. Laboratory versus field behaviour. To be sure that phenomena observed in the laboratory are also reproduced in nature, laboratory studies should be compared with field studies. As part of the Pulau Tekong project carried out by the Housing and Development Board, four pilot test areas were created, TA1, TA2, TA3 and TA4 as shown in Fig.4.30. Among these four test areas, TA2 and TA4 are used to check the effect of size of lumps. In TA2, big clay lumps obtained from sea bed were used, whereas in TA4 the soil obtained from underground construction, which produces lumps of smaller size, were used. With void ratio corresponding to the water content at the liquid limit state as a reference parameter, it has been shown that the generalized soil state parameter reflects the clay fabric conditions rather than void ratio( Nagaraj and Miura, 2001). The generalized soil state parameter, e/eL, is used in analyzing the lumpy fill behaviour under surcharge load. The void ratio data of laboratory studies on 50mm cubical lumps with different degree of swelling from their liquid limit state to different consolidation stress levels were compared with void ratio data obtained from pilot test area TA4 using ND- 127 CPT. In Fig.5.15 the data of e/eL-log σv’ for laboratory and field is presented. When void ratios at different consolidation pressures are normalized by void ratio corresponding to the liquid limit water content of that soil, it can be seen that all the normalized points fall on or above the ICL line under smaller consolidation pressure, and closer to the ICL for higher pressure. This means that the soil is in a normally consolidated state under small loading pressure and this is in agreement with the void index results of 200mm cubical lumps which was discussed in Chapter 4. Lunne et al.(1997) have reported the possibility of identifying the deposit as under consolidated or over consolidated by considering the deviations of the cone penetration resistance path from the linear band with depth, as shown in Fig.5.16a. If the qt profile is left to this line, the soil is regarded as underconsolidated, if it is to the right of this line, the soil is regarded as overconsolidated soil. Fig. 5.16a shows the typical cone penetration resistance obtained from test area TA4, since the cone resistance generally coincides with the linear line reported by Lunne et al.(1997), the soil is regarded as being in a normally consolidated state and this agrees with the results obtained from e/eL concept in the Fig.5.15. Also the variation of pre-consolidation pressure for the test area TA4 and laboratory results with 200mm cubical lumps under 50kPa and 200kPa are shown in Fig.5.16b. In the case of TA4, the pre-consolidation pressures were measured before the sand surcharge was placed over it. From the above figure, it was noted that, the pre-consolidation pressure of the lumpy fill increases with increase in surcharge load pressure. In the previous chapter, the effect of lump size was studied by using 12.5mm, 25mm and 50mm cubical lumps and found that the size has no significant effect on the 128 consolidation behaviour of lumpy fill for above size range of lumps. In the field study, test area TA2 and TA4 data were used to study the effect of size of lumps, as the size of the lumps used were very different in the two test areas. The variation of cone resistance, wet density, water content and void ration variation along the depth of fill at the typical location for the test area TA2 and TA4 are shown in Fig.5.17. From this figure, it was observed that there is a some difference in the magnitude of the above consolidation parameters for the two test areas. However, considering that the degree of swelling is likely to be also different, this suggests that the size of the lumps may not play significant effect on the consolidation characteristics of lumpy fill. 5. 6. Summary In this chapter, experimental results as discussed in the Chapter 4 were used to study the possible structural mechanism developed during consolidation in the lumpy fill using the Olsen (1962) cluster model theory. Results from the interpretation of laboratory experiments shows that there are three likely stages of consolidation in a lumpy fill system. It was found that the compressibility of individual lumps is small at initial loading pressure, so compression is accompanied by reduction in inter lump voids, but with little change in intra lump void ratio. From the shear strength measurements, it is noted that the pre-consolidation pressure and size of lumps has little influence on the compressibility of lumpy fill. In order to make sure that the observations from small scale laboratory experimental results were applicable to the filed, the results were compared with some geotechnical parameters obtained from field test in pilot test areas TA2 and TA4 in 129 Stage-1 of the Pulau Tekong reclamation project where different lumps sizes were used for the fill. From this study, a reasonable agreement between laboratory and field behaviour of the lumpy fill was observed. 130 Fig.5.1 Cluster model for permeability prediction (after Olsen,1962) Fig. 5.2 Assumed relationship between the total, cluster, and inter-cluster void ratios (after Olsen, 1962) 131 4.0 a) 12.5 mm Void ratios, eiv and el 3.5 Assumed points at which Lumps begins to compress 3.0 2.5 2.0 eiv 1.5 1.0 el y = 0.7558x - 3E-16 0.5 0.0 0.0 1.0 2.0 3.0 4.0 Total void ratio, es Void ratios, eiv and el 4.0 b) 25 mm 3.5 Assumed points at which Lumps begins to compress 3.0 2.5 eiv 2.0 1.5 1.0 y = 0.7598x 0.5 el 0.0 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 Total void ratio, es Void ratios, eiv and el 4.0 c) 50 mm 3.5 Assumed points at which Lumps begins to compress 3.0 2.5 eiv 2.0 1.5 1.0 y = 0.7486x 0.5 el 0.0 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 Total void ratio, es Fig.5.3 Assumed relationship between the total, inter and intra lump void ratios a) 12.5 mm b) 25 mm c) 50 mm 132 Ratio of Estimated flow rates in inter to intra lump void ratio 100 12.5 mm 25 mm 50 mm 10 1 0.1 1 10 100 1000 Consolidation pressure, kPa Ratio of estimated flow rates in inter to intra lump void ratio Fig. 5.4 Permeability discrepancies according to the cluster model for different size of lumps. 100 10 1 Us = 0% Us = 50% Us= 100% 0.1 1 10 100 1000 Consolidation pressure, kPa Fig. 5.5 Permeability discrepancies according to the cluster model for different degree of swelling. 133 1.0E-04 Permeability,k( m/s) S-1 A 1.0E-05 B 1.0E-06 1.0E-07 S-2 1.0E-08 1.0E-09 C S-3 D 1.0E-10 1.0E-11 1 10 100 1000 Consolidation pressure, kPa Void ratios, inter and intra lump void ratio Fig.5.6 Typical Permeability variation in the lumpy fill under consolidation pressure 4.0 3.5 es-era = eer×(1+era) 3.0 Plasticity + Consolidation 2.5 Plasticity 2.0 Consolidation 1.5 1.0 Variation of intra Lump void ratio 0.5 0.0 A B B C B D B 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 Total void ratio, es Fig.5.7 Modified assumed relationship between the total, inter and intra lump void ratios. 134 4.0 3.5 12.5 mm Void ratio 3.0 2.5 es-era = eer×(1+era) 2.0 B A C 1.5 y = 0.3584x + 0.9314 1.0 y = 0.9236x 0.5 D 0.0 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 Total void ratio, es 4.0 25 mm 3.5 Void ratio 3.0 2.5 es-era = eer×(1+era) 2.0 B C 1.5 A y = 0.4879x + 0.675 1.0 0.5 y = 0.897x D 0.0 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 Total void ratio, es 4.0 50 mm 3.5 Void ratio 3.0 2.5 es-era = eer×(1+era) 2.0 B 1.0 0.5 0.0 A C 1.5 y = 0.4096x + 0.7846 y = 0.9x + 3E-16 D 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 Total void ratio, es Fig.5.8 Assumed relationship between total, inter and intra lump void ratios Note: AB Plastic deformation, BC Plastic deformation + Consolidation, CD Consolidation 135 1.0E-04 A B Permeability, k(m/s) 1.0E-05 B 12.5mm/2.86/1.56/1.30 B 25mm/3.11/1.75/1.36 1.0E-06 50mm/3.14/1.83/1.31 1.0E-07 1.0E-08 C 1.0E-09 C D C 1.0E-10 D 1.0E-11 1 10 100 1000 Consolidation pressure, s' v (kPa) Fig.5.9. Permeability of lumpy fill made of 12.5mm, 25mm and 50mm cubical Lumps with 100% degree of swelling. 136 Water content water content 0.45 0.5 0.55 0.6 0.4 0 2 50 kPa 100kPa 200 kPa 0.55 0.6 0.4 2 8 0.5 0.55 0.6 50kPa 100kPa 100kPa 2 200kPa 4 6 0.45 0 50 kPa Depth, cm Depth, cm 8 0.5 0 4 6 0.45 Water content 200kPa 4 Depth, cm 0.4 6 8 10 10 10 12 12 12 14 14 14 Fig.5.10. Water content variation along the depth of lumpy fill for 100% swelling under different consolidation pressure for (a) 12.5mm (b) 25mm and (c) 50mm 137 2.5 12.5 mm Void ratio 2.0 1.5 1.0 e ra eer oedo 0.5 0.0 1 10 100 1000 Consolidation Pressure, kPa 2.5 25 mm Void ratio 2.0 1.5 1.0 era eer Oedo 0.5 0.0 1 10 100 1000 Consolidation Pressure, kPa 2.5 2.0 Void ratio era eer oedo 50 mm 1.5 1.0 0.5 0.0 1 10 100 1000 Consolidation pressure, kPa Fig.5.11. Variation of inter and intra lump void ratio with consolidation pressure for lumpy fill made of 12.5mm, 25mm and 50 mm cubical lumps. 138 2.5 US = 0% Void ratio 2.0 1.5 1.0 era e er 0.5 oedo 0.0 1 10 100 1000 Consolidation Pressure, kPa 2.5 US = 50% Void ratio 2.0 1.5 1.0 era 0.5 eer Oedo 0.0 1 10 100 1000 Consolidation Pressure, kPa 2.5 2.0 Void ratio era eer oedo US = 100% 1.5 1.0 0.5 0.0 1 10 100 1000 Consolidation pressure, kPa Fig.5.12. Variation of inter and intra lump void ratio with consolidation pressure for lumpy fill made of 50 mm cubical lumps with 0%, 50% and 100% degree of swelling. 139 Shear strength,kPa 70 60 (a) 50 40 30 20 10 Pc =200kpa,25 mm, 0% Pc=200kpa,50 mm, 0% 0 0 100 200 300 400 500 Normal stress,kPa 70 (b) Shear strength, kPa 60 50 40 30 Pc=200,50mm,0% 20 Pc=200kpa,50mm,50% 10 Pc=200kpa,50mm,100% 0 0 100 200 300 400 500 Normal stress, kPa Shear strength, kPa 70 (c) 60 50 40 30 pc=100kPa,50mm,0% 20 pc=200kPa,50mm,0% 10 0 0 100 200 300 400 500 Normal stress, kPa Fig.5.13. Triaxial shear strength for (a) Different size of lump, (b) Different degree of swelling and (c) Different pre-consolidation pressure lumps 140 Shear strength, kPa 0 20 Shear Strength,kPa 40 60 0 0 20 40 80 100 120 0 (b) Pc = 200kPa (a) Pc=100kPa 2 2 4 4 6 6 Depth ,cm Depth, cm 60 8 8 10 10 12 12 0.5Pc 14 1.0Pc 2.0Pc 16 0.5Pc 14 1.0Pc 2.0Pc 16 Fig.5.14. Vane shear strength for lumpy fill made up of 50 mm cubical lumps with 0% degree of swelling for (a) Pc =100 kPa and (b) Pc=200 kPa. 1.8 1.6 1.4 e/eL 1.2 1.0 NC 0.8 0.6 50mm/100% 50mm/50% 0.4 50mm/0% TA4-1-1 0.2 ICL OC 0.0 1 10 100 1000 Consolidation pressure, kPa Fig.5.15 Generalization of compression paths for laboratory and field data 141 Corrected cone resistance, qt(MPa) Preconsolidation pressure, kPa 0 0 0 2 0.2 4 0.4 6 8 10 12 q = 2.75σ'v0 (corresponding to NC) Depth ratio, d/D Depath (m) 0.00 0.25 0.50 0.75 1.00 1.25 1.50 100 200 300 400 0.6 0.8 1 1.2 1.4 TA4-1-1 14 1.6 16 1.8 200mm/50kPa 200mm/200kPa Fig.5.16 (a) Typical cone resistance for Test area TA4. (b) Comparison of pre-consolidation pressure measured by the ND-CPT in TA4 with the laboratory experiment results using 200mm cubical lumps. 142 Wet density(KN/m ) Corrected cone resistance, qt(MPa) 0.0 0.5 1.0 1.5 0 Depth(m) 2 2.0 12 14 16 18 0 TA2-1-46 TA2-1-46 2 TA4-1-18 Void ratio, e 20 0.0 2.0 3.0 4.0 5.0 0 TA4-1-18 2 4 4 6 6 6 8 8 10 10 12 12 14 14 8 1.0 4 10 12 14 16 TA2-1-46 TA4-1-18 18 16 16 Fig.5.17. Typical cone resistance, wet density and void ratio profiles at location TA2-1-46 and TA4-1-18. 143 CHAPTER 6 Summary and Conclusions 6.1 Summary of Work In Singapore, large volume of clay lumps are obtained from underground construction activities like deep basements for buildings, excavations and tunneling, etc. Further, dredging is frequently done for maintenance of navigation channels and port construction activities. In addition, large quantities of dredged clay lumps are also obtained from excavation done in the seabed for the construction of sand keys and sand bunds along the periphery of the land reclamation projects. The disposal of the clay lumps is an almost intractable problem in Singapore which has no landfill sites since the mid-1990s. One ideal solution is to be able to use these unwanted soils as an alternate fill for the reclamation (Hartlen and Ingers, 1981 and Karunaratne et al., 1991). The use of dredged materials for land reclamation is an attractive proposition because it will help resolve an almost intractable environmental problem associated with the disposal of these materials and provide an alternate fill for land reclamation in land-scarce Singapore. When these big dredged clay lumps are used for reclamation, the reclaimed ground initially will have large inter lump voids and these inter-lump voids (void between lumps) could be partly filled with small lumps or clay slurry. The current research work focused on how the inter and intra lump voids were closing with surcharge load by using experimental results with lumps of size varying from 12.5mm to 200mm. A normalized 144 parameter called void index, Olsen (1962) cluster model, pore water and shear strength measurements were various parameters used to evaluate the above issue. Detailed conclusions were presented at the end of each chapter and a general summary is given here. 6.2 Conclusions The following conclusions can be drawn from the preceding discussions in Chapter 4 and Chapter 5. It is found that the results obtained from consolidation experiments are very useful to provide the kind of results needed to use the intrinsic frameworks proposed by Burland (1990) to capture the compressibility characteristics of the natural deposit. This intrinsic framework, with a little modification (discussed in chapter 4), is also capable of describing the compressibility characteristics of the lumpy fill layer. The effect of degree of swelling on consolidation behaviour of lumpy fill was investigated using three different degree of swelling 0% (immediately after filling), 50% (after some time) and 100% (if left alone for a long period). Based on the results, it was noted that the degree of swelling has a strong influence on the compressibility behaviour of the lumpy fill. The surcharge pressure required to close the inter lump voids of fully swollen lumps is considerably smaller than that required for lumps that were subjected to partial swelling. The influence of size of lumps on the compressibility behaviour was investigated by using 12.5mm, 25mm and 50mm cubical lumps. The results showed that there 145 is no significant effect of size on the compressibility behaviour of the lumpy fill system within the range of lump sizes investigated, when the lumps were subjected to 100% swelling. The influence of initial packing arrangement on compressibility of the lumpy fill was investigated by using 50mm size of the lump with inter lump void ratio in the range of 1.41 to 2.25. The results showed that the packing arrangement of the lumps has significant influence on the consolidation behavoiur of lumpy fill. For consolidation pressure less than 50kPa, the shape of the time-compression curve in the logarithmic plot is concave, in marked contrast to those observed for homogenous clays, following the classic Terzaghi’s model. Beyond a consolidation pressure of 50kPa, the rate of settlement is much slower due to the closure of the inter-lump voids and the shape of the curve started changing. Typical S-shaped curves are obtained in the logarithmic plot for pressure beyond 100kPa. It is noted that for lumpy fill made up of spherical lumps experienced less settlement strain as compared to the cubical lumps due to its denser packing. It is also observed that the surcharge pressure required to close the inter lump voids is less in the case of lumpy fill made of spherical lumps. For Us = 0%, pre-consolidation pressure will significantly affect the initial consolidation behavoiur of the lumpy fill system. However, the effect of preconsolidation pressure is not significant for the other two degrees of swelling investigated. 146 It was observed that the rate of settlement for the lumpy fill system the inter-lump voids filled with slurry is considerably slower than that filled with water. Also it was noted that, the inter-lump voids filled with slurry in the lumpy fill is not significantly different from that with water for higher consolidation pressure. From the evaluation of shear strength using the cone penetration test indicates that the fill is highly heterogeneous under a pressure of 100kPa. When the consolidation pressure is equal to or greater than the pre-consolidation pressure of the lumps, the strength profile is uniform. The coefficient of secondary compression of the lumpy fill is comparable to the homogeneous clay indicating that inter-lump voids were effectively closed and secondary compression is no more significant than for the undisturbed clay. The strength and deformation characteristics of the lumpy fill layer were found to be variable with normally consolidated and over consolidated zones. The normally consolidated zones are due to existence of disintegrated material in the initial inter-lump and the over consolidated zones could be part of the original lumps. In order to make sure that the observation seen in the small scale laboratory experimental results were compared with few geotechnical properties of field at typical locations of test areas TA2 and TA4 in stage-1. From this study, it was observed a significant agreement between laboratory and filed behaviour of lumpy fill. 147 REFERENCES Al-Tabbaa, A.(1995), “Excess pore pressure during consolidation and swelling with radial drainage”, Geotechnique, Vol.45, No.4, pp.701-707. ASTM D 2435.(1989), “Standard test method for one-dimensional consolidation properties of soils”. Annual Book of ASTM standards, Vol. 04-08, pp. 283-287. Bo, M. W., Bawajee, R. and Choa, V.(2001), “Reclamation using dredged materials”. Proceedings of the international conference on Port and Maritime R & D and Technology, Singapore, Vol.1, pp.455 – 461. Bjerrum, L.(1967), “Progressive failure in slopes of over consolidated plastic clay and clay shales”, ASCE, Vol.93, SM5, pp.1-49. British Standards Institution (1990), “Methods of test for soils for civil engineering purposes: Compressibility, Permeability and Durability tests”, BS1377: Part 5, Milton Keynes:BSI. Burland, J.B. (1990), “On the compressibility of and shear strength of natural clays”, Geotechnique, Vol.40, No.3, pp.329-387. Casagrande, A.(1949), “Soil Mechanics in the Design and Construction of the Logan Airport”, Journal of the Bostan Society of Civil Engineers, re-printed in Contributions to Soil Mechanics, 1941-1953, BSCE(1953), pp.176-205. Choa, V. and Bo, M. W. (2001), “Evaluation of shear strength parameters of Singapore marine clay”. Proceedings of Underground Singapore 2001, Singapore, pp.315-322. Chong, Y.L., Leung, C.F., Tan, S.A., Lim, T.H. and Chua, S.G.(1998), “Performanc of Dredged Clay as Reclamation Fill”, Thirteenth Southeast Geotechnical Conference,16-20 November, Taipei, Taiwan, Roc, pp.763-768. David Muir Wood. (1990), “Soil Behaviour and Critical State Soil Mechanics”, Cambridge University Press 1990. Dobbie, M.J. A study of cone penetration tests in the Singapore marine clay, in Proceeding of the 1st International Symposium (ISOPT-1) on penetration Testing, Orlando, Fla., 20-24 March 1988. A.A. Balkema, Rotterdam, The Netherlands, pp.737744. Feda, J. (1998),“Fragmentary clay-a difficult waste material”, Engineering Geology, vol.1, pp.77-88. 148 Hartlen, J. and Ingers, C. (1981), “Land Reclamation using fine-grained dredged material”.Proceedings of the tenth international conference on Soil Mechanics and Foundation Engineering, Stockholm, Vol.1, pp.145 – 148. Hee, A.M., Lim, T.H., Chong, Y.L., Leung, C.F. and Tan, S.A. (1999), “Tuas Reclamation Project: A case study, Field Measurmentss in Geomechanics”, Leung, Tan and Phoon (eds), Ballkema, Rotterdam, ISBN 90 5809 0663, PP.467-473. Yang Jun Wei (2003),” Consolidation charactistics of lumpy fill”, M.Engg. Thesis, National University of Singapore, Singapore. Karthikeyan, M., Dasari, G. R. and Tan, T. S. (2004), “In-situ characterization of a land reclaimed using big clay lumps”, Canadian Geotechnical Journal, Vol.41, pp.242256. Karthikeyan, M. Application of radioisotope cone penetrometer to characterize a lumpy fill. Ph.D thesis, National University of Singapore, Singapore.2005. Karunaratne, G. P., K. Y. Yong, T. S. Tan, S. A. Tan, K. M. Liang, S. L. Lee, and A.Vijiaratnam.(1990), “Layered clay-sand scheme reclamation at Changi South Bay”, Proceedings of 10th Southeast Asian Geotechnical Conference, 16-20 April, Taipei, Vol.1, pp.71-76. Karunaratne, G. P., K. Y. Yong, T. S. Tan, S. A. Tan, S. L. Lee, and A.Vijiaratnam.(1991), “Land Reclamation using Layered Clay-Sand Scheme”, In proceedings of the International Conference on Geotechnical Engineering for Coastal Development: Theory and Practice on Soft Ground, Yokohama, Japan, pp.335-340. Khoo, E., Okumura, T and Lee, F.H.(1994), “Side friction effects in plane strain models”, Proc.of the International Conference Centrifuge 94, Singapore, pp.115-120. Ladd, C. C., Foott, R., Ishihara, K., Schlosser, F. and Poulose, H. G.(1977),“Stressdeformation and strength characteristics”, SOA report. Proceedings, 9th International Conference on Soil Mechanics and Foundation Engineering, Tokyo, Vol.2, pp.421-494. Lee, S.L., Karunaratne, G.P., Lo, K.W., Yong, K.Y. and Choa, V.(1985), “Development in Soft Ground Engineering in Singapore”, In Proceedings of the Eleventh International Conference on Soil Mechanics and Foundation Engineering, San Francisco, USA, Vol.5, pp.1661-1666. Lee, S.L., Karunaratne, G.P., Yong, K.Y. and Ganeshan, V. (1987) “Layered ClaySand Scheme of Land Reclamation”, Journal of Geotechnical Engineering Division, ASCE, 113, PP.984-995. Lee, S.L., Karunaratne, G.P., Yong, K.Y., Chow, Y.K. and Chew, S.H.(1988), “Consolidation of Dredged Clay in Reclamation”, Soils and Foundations, Vol.28, pp.12-25. 149 Leung, C. F., Wong, J. C., Manivanann, R. and Tan, S. A. (2001), “Experimental evaluationof consolidation behaviour of stiff clay lumps in reclamation fill”,Geotechnical Testing Journal, GTJODJ, Vol.24, No.2, pp.145 - 156. Lianhe Zaobao. (2000), Press report on 20 August 2000. Singapore press Holdings, Singapore (In Chinese). Lord, E.R.F. and B.B.A.Isaac.(1988), “Geotechnical Investigation of Dredged Onverburden at the Syncrude Oil Sand Mine in Northern Alberta, Canada”, Canadian Geotechnical Journal, Vol.26, pp.132-153. Lui, P. C. and Tan, T. S. (2001), “Building Integrated large-Scale Urban Infrastructures: Singapore’s Experience”, Journal of Urban Technology, Vol.8, No.1, pp.49-68. Lunne, T., Robertson, P. K. and Powell, J. J. M. (1997), “Cone penetration testing in geotechnical practice”, Blackie Academic & Professional, London. Manivannan, R., Wong, J. C., Leung, C. F. and Tan, S. A. (1997), “Consolidation characteristics of lumpy fill”, Centrifuge 98, Balkema, Rotterdam, pp.889-894. Mendoza, M. J. and Hartlen, J. (1985), “Compressibility of clayey soils used in land reclamation”, Proceedings of the Eleventh International Conference on Soil Mechanics and Foundation Engineering, San Francisco, Vol.2, pp.583-586. Metchell, J.K. (1997), “Fundamentals of soil behavior”, second edition, John Wiley & Sons, Inc. Mesri, G. and Godlewski, P. M. (1977), “Time and stress compressibility interrelationship”, Journal of Geotechnical Engineering Division, ASCE, Vol.103, pp.417-430. Mesri, G. and Castro, A.(1987), “Cα/Cα concept and Ko during secondary compression”, Journal of Geotechnical Engineering, ASCE, Vol.103(GT-3), pp.230-247. Ministry of Information and Communication. stretched, he Mirror, 22 (3), pp.16.1989. Land Reclamation: Our resources Nagaraj, T.S. and Miura, N. (1990), “Soft clay behaviour analysis and assessment”, A.A.Balkema/Rotterdam/Brookfiels/2001. Nishimura, S., Fujii, H. and Wakitani, Y.(2003), “Consolidation Properties of Sludge Cakes”, Proceedings of the Twelfth Asian Regional Conference on Soil Mechanics and Geotechnical Engineering, Singapore, Vol.1, pp.183-186. Nogami, J.L., Wang Weidong and Li Maoxing. (2001), “Consolidation of Lumpy Clay Fills”, In proceedings of the Inaugural International Conference on Port and Maritime 150 R&D and Technology, ed by N.Jothisankar, Toh Ah Cheong and Lin Pengzhi, October 2001, Singapore, pp.483-489. Olsen, H.W.(1962), “Hydraulic flow through saturated clay”, Proceedings of the Ninth National Conference on clays and clay minerals, pp.131-161. Robinson, R. G., Dasari, G. R. and Tan, T. S. (2004), “Three-dimensional swelling of clay lumps”, Géotechnique, Vol.54, No.1, pp.29-39. Robinson, R. G., Tan, T. S., Dasari, G. R. and Phoon, K.K. (2003), “Comparative Study on the Compressibility of Lumpy Fill in Water and in Clay Slurry for Land Reclamation”, Proceedings of the Twelfth Asian Regional Conference on Soil Mechanics and Geotechnical Engineering, Singapore, Vol.1, pp.233-236. Tan, T. S., Leung, C. F. and Yong, K. Y. (2000), “Coastal reclamation projects in Singapore”, In Proceedings, Coastal Geotechnical Engineering in Practice, Vol.2, ISYokohama, Japan, Edited by A.Nakase and T.Tsuchida, pp.119-125. Tan, T. S., Phoon, K. K., Lee, F. H., Tanaka, H., Locat, J. and Chong, P. T. (2002), “A Characterization Study of Singapore Lower Marine Clay”, Characterization and Engineering Properties of Natural Soils (eds Tan, T. S., Phoon, K. K., Hight, D. W. and Leroueil, S.), Singapore, pp.429-454. Tanaka, H., J.Locat, S. Shibuya, T.S.Tan. and D.R.Shiwakoti. Characterization of Singapore, Bangkok, and Aariake clays, Canadian Geotechnical Journal, 38, pp.378400.2001. Terzaghi, K.(1931), “The influence of elasticity and permeability on the swelling of two phase system”, Colloid Chemistry, ed.J.Alexander, Vol.3, Chemical Catalog co., New York, pp. 65-88. The Strait Times. Singapore Grow Bigger, Press report on 28th August 2000. Singapore press Holdings, Singapore. Whitman, R.V. (1970), “Hydraulic Fills to Support Structural Loads”, Journal of the Soil Mechanics and Foundations Divisions, ASCE, Vol.96, pp.23-47. Wong, .J.C. (1997), “Model studies of lumpy fill”, M.Engg. Thesis, National University of Singapore, Singapore. Yang, L. A., Tan, T. S., Tan, S. A. and Leung, C. F. (2002), “One-dimensional selfweight consolidation of a lumpy clay fill”, Géotechnique, Vol.52, No.10, pp.713-725. Yang, L. A.and Tan, T. S.(2004), “One-dimensional of a lumpy clay with non-linear properties”, Géotechnique, Vol.61, No.8, pp.511-525. Yap, C.H. (2001), “New container terminal case study – performance of lumpy fill”, B.Engg. thesis, National University of Singapore, Singapore. 151 APPENDIX Inter-relationship between inter, intra and system void ratios: VT VS Vra VL Ver es eer era = = = = = = = = Total volume Volume of solids Volume of Intra-lump voids Volume of lumps = VS + Vra Volume of inter-lump voids = VT - VL System void ratio Inter-lump void ratio Intra-lump void ratio System Void ratio, es = VT − VS VS Inter lump void ratio, eer = VT − VL = VL eS – eer = VT − VS VS = VraVT VS (VS + Vra ) = era × es - eer es - era VT VL VT − (VS + Vra ) (VS + Vra ) VT − (VS + Vra ) (VS + Vra ) = V − V L V L  era  T +  VL   VL = era (eer+ 1) = eer (1+era) 152 [...]... 2001) The use of large clay lumps in a major way is now being implemented in reclamation at the eastern coast of 9 Singapore Island and the present research is an integrated part of that project The literature review relating to the consolidation behaviour of land reclamation using lumpy clay fill is discussed in the following sections 2.2 Experimental studies on lumpy clay fill in land reclamation Casagrande... size of lump, degree of swelling, pre-consolidation pressure of lumps and packing arrangement were studied using modified burland frame work The burland intrinsic frame works were discussed in detail in the next page 2.2.1 Void Index and Intrinsic Compression Line In the characterisation of soil structure in terms of compressibility, Burland(1990) referred to the intrinsic compression curve (ICC) obtained... curve of lumpy fill made up of 50mm cubical lumps under consolidation pressure of 50 kPa 95 Fig.4.22 Variation of voids index with consolidation pressure for different degree of swelling for lumpy fill made of 50mm size cubical lumps 96 Fig.4.23 Variation of consolidation pressure required to close the inter-lump voids of lumpy fill with degree of swelling 96 Fig.4.24 Photographs of lumpy fill made up of. .. presence of clay slurry in the inter-lump voids of a lumpy fill Comparison was made with the results obtained on lumpy fill whose inter-lump voids are filled with water 12 Consolidation tests conducted in the laboratory on lumpy fill made of cubical lumps of 25 mm suggest that the presence of slurry in the inter-lump voids has significant influence on the consolidation behaviour during the early stages of. ..LIST OF FIGURES Figure No Title Page No Fig.1.1 Original Land Profile versus exiting and future reclamation Profiles (After The Straits Times, 2000) 6 Fig.1.2 Fill materials used in Singapore land reclamation projects (after the Ministry of information and communication, 1989) 6 Fig 1.3 Dredging of seabed using clamshell grab 7 Fig.1.4 Dredged lumps placed in a barge 7 Fig.1.5 Dumping of clay lumps... out using centrifuge modeling technique to evaluate the effect of lump size and shape and original in- situ shear strength of clay on the performance of lumpy fills From this study, it was observed that the lumpy fill made up of spherical lumps settles less than the irregular and cubical lumpy fill due to the denser packing of spherical lumps in the lumpy fill (Fig 2.4) As the size of the lumps increases,... area of 662 square kilometers today, an increase of 17 percent (Lui and Tan, 2001) Land reclamation has become an integral part of Singapore’s development In the beginning, hill-cut materials and sand deposits in the surrounding sea-bed have been used as fill materials Fig.1.2 shows the fill material used in previous Singapore land reclamation projects (Ministry of Information and Communication, 1989)... that the rate of softening decreases with increasing lump size and increases with increases of surface area being exposed to surrounding water Also, Robinson et al (2004) investigated the swelling of clay lumps with free lateral boundary with all round drainage To examine this aspect a series of experiments was carried out in which reconstituted cylindrical clay lumps of 105, 205 and 400 mm in diameter... finding suitable dumping ground for the disposal of these materials, as well as creating new land for land- scarce Singapore 1.2 Land reclamation in Singapore using Lumpy fill In the 1500 hectares propose Pulau Tekong reclamation, clay lumps from both seabed dredging and from construction activities are used as fill materials The dredging of in- situ clay from the seabed is carried out using large clamshell... swelling of clay lumps, packing of lumps and size of lumps An understanding of the behaviour of lumpy fill and the influence of these factors at the ultimate state would provide significant opportunity to further optimize the design 1.3 Objectives of this research In connection with the Pulau Tekong reclamation project using lumpy clay, a study of settlement and consolidation behaviour of such lumpy ... problem of finding suitable dumping ground for the disposal of these materials, as well as creating new land for land- scarce Singapore 1.2 Land reclamation in Singapore using Lumpy fill In the.. .CHARACTERIZATION OF LUMPY FILL IN LAND RECLAMATION VIJAYAKUMAR ALAPAKAM B.Tech (SVU), M.S.(IIT Madras) A THESIS SUBMITTED FOR THE DEGREE OF MASTER OF ENGINEERING DEPARTMENT OF CIVIL ENGINEERING... ascertained For subsequent engineering design purpose, the engineering properties of the fill after stabilization also need to be established An understanding of the behavior of lumpy fill and the influence