DEVELOPMENT OF A TIMBER RAFT & PILE...

DEVELOPMENT OF A TIMBER RAFT & PILE

FOUNDATION FOR EMBANKMENTS

ON SOFT GROUND

September, 2009

Division of Engineering Systems and Technology Graduate School of Science and Engineering

Saga University, Japan

PONGSAGORN POUNGCHOMPU

DEVELOPMENT OF A TIMBER RAFT & PILE FOUNDATION FOR EMBANKMENTS ON SOFT GROUND

A Dissertation submitted to the division of engineering systems and technology, Graduate school of science and engineering, Saga University in partial fulfillment of the requirements for the degree of Doctor of Philosophy in Civil Engineering

by

PONGSAGORN POUNGCHOMPU

Nationality Thai Previous degree Bachelor of Engineering Khon Kaen University, Khon Kaen, Thailand Master of Engineering Saga University, Japan

Division of Engineering Systems and Technology Graduate school of Science and Engineering

Saga University JAPAN

September, 2009

Examination Committee

Professor Shigenori HAYASHI (Chairman) Institute of Lowland Technology, Saga University, Saga, Japan

Professor Jin-Chun CHAI Department of Civil Engineering, Faculty of Science and Engineering, Saga University,

Saga, Japan

Associate Professor Akira SAKAI Department of Civil Engineering, Faculty of Science and Engineering, Saga University,

Saga, Japan

Associate Professor Daisuke SUETSUGU Institute of Lowland Technology, Saga University, Saga, Japan

Associate Professor Marolo C. ALFARO Department of Civil Engineering, Faculty of Engineering, Manitoba University,

Manitoba, Canada

ABSTRACT

In recent years, the construction has been under pressure to operate with greater finesse and accuracy. There were two main reasons for this. Firstly, with the economy becoming more mature and environmental issue gaining importance, the constraints on construction were gradually increasing. Secondly, it was recognized that effects on the environment must be minimize to avoid pollution problem. This pressure accelerates the advance of technology in the construction field.

For a road embankment constructed on soft subsoil such as Ariake clay, large

settlement and lateral deformation will occur. These deformations will be transferred to the surrounding ground surface and subsoil and will damage the adjoining buildings and agricultural lands along the highways. Moreover, among the various techniques for reinforcement embankment on soft ground. Most of reinforcement embankment methods use the artificial geosynthetics as the reinforcement and/or chemical admixture as additive to improve the soil strength. The use of the artificial material for long term for improving the ground may work out expensive and may create environmental problem. Timber Raft & Pile method is considered as one of the most versatile, cost effective and environmental friendly methods, which will not pollute the environment when compared to other methods, in which the artificial geometries are used.

In order to evaluate the effectiveness of the Raft & Pile foundation on the basis

consideration of both stability and deformation of clay foundations. The laboratory model tests are conducted, in which width of raft and length of pile and the connecting manner with tie rod for pile line. Simple index for evaluate effectiveness of Raft & Pile is purpose. The result show that effective of Raft & Pile foundation can be reduced both settlement at embankment canter and heave beyond toe of embankment significantly. The results of the model test need to be validated by full-scale tests in the field.

The scale model of a new pattern of raft that consisted of timber in circle shape diameter 6mm and composed together by using screw, in order to mirror the true physical flexural behavior of an original phenomenon, or prototype level. In total twenty test cases of tree-point bending test on composites Raft were conducted to investigated firstly, the effect of No. layer of raft and effect of pattern of axial raft element, secondly, to evaluate effectiveness of Raft composites pattern support to design criteria, thirdly, to links between the composites short timber Raft (simulated

ABSTRACT

iv

field condition) and composites long timber Raft (simulated FEM analysis). Based on the result, composites raft very flexural response that can be use as flexible foundation support road embankment, number of raft layer and No. of axial raft element were significant on rigidity of composites raft for construction. Links between the composites short timber Raft (simulated field condition) and composites long timber Raft (simulated FEM analysis) and relation of EI between design and FEM analysis in term of layer Raft are purposed.

To verify the effectiveness of Raft & Pile foundation that depending on the condition of ground and loading in each situation of road embankment construction. Finite element analysis is one of the useful approaches to make clear the effects with various analysis conditions. FEM simulation embankment height 5.5m constructed over soft Ariake clay ground using a computer code PlaxisV.8. One most important aspect of design of Raft & Pile foundation, it is required to estimate a residual settlement of the road surface. Based on the result of FEM analysis, the effective of timber Raft & Pile can be reducing both settlements at embankment center and heave beyond toe of embankment, effective of timber Raft & Pile can be reducing development of displacement and excess pore pressure in the subsoil with time response, also clearly that embankment load was carried by raft, the contribution of the piles was found to decreases at large displacements, Raft only can be reducing differential settlement beneath embankment more effective with pile in reducing differential settlement at transition zone and heave beyond toe of embankment and investigated Raft & Pile can be support design criteria such as necessary layer of raft, optimum pile length, thickness of corrected filled material and differential settlement inside and outside toe of embankment.

List of figures

ACKNOWLEDGEMENTS

The author would like to express his deep sense of gratitude and indebtedness to his supervisor, Professor Shigenori Hayashi of the Institute of Lowland Technology, Saga University, Japan for his constant supervision, continuous guidance, helpful criticism, suggestions and encouragement given throughout the course of this research. Even with his tight and hectic schedules, he always gave the author his time and the support every time the author needed. His high appreciations, positive attitudes and generous comments about the author’s research ability, played an important role in the outcome of this dissertation.

Profound gratitude is expressed to the members of the examination committee of

this dissertation, Professor Jin-Chun Chai, Department of Civil Engineering, Saga University, Japan Associate Professor Akira Sakai, Department of Civil Engineering, Saga University, Japan, Associate Professor Daisuke Suetsugu, Institute of Lowland Technology, Saga University, Japan and Associate Professor Marolo C. Alfaro, Department of Civil Engineering, Manitoba University, Canada for their support and valuable suggestions and comments.

Sincere thanks are also extended to Prof. D.T. Bergado, Asian Institute of Technology, Thailand, Prof. A. Sridharan, Formerly Professor, Indian Institute of Science, India, Prof. S. L. Shen, Shanghai Jiao Tong University, China, Prof. J.T. Shahu, Indian Institute of Technology, India for their cordial and sincere efforts to teach the author the fundamental soil behavior and many geotechnical issues during stay in Institute of Lowland Technology, Saga University as Visiting Professor.

Special thanks to Prof. Y.J. Du, Institute of Lowland Technology, Saga University, Japan for guidance me with the beginning experiment and data collection and giving a favorable information and advantageous data.

Thank are also extended to the member of Civil Engineering Department, Saga University and Institute of Lowland Technology for the use of the facilities in their laboratories.

Appreciation is also extended to formerly students of Saga University especially Dr. X. J. Chai, Dr. K. Sinat, Dr. M. A. Moqsud, Dr. H. Tri, Dr. I. Patcharaporn, Dr. J.

ACKNOWLEDGEMENTS

vi

Teeraporn and thank are also due to formerly visiting reseach student from Asian Institute of Technology, Thailand , Dr. T. Tawatchai and Mr. S. Jaturong for their laboratory assistance and their friendship.

Appreciation is also due to all the former and present members of Hayashi sensei’ Laboratory, Hino sensei’ Laboratory and Chai sensei’ Laboratory for their friendly behaviors, unconditional help and encouragement provided to the author in various aspects from the beginning of his living in Saga and in his research work. The facilities obtained from Saga University and the encouragement received from the faculty members during the course of this study is gratefully acknowledged. The author wishes to appreciate very much the encouragement and warmth social gathering rendered by all the former and present Thai students who lived and are living in Saga.

The author wishes to place on record his sincere thanks to Mrs.Komori Ikuko and My owner apartment “Tsuruya Manshion” for their sacrifice, supported my family since arriving in Japan.

The most gratefulness to my magnificent parents and Dr. Suriya Koosirivichian for always been supporting far from Japan. Special thanks go to my elder sister, Dr. Ornanong Poungchompu for all helps and encouragements in my life and study. I would like to thank my wife, Dr. Supaporn Poungchompu for her wisdom in understanding the situation and her very well tolerance everything and to my two daughters, for giving a cheerful and heavenly atmosphere at home and making me to be more powerful in my life.

List of figures

vii

CONTENTS ABSTRACT iii ACKNOWLEDGEMENTS v

LIST OF TABLES xii

LIST OF FIGURES xiii

CHAPTER Ⅰ

INTRODUCTION

1.1. General 1

1.2. Research Background 2

1.2.1. Ariake sea coast road embankment 2

1.2.2. Strategy of wood use as construction material in Ariake sea coast region 3

1.2.3. Safety-high durability and environment friendly of timber Foundation

on soft ground 3

1.2.4. Develop a non-polluting construction technology to minimize

environment impact 6

1.3. Objective of this research 11

1.4. Scope of this dissertation 12

CHAPTER Ⅱ

LITERATURE REVIEW

2.1. General 14

2.2. Problems associated with the construction of embankments on

compressible soil around the world 15

CONTENTS

viii

2.3. Subsoil characteristics and geotechnical problems in

Chikushi plains, Japan 16

2.3.1 Subsoil Characteristics of Chikushi Plain 16

2.3.2 Some Geotechnical Problems 16

2.4. Reinforcement embankment on soft ground 17

2.4.1. The first uses of geosynthetics 17

2.4.2. Embankments on soft ground or cavities by laying reinforcements

and/or pile foundations 19

2.4.3. Design or modeling for the stabilized embalmment by geosynthetic

reinforcements 19

2.4.4. New applications and approach as the embankment reinforcement 20

2.4.5. Stability of the two test facilities from the deformation characteristics

of soft foundation 21

2.5. Foundation for building 22

2.5.1. Pile-Raft system 22

2.5.2. Stabilized box foundation (SBF) method 22

2.6. Finite element analysis on earth reinforcement 24

2.6.1. Contrast between LEM and FEM 25

2.6.2. Conerstones of analysis 26

2.7. General model aspects of Plaxis version 8 28

2.8. Summary 29

References 29

CONTENTS

ix

CHAPTER Ⅲ

SMALL-SCALE MODEL TEST OF RAFT & PILE FOUNDATION

3.1. General 32

3.2. Laboratory test program 32

3.2.1. Test set up 33

3.2.2. Model ground 35

3.2.3. Test procedure 36

3.3. Test result and discussions 40

3.3.1. Settlement at embankment center 40

3.3.2. Subsoil movement 41

3.3.3. Lateral displacement 45

3.4. Summary 48

References 49

CHAPTER Ⅳ

INVESTIGATION ON RAFT COMPOSITES FOR CONSTRUCTION

4.1. General 50

4.2. Bending test for Raft composites 51

4.2.1. Test set up 51

4.3. Test result and discussions 53

4.3.1. General observation 53

4.3.2. Effect of Raft layer 58

4.3.3. Effect of bending member 58

4.3.4. Effect of Raft pattern 59

CONTENTS

x

4.3.5. Effect of axial member 60

4.4. Links between the bending test, FEM analysis, and design

of Raft composites 62

4.5. Summary 64

References 64

CHAPTER Ⅴ

FINITE ELEMENT MODELING OF RAFT & PILE FOUNDATION OF

EMBANKMENT ON SOFT GROUND

5.1. General 50

5.2.Numerically examining an embankment on soft ground support

with Raft &Pile foundation 65

5.2.1. Problem considered and numerical model 65

5.2.2. Boundary condition and model parameters 67

5.2.3. Settlement at embankment center with time response 71

5.2.4. Displacement vector in subsoil 71

5.2.5. Effects of Raft 73

5.2.6. Effects of Pile 74

5.2.7. Effects of Raft & Pile 74

5.2.8. Surface vertical displacement with time response 81

5.2.9. Lateral displacement with time response 82

5.3. Evaluate effectiveness of Raft & Pile Foundation with

subsidence correction 83

5.4. Performance of road embankment design height 5.5m above the

CONTENTS

xi

ground level with Raft & Pile foundation 84

5.5. Summary 88

References 89

CHAPTER Ⅵ

CONCLUSIONS AND RECOMMENDATIONS

6.1. Conclusions 91

6.2. Recommendations for future research 93

PUBLICATIONS

List of figures

xii

LIST OF TABLES

CHAPTER Ⅰ

INTRODUCTION

Table 1.1: Change in greenhouse gas emissions, 1990-2004 (Kyoto Protocol) 7

Table 1.2: Cements and comparisons in amount of CO2 exhaust with

steel material, wood, and promotion woods (Numata et al 2006) 8

CHAPTER Ⅲ

SMALL-SCALE MODEL TEST OF RAFT & PILE FOUNDATION OF

EMBANKMENT

Table 3.1: Detail for test cases 39

CHAPTER Ⅳ


Table 4.1: Detail for test cases 51

Table 4.2: Purpose relation of EI between design and analysis in term of layer Raft 63

CHAPTER Ⅴ



Table 5.1: Material Properties 69

Table 5.2: Detail for numerical simulation of embankment (8cases, H=5.5m) 70

Table 5.3: Summarize the results of embankment heigh 5.5m 72

Table 5.4: Detail for numerical simulation of embankment 85

List of figures

xiii

LIST OF FIGURES CHAPTER Ⅰ

INTRODUCTION

Figure 1.1: Ariake sea coast road embankment 2

Figure 1.2: A bridge model of Nishidahashi, Kakoshima, Japan 4

Figure 1.3: Field investigation of Nishida Bridge foundation (1996) 4

Figure 1.4: Kurosaki embankment excavation investigation, Saga prefecture, Japan 5

Figure 1.5: Progressing of Global warming 6

Figure 1.6: Transition of stock amount feeling calculation of cement, steel

and wood product of Japan (Numata et al 2006) 9

Figure 1.7: CO2 fixation by CO2 absorption , afforestation and deforestation

by wood use (Numata et al 2006) 10

Figure 1.8: Result of investigation on durability of wood use to basic work 10

Figure 1.9: Use wood type under underground water condition (Numata et al 2007) 11

Figure 1.10: Flow diagram of the main parts of this study 13

CHAPTER Ⅱ

LITERATURE REVIEW

Figure 2.1: Illustration of problems in soft ground (Hayashi et al 1997) 16

Figure 2.2: The Construction controlling diagram by Matsuo et.al 22

Figure 2.3: Simplified representation of a piled-raft unit (Poulos, 2001) 23

Figure 2.4: Concept of stabilized box foundation method

(Koumoto, 1992) 23

Figure 2.5: Static on numerical analysis for earth reinforcement

(after A. Yashima, 1997) 25

Figure 2.6: Contrast between LEM and FEM 26

Figure 2.7: Four cornerstones of analysis (after Bolton, 1991) 27

Figure 2.8: Typical of element, node and stress point during mesh generation

(a)15-node element (b) 6-node element (Plaxis manual) 28

LIST OF FIGURES

xiv

CHAPTER Ⅲ


Figure 3.1: Schematic diagram of embankment on very soft clay

supported by Raft & Piles as foundation (a) cross section (b) isometric view 33

Figure 3.2: Test set up used in Raft and Pile model test 34

Figure 3.3: Photo showing grid and circles marked on the hard glass plate

and latex rubber membrane respectively for measuring displacements 34

Figure 3.4: (a) Variation of water content with depth (b) Variation of qc

with depth (cone penetration test) content with depth was measured 37

Figure 3.5: Isometric view of raft (made out of reapers), piles, sidereapers

and tie members 38

Figure 3.6: Embankment loading with time 38

Figure 3.7: Settlement at center of embankment 42

Figure 3.8: Surface vertical displacement for case (a) without any support MT-0,

collapsed under 10 kPa (b) support with raft MT-R1, under 30 kPa

(c) support with raft-pile MT-RP2, under 30 kPa 43

Figure 3.9: Displacement in ground (a) without any support MT-0 (b) support

with raft MT-R1(c) support with raft-pile MT-RP2 44

Figure 3.10: Lateral displacements at the end of final loading of 30 kPa for

raft and raft + pile 45

Figure 3.11: Vertical displacement at center of embankment for raft + pile 46

Figure 3.12: Effectiveness of Raft & Pile Foundation

(L=installed depth of piles) 47

Figure 3.13: Effectiveness of longer pile 48

CHAPTER Ⅳ


Figure 4.1: Test set up 52

Figure 4.2: Detail of scale model pattern Raft 52

Figure 4.3: Detail of scale model 53

Figure 4.4: Composites Raft case 1R12cm,11n-13L (side view) 53

LIST OF FIGURES

xv

Figure 4.5: Composites Raft case 1.5R12cm,11n-13L (side view) 54


Figure 4.7: Composites Raft case 2.5R12cm,11n-13L (side view) 55


Figure 4.9: Composites Raft case 3R12cm,11n-13L (Top view) 56




Figure 4.13: Effect of No. of Raft layer pattern A(long timber) 58

Figure 4.14: Effect of No. of Raft layer pattern B(short timber) 59

Figure 4.15: Effect of bending member raft pattern A 59

Figure 4.16: Effect of Raft pattern 60

Figure 4.17: Effect of No. axial member for raft composites pattern B(short timber) 60

Figure 4.18: Evaluate the effect of pattern of Raft composites for constructions 61

Figure 4.19: Considered effect of short timber and long timber for composites Raft 62

Figure 4.20: Detail for calculation EI of Raft (2D FEM analysis) 63

CHAPTER Ⅴ



Figure 5.1: At test site in Saga Prefecture, Japan (a) Boring test locations

(b) Cross section (c) Soil profile of Kaseminami and Kubota 66

Figure 5.2: Situation of road embankment on soft soil 68

Figure5.3: Typical finite element mesh for embankment height 5.5m 72

Figure 5.4: Settlement at embankment center (embankment height 5.5m) 73

Figure 5.5: Displacement for embankment height 5.5m (a) with 2R (b) with 2R6P 75

Figure 5.6: Effects of Raft (a) displacement for case 1R (b) displacement

for case 1.5R(c) displacement for case2R 75

Figure 5.7: Effects of Raft on excess pore pressure (a) excess pore pressure

for case 1R(b) excess pore pressure for case 1.5R

(e) excess pore pressure for case 2R 76

LIST OF FIGURES

xvi

Figure 5.8: Effects of Pile (a)displacement for case 2R (b)displacement

for case2R4P (c)displacement for case2R6P 77

Figure 5.9: Effects of Pile on excess pore pressure (a) excess pore pressure for

case 2R (b) excess pore pressure for case 2R4P (c) excess pore pressure for

case2R6P 78

Figure 5.10: Effects of Raft & Pile (a) displacement for case 1R7P

(b) displacement for case 1.5R, 6-0.5P (b) displacement for case 2R6P 79

Figure 5.11: Effects of Raft & Pile on excess pore pressure (a) excess pore pressure

for case 1R7P (b) excess pore pressure for case 1.5R, 6-0.5P (c) excess pore

pressure for case 2R6P 80

Figure 5.12: Surface vertical displacement with time for embankment height 5.5m 81

Figure 5.13: Lateral displacement at toe with time for embankment height 5.5m 82

Figure 5.14: Evaluation of the effectiveness of Raft & Pile Foundation with

subsidence correction 83

Figure 5.15: Typical Finite element 85

Figure 5.16: Settlement at embankment center 86

Figure 5.17: Displacement for case with Raft & Pile (a) Total displacement after

end of construction (25days) (b) Surface vertical displacement (c) Lateral

displacement at toe of embankment 87

Figure 5.18: Excess pore pressures for embankment design height 5.5m 87

Figure 5.19: Cross section of embankment 88

List of figures

1

CHAPTER

Ⅰ

INTRODUCTION

1.1. General

In recent years, the construction has been under pressure to operate with greater

finesse and accuracy. There were two main reasons for this. Firstly, with the Japanese

economy becoming more mature and environmental issue gaining importance, the

constraints on construction were gradually increasing. Secondly, it was recognized that

effects on the environment must be minimize to avoid pollution problem. This pressure

accelerates the advance of technology in the construction field.


settlement and lateral deformation will occur. These deformations will be transferred to

the surrounding ground surface and subsoil and will damage the adjoining buildings and

agricultural lands along the highways. Moreover, as an alternative if there are

piled-structures such as bridge abutments and/or box culverts across the road, there will

be large differential settlement problem between approach and the piled structures. This

differential settlement will affect the vehicle ride quality and/or even obstructs the

vehicle movement. Therefore, the soft subsoil should be treated throughly applying

CHAPTER I INTRODUCTION

2

various methods such as sheet-pile method, basal reinforcement of embankment, deep

mixed columns, and light-weight fill materials. Most of these methods use either

geosynthetics as reinforcements and/or chemical admixture as additive to improve the

soil strength. The use of the artificial material for long term for improving the ground

may work out expensive and may create long-term environmental problems.

The dissertation proposes a new foundation improvement method for an

embankment on soft subsoil to reduce the deformation of embankment by using natural

and environmentally-friendly materials.

Figure 1.1: Ariake sea coast road embankment.

1.2. Research Background

1.2.1. Ariake sea coast road embankment

The regional high standard road and the disaster prevention road of the

extension(~53km) where Fukuoka Prefecture Omuta City is made a starting point in the

Ariake Sea coast.

Saga Prefecture

Omuta

Kashima

Fukuoka Prefecture

Ariake Sea


3

1.2.2. Strategy of wood use as construction material in Ariake sea coast region

Management forest and land use of mountains area for five Prefectures around the

Ariake sea coast, a huge amount of forests in the state of not being maintained exist

(acts such as planting, the weeding, the improvement cuttings, and thinning). When

even only Saga Prefecture reaches 53,000 - 83,000m3 a year (2001 fiscal year results).

Since timber locally-available to be use as Raft & Pile, free except for the

transportation fee, which is an economic way but also to aim at the reproduction of the

forest in the Ariake Sea coast region.

1.2.3. Safety-high durability and environment friendly of timber foundation on soft

ground

The proposed Raft & Pile method uses the natural material thinning wood and

revives the principle of an ancient technology such as "Ladder Douki", "Well crib and

Citigou" that has been often used as a stone wall base of a flat castle , and "Floating

foundation" at the present age.

When building stone structures on soft ground, it was first necessary to drive piles

in the soil. Ladder-shaped foundations of wood were then placed on the piles before

stone could be laid. Clearly, this type of work requires a deep understanding not only of

stone, but also of how soil and water behave. The principles of marsonry were

transmitted word-of-mouth though history. Phrases such as “the strength of mutsume

(stone with six contacting forces)” and “backfilled stone walls last longer” are still use

today (Penta Ocean Construction, 2000).

Here, some examples of the wood use are enumerated in engineering works.

Figure 1.2 show the process of construction a stone arch bridge Nishidahashi,,

Kagoshima city, Japan up to now of about 160 years ago. Ladder Kiriki method was

especially spread in Nishidahashi and the location of timber raft foundation shown in

Fig. 1.2. Field investigation of Nishida bridge foundation in 1996 (see Fig. 1.3) was

indicated that timber raft foundation was safety, high durability, and environment

friendly since timber are natural material, the raft timbers will not give pollute to

environment.


4

Figure 1.2: A bridge model of Nishidahashi, Kagoshima, Japan.

Figure 1.3: Field investigation of Nishida Bridge foundation (1996).

(Construction year:1846)

Nishida Bridge

(Construction year:1846)

Nishida Bridge


5

Figure 1.4: Kurosaki dike embankment excavation investigation, Saga prefecture, Japan.

A large amount of woods used for the base of the old willow river clan and the

Kurosaki dike embankment height of about 6m shown in Fig 1.4, wood had been used

from the small branch to the leaf with minimal waste. Moreover, wood is placed

beneath the embankment, and Pile of the proposed Raft&Pile method support at toe

both side of the embankment. It have been constructed about 350 years ago.


6

1.2.4. Develop a non-polluting construction technology to minimize environment

impact

The increasing of demand of our civilization has been under pressure to operate

with greater safety and reliability. Indeed, Earth’s atmosphere is so thin we have vastly

increased the amount of carbon dioxide, the most important of the so – called

greenhouse gases (Algore 2007). Recently, the global warming is progressing in terms

of green house gas emission (Fig. 1.5). Japan has the obligation to reduce CO2

emissions 6% less than the amount of 1990 by average amount of 2008~2012 (Kyoto

Protocol) as shown in Table 1.1.

Cement is multiused to the industrial method used as a main current now. When

the raw material obtains one quicklime (calcium oxide and CaO) from the lime stone

(calcium carbonate and CaCO3) in manufacturing it cement, a large amount of carbon

dioxides (CO2) are discharged. Carbon dioxide is one of the heat-trapping gases, and is

included by the main factor of global warming.

(a) (b)

Figure 1.5: Progressing of global warming (a) Average global temperature, 1880-2005

(b) Global atmospheric concentrations of Carbon Dioxide, 1760-2004


7

Table 1.1: Change in greenhouse gas emissions, 1990-2004 (Kyoto Protocol)

-8%N/A-2.6%-0.8%EU-15

-8%+27%+28.9%+41%Portugal

-8%-12.5%-14.8%-14%United Kingdom

-6%N/A+5.2%+6.5%Japan

-8%+13%+22.7%+23%Ireland

-8%+25%+25.3%+27%Greece

-8%0%-6.1%-0.8%France

0%N/A+17.9%+21%New Zealand

+1%N/A-18.7%+10%Norway

-7%*N/A+21.1%+16%United States

-8%+15%+50.4%+49%Spain

+8%N/A+5.2%+25%Australia

-6%N/A+62.2%+27%Canada

-8%-21%-18.2%-17%Germany

Treaty Obligation 2008-2012

EU Assigned Objectivefor 2012

Change in greenhouse gasEmissions (1990-2004)

including LULUCF


excluding LULUCFCountry

-8%N/A-2.6%-0.8%EU-15

-8%+27%+28.9%+41%Portugal

-8%-12.5%-14.8%-14%United Kingdom

-6%N/A+5.2%+6.5%Japan

-8%+13%+22.7%+23%Ireland

-8%+25%+25.3%+27%Greece

-8%0%-6.1%-0.8%France

0%N/A+17.9%+21%New Zealand

+1%N/A-18.7%+10%Norway

-7%*N/A+21.1%+16%United States

-8%+15%+50.4%+49%Spain

+8%N/A+5.2%+25%Australia

-6%N/A+62.2%+27%Canada

-8%-21%-18.2%-17%Germany

Treaty Obligation 2008-2012

EU Assigned Objectivefor 2012


including LULUCF


excluding LULUCFCountry

In addition, it was thought whether global warming could be controlled using

natural materials in civil engineering such as wood. The concrete reduction target of the

heat-trapping gas is shown in the Kyoto Protocol issued on February 16, 2005, and it

increases by 8.3% compared with fiscal year 1990 in fiscal year 2003 though it is 6% in

our country compared with fiscal year 1990. In other words, the reduction in 14.3%

total is needed. It is a method of absorbing fixing CO2 to the forest that want to pay

attention especially though there are various methods of heat-trapping gas reductions. If

it is achieved to afforest this a large amount of stable method, the heat-trapping gas can

be reduced extremely naturally without impossibility. The voice whether to assume

"Use of the construction industry" that starts using the manpower of the construction

industry with high possibility of becoming a long-term depressed industry for the forest

business rises as a mechanism to this. In this case, stable, a large amount of demand for

wood comes to the necessary condition. If a large amount of using it becomes possible,

reducing the heat-trapping gas from the decrease of the amount of the exhaust of the

heat-trapping gas further becomes possible without stably decaying wood for a long

term. Then, the verification of the effect of wood of the CO2 reduction and the durability

of the use wood is described next. Table 1.2 is an amount of the CO2 exhaust of cement

and the steel material and the amounts of the CO2 stock in wood and the promotion


8

woods. It is understood to change from the CO2 exhaust side to the stock side surely by

at least replacing cement and the steel material with wood though this cannot be simply

compared as it is. Figure 1.6 show the transition compared with 1964 of cement, the

crude steel, and the wood of Japan production. The one that the amount of 1.1 of the

stock of CO2 a material ton for each was multiplied is shown.

Table 1.2: Cements and comparisons in amount of CO2 exhaust with steel material,

wood, and promotion woods (Numata et al 2006).

−７.２５ (t-CO2/ha/year) (absorbed amount)Promotion woods(ha/year)

−１.６４ (t-CO2/material (t)) (amount of stock)Wood

１.２６ (t-CO2/material (t))Steel material (type steel)

０.７７ (t-CO2/material (t))Cement (Port land cement)

Amount of CO2exhaust (amount of stock)Material

−７.２５ (t-CO2/ha/year) (absorbed amount)Promotion woods(ha/year)

−１.６４ (t-CO2/material (t)) (amount of stock)Wood

１.２６ (t-CO2/material (t))Steel material (type steel)

０.７７ (t-CO2/material (t))Cement (Port land cement)

Amount of CO2exhaust (amount of stock)Material

Here, because each value is different, it regularizes it by the value in 1964 from

which Tokyo Olympics is held. The target value of 6% of the amount of the CO2

exhaust in 1990 shown in the Kyoto Protocol as a standard was indicated in Fig 1.6.

The amount of the stock of wood exceeds the total of the amount of the CO2 exhaust of

cement and the steel material, and this gross weight of three people is positioned on the

CO2 stock side if paying attention in 1950's. This relation is completely reversed

through high economic growth afterwards, and the total of the amount of the CO2

exhaust of cement and the steel material is about ten times or more the amount of wood

of the CO2 stock. The CO2 reduction becomes possible though it might be seemed that

global warming is promoted by using a large amount of woods if it does well

afforestation and deforestation the balance Japanese larch and the Japan cedar absorb a

large amount of CO2 between the woods ages until 15 years and the absorbed amount

decreases gradually.


9

-200

-150

-100

-50

0

50

100

1950 1960 1970 1980 1990 2000 2010

1964 Tokyo Olympic1964 Tokyo Olympic

CO2stock

Year

(million ton) Construction material in Japan(1950-2005)

Steel

Cement

Wood

Source: Numata et al(2006)

Figure 1.6: Transition of CO2 stock amount feeling calculation of cement, crude steel,

and wood production of Japan (Numata et al 2006).

Then, if it is repeated to deforest before it makes it to aged and to afforest it at the

same time as preserving the wood, it newly stores and, maybe, supplement can be

matched (see Fig. 1.7). As the condition, it is a forest where it is newly afforested after

1990, and appropriate forest management (acts such as planting, the weeding, the

improvement cuttings, and thinnings) is done. In Japan, the promotion woods becomes a

calculation that absorbs 1.77 tons a hectare year (carbon conversion) as an average

numerical value. Hence, it can be said that it is possible to contribute also to the global

warming prevention reduction of greenhouse gases furthermore by developing and

using the Raft&Pile industrial method.

Moreover, there is a result of investigating the state of decay and the tree kind of

wood that passes 1-100 year (Fig. 1.8) when the investigation of results of the wood use

to a basic worker is seen. The investigation result that it is an underground water table

100% healthy condition while the wood of 40% has decayed has come out in two

sections of ground water table level, underground water title change region when the

inside of the ground was divided into underground water table change, underground

water table change region and underground water title change and the ratio of the tree

kind used and wood type under underground water condition was shown figure 1.9 In

addition, it is assumed that all building base that passes about 400 years of the Venice

urban area is a tree paling. Durability is proven also under the seawater environment in

a word. Durability is demonstrated from the above-mentioned by underground water


10

table sinking wood. Moreover, it is thought that the controlling effect of global warming

can be expected by maintain of the forest advancing and absorbing CO2 by using wood.

Figure 1.7: CO2 fixation by CO2 absorption , afforestation and deforestation by wood

use (Numata et al 2006)

腐朽

40%

健全

60%

15

腐朽

40%

健全

60%

15

Above ground water table Ground water table variation zone Underground water table

腐朽

0%

23

Healthy( 60%)

Decay( 40%)

Decay( 40%)

Healthy( 60%)

Healthy( 100%)

Use timber under groundwater condition

Above ground water table

Ground water table variation zone

Under ground water table

Use timber under groundwater condition

Above ground water table

Ground water table variation zone

Under ground water table

Figure 1.8: Result of investigation on durability of wood use to basic worker

(Numata et al 2007)

0 〜 20 〜 60year

0 〜 20 〜 60year

0 〜 20 〜 60year

Deforestation2Deforestation3

Tre

e ab

sorp

tion

and

amou

nt

of st

ock

of C

O2

(ton

/ha)

turning phase

• CO2 fixation CO2 by absorption (promotion wood and wood use as construction material)

• Afforestation(~60years till turning phase)• Deforestation(to use wood as construction material)

Amount of CO2 absorptionin promotion woods7.25 t-CO2 /ha/year

turning phase

turning phase

Afforestation1

Afforestation2

Afforestation3

Deforestation1

total stock1

total stock2 total stock3

total stock of C

O 2

(ton)


11

Pinko Matsu( 4%)

カラマツ

4% アカマツ

9%

松(松or松杭)

35%

スギ(オビスギ含む)

26%

不明

17%

ベイマツ

9%

その他(ヒノキ，サザ

ンイエローパイン等)

0%

23

最大26年

(沈床) 最大55年

(橋脚基礎)

最大76年

(ビル基礎)

最大86年

(樋門基礎)

最大26年

(柵支柱)

Others (Japanese cypress and Southern Yellow Pine, etc.)

Uncertainty( 17%)

Larch( 4%)

26 years or less( hedge prop)

86 years or less( Sluiceway base)

Bay Matsu( 9%)



Cryptomeria( 26%)


Pine (Pine or matsci)( 9%)

Figure 1.9: Use wood type under underground water condition (Numata et al 2007)

1.3. Objective of this research

Given the benefits of wooden materials for in civil engineering constructions, there

is need to provide engineers the design and construction guidelines of using Timber Raft

and Pile Foundation method in supporting road embankments on soft clay foundations.

This requires research to determine the mechanical properties of the composite

materials, to obtain improved understanding of the mechanisms involved under

operating conditions, and to evaluate the overall performance of the technique in field

conditions.

The specific objectives of this study are as follows:

1. To develop a laboratory test apparatus for testing the laboratory-scale Raft &

Pile foundation subjected to embankment loading.

2. To understand the deformation behavior of ground and embankment

reinforced by Raft & Pile foundation during embankment loading.

3. To evaluate the effectiveness of Raft & Pile methods in terms of reducing

differential settlements and lateral spreading.

4. To evaluate flexural behavior of raft composites for design and construction

calculations.

5. To develop design and construction guidelines of using Timber Raft & Pile

foundation for supporting embankments in soft clay foundations


12

1.4. Scope of this dissertation

This dissertation contains six chapters. Chapter I is the introductory chapter and it

describes the objectives and scope of the work. Chapter II is literature review that starts

with ing looking at problems associated with the construction of embankments on

compressible soil and the various ground improvements remediating the problems.

Links between testing, modeling, and design also reviewed. Critical and positive

comments on previous works are given in relation to the work that was proposed in this

study. The third chapter describes the effectiveness of Raft & Pile foundation

subjected to embankment loading based on laboratory model tests. At first, the detailed

experimental program and procedures, and soil properties determined in the laboratory

are introduced. Then, the test results are presented in terms of their deformation

characteristics. Chapter IV presents the results of the experimental study investigating

the flexural behavior of composites Rakft for construction. Links between the bending

test (chapter IV), FEM analysis and design of Raft composites discussed. Chapter V

presents the finite element modeling of Raft & Pile foundation of embankment on soft

ground. FEM simulation embankment height 5.5m constructed over soft Ariake clay

ground using a computer code PlaxisV.8. In order to investigate the effects of raft size,

pile length, and residual settlements of the road surface in case of embankment height

5.5m. Based on the results in Chapters III, IV, and V, design, and construction

guidelines method for Raft & Pile foundation are also given. Finally, the main

conclusions drawn from this study are given in Chapter VI.

The flowchart showing links between the six chapters is given in Figure 1.10.


13

DEVELOPMENT OF A TIMBER RAFT & PILE FOUNDATION FOR EMBANKMENTS ON SOFT GROUND

INTRODUCTION

INVESTIGATION ONRAFT COMPOSITESFOR CONSTRUCTION

FINITE ELEMENT MODELING OF RAFT &PILE FOUNDATION OF EMBANKMENT ONSOFT GROUND

CONCLUSIONS AND RECOMMENDATIONSCONCLUSIONS AND RECOMMENDATIONS

LITERATURE REVIEW

Chapter IIIChapter III

Chapter IChapter I

Chapter IIChapter II

Chapter IVChapter IV

Chapter VIChapter VI

Chapter VChapter V

SMALL-SCALE MODEL TESTOF RAFT&PILE FOUNDATION

Figure 1.10: Flow diagram of the main parts of this study.

List of figures

14

CHAPTER Ⅱ

LITERATURE REVIEW

2.1. General

Amongst the various techniques of constructing embankments on soft ground, the

use of geosynthetic basal reinforcements and/or chemical admixtures as additive to

improve the soil strength are commonly used. The use of the artificial material for long

term improvement of the ground can be expensive and may create environmental

problems. Timber Raft & Pile method is a versatile, cost effective and environmental

friendly alternative.

This literature review starts with the statement of the problems associated with the

construction of embankments on compressible soil as viewed by engineers in research

and practice. Various soft ground improvement techniques to rectify the problems are

also reviewed. The links between the testing, modeling, and design are discussed based

on the works of other researchers. Attention is paid to the development of

environmentally friendly construction technology. Critical and positive comments are

given to the reviewed materials as they relate to this study.

CHAPTER II LITERATURE REVIEW

15

2.2. Problems associated with the construction of embankments on compressible

soil around the world

Embankments are among the most ancient forms of civil engineering construction

but are also among the most relevant today. Economic and social development has

brought about a considerable increase in the construction of this type of structure since

the middle of the nineteenth century, and particularly since the 1950s. Embankments are

required in the construction of road, motorway and railway networks (linear

embankments, access embankments, and embankments across valleys), in hydroelectric

schemes (dams and retention dykes), in irrigation and flood control works (regulation

dams), harbor installations (seawalls, quays and breakwaters) and airports (runways).

Problems associated with the interaction between embankments and structures are

among the most difficult to control. Most embankments on compressible soils are

crossed by channels and hydraulic works intended to both re-establish the normal flow

regime of the water (linear road embankments) and ensure the proper functioning of the

structure (drainage galleries in dykes and dams). Considerations of settlement influence

not only the choice of type of construction (flexible channels in preference to rigid

structures) but also the specification of their longitudinal profile (necessity for reverse

camber in order to ensure a camber compatible with the requirements of flow when

settlement has ceased). There may also be a need to overdesign the structures in order to

guarantee safe and more reliable performance of these structures during their design

life.

The interaction between two embankments, old and new, poses problems for which

it is difficult to find a satisfactory solution. In the case of enlarging an old embankment,

the settlements under the new embankment will inevitably produce disturbance in the

old one, and this will occur over a prolonged period. The linking of a new alignment

with an existing road poses the same type of problem, but at a point, so that solutions,

such as the placing of a curtain of sheet piles in order to isolate the foundation soils of

the two embankments, can be conceived. Serious difficulties can also arise in the

extension of runways. In all cases the designer must, from the start of the project,

attempt, as far as possible, to foresee and to plan the arrangement of embankments that

will be needed in the long term.


16

2.3. Subsoil characteristics and geotechnical problems in Chikushi Plain, Japan

2.3.1 Subsoil Characteristics of Chikushi Plain

The Ariake Sea is an inner sea, which is long in the north-south direction and has a

small outlet. Due to these geographical characteristics, the tide range is large (maximum

about 6 m) which creates special water movement in the sea, and has a significant effect

on the depositional environment of the Ariake clay. The soil particles carried by

Chikugogawa (the largest river in Kyushu Island), Kikuchikawa, Sirakawa, and

Midorikawa rivers with Aso Mt. volcano region as a source, first move offshore, then

with high tide, the suspension moves toward the shore with counterclockwise tidal

current, and the particles get deposited gradually along the shore area. This process was

repeated year by year, the tidal land developing gradually, and forming the soft deposit

in Chikushi plains. The fine particles carried by the rivers are volcano ash from Kuju

mountain and Aso mountain area, and the diatom deposited in Jakes and swamps is

from the upper regions of the rivers. The high water content and high sensitivity of

Ariake clay can be explained from the properties of these particles.

2.3.2 Some Geotechnical Problems

The Ariake clay which formed the Chikushi plains has a water content of 80 to

160%, high compressibility, and high sensitivity and categorizes it as quick clay. The

Saga plain is the softest area in Chikushi plains. In Saga area, the shoreline was

advanced by about 10 km for the above mentioned deposit and land reclamation, and

forming a very young and soft deposit. The geotechnical problems of this kind lowland

region are illustrated in Fig. 2.1.

Figure 2.1: Illustration of problems in soft ground (Hayashi et al 1997).


17

The subsidence has become a serious problem since 1955. At that time, large

mounts of ground water were pumped out due development of deep well techniques.

At present, in Saga prefecture, the area of subsidence is about 320 km2, i.e. about 46%

of the total area of Saga pref. At the northeast are of Chikushi plains (Morotomi, Saga,

Kawazoe), he ground water was mainly used by industries and for people’s daily living,

while at the northwest area of Chikushi plains (Fukutomi, Ariake, Shiroishi), has been

utilized still for agriculture and domestic use.

From 1975, with construction of dams and restriction on ground water pumping the

surface water has been used for industry instead of ground water, and in the northeast

area, the subsidence has been almost arrested. However, in the northwest area, the

agriculture still depends on ground water, and subsidence is continuing, especially, in

the years of low rainfall (in 1978, 1994), when argue mounts of settlement occurred.

Furthermore, with the popularization of automobiles since 1960’s, there are

problems for construction of roads on soft ground along with the settlement caused by

the traffic load. The settlement of the road, the deformation of the surrounding area, and

the settlement gaps between pile supported structure and embankment on natural

ground, have become some of the major problems in Saga plain.

2.4. Reinforcement embankment on soft ground

2.4.1. The first uses of geosynthetics

R. Holtz stated that A. Casagrande, in his lectures at Harvard University,

mentioned the possibility of reinforcing earth dams by using steel rods. In 1967, J.

Kerisel used wire mesh in the construction of the Arezal dam. This 13 meter-high dam

was built on 24 meters of silt in a tidal estuary. A 2-metre layer of sand was placed on

the 3.5kPa concessive silt. This layer served as a drainage layer for the vertical drains in

the sand. In order to prevent this sand layer from being punctured by the riprap in the

layer above and to improve the resistance of the structure to horizontal spreading, wire

mesh with a resistance of 270 kN/m was installed. During construction and because of

acceleration in works, failure occurred.

It would appear that the Dutch were the first to use geotextiles in embankments: in

1953, the authorities used geotextiles when reconstructing the dykes destroyed by a tidal

storm.


18

In France, Devaux et al. reported using a non-woven geotextile as a

non-contamination sheet and for load distribution as early as 1970: on marshy ground

(2.50 m of peat), non-woven geotextile was inserted under the embankment, which

reached a height of 3 meters above the grown soil. Settlement was as much as 1 meter,

but the measured elongation of the geotexiile was no more than a few percent.

In Japan, the 1967 report by Fukuzumi and Mishibayashi, of the Obayashi

Research Institute is a surprising forerunner in this field. It describes the use of vinylon

sheets with a resistance of about 25kN/m at bottom of embankments on soil with a very

low bearing capacity (cohesion less than 1 kPa). This technique is known as “Fagot

sheets”, where the membrane effect of the textile is used. An application involving a

hydraulic embankment at the Nippon-Known as “Fagot sheets”, where the membrane

effect of the textile is used. An application involving a hydraulic embankment at the

Nippon-Kokan site in Fukuyama had been documented: without the textile, the soil in

question sank to a depth of 10 meters. When textile sheets were used, settlement was

limited and savings of 400% in the volume of full were made.

In Sweden, Wager et al. invented a system for reinforcing embankments whereby two

rows of short piles under each crest of the embankment were connected by steel tie rods

in order to limit spreading of the embankments. This system was used in Sweden in

1968 and was later based on the use of geotextiles (1971).

It was also in Sweden (1975) that a combination of piles and geotextiles was first

used (Holtz et al). The structure in question was an approach embankment, to a bridge

founded on piles, on a site composed of sensitive clays. The timber piles, 0.17 m in

diameter and spaced 1.5 meters apart, were surrounded by a paper-plastic drain and

equipped with a concrete cap to facilitate transfer of the embankment load to the piles.

Three sheets of woven polyester textile were placed in the foundation layer of the

embankment in order to limit horizontal movement.

In the United States, the first reinforced embankment is attributed to J. Bell et al.

(1976). The embankment was constructed in Alaska on peat and had a maximum height

of 2.4 m (3 m average thickness of the peat and cohesion of 12 kPa). Savings in fill

material were estimated at 28% and deformation of the non-woven geotextile, which

was slight at the time of construction, was as much as 50% in places later.


19

2.4.2 Embankments on soft ground or cavities by laying reinforcements and/or pile

foundations

Countermeasure for embankments on soft ground or cavities are chosen by laying

reinforcements on surface of base ground with or without pile foundations as supporting

embankment loads.

Stolarski, G. et. al. introduced road embankment over old waste dump, chemical

contaminated and heterogeneous layer of 4-6m thickness by laying georgics. From field

measurement it is confirmed the effect of decreasing the settlement and differential

settlement of the embankment and protecting environmental contamination.

Ast, A. et al. introduced over bridging system, geogrids 2 layer, for embankments

over cavity by field experiments. Water filled cushions was used for cavity model.

Results of monitoring and some aspects of design were presented as a safe and

economic engineering structure.

Sa, C. T. et al. assessed the behavior of reinforced piled embankments and

reinforced retaining walls on soft soils by numerical parametric simulations (FLAC).

Parameters as pile spacing foundation reinforcements were varied. The behavior of case

histories was supported the analyses.

Alexiew, D. et. al introduced case history of piled railway embankment,

Interesting case of projects were concerned on renewal the existing embankment to high

speed train embankment by removing top part of embankment, installation cemented

stone columns, horizontal geosynthetic reinforcement and new track embankment. It is

reasonable for eliminate additional embankment load against soft soils.

2.4.3. Design or modeling for the stabilized embankment by geosynthetic

reinforcements

Experiences and know-how can be obtained from etching performances of

reinforced soil structures. However, it is necessary to convert quantitatively as some

suitable formulae expressed reinforced mechanism for ease to use. The target of all

research activities is to constitute design for various reinforced structures evaluated

reinforcing mechanism.

Madhav, M. R. et al Proposed modeling on response of geosynthetic reinforcement

to transverse force assuring a simple Winkler taper model. Fundamental concept was

that the response to the applied force depends not only on the interface a hear


20

characteristics of the reinforcement but also on the deformational response of ground. A

parametric study quantified the contributions of length and interface characteristics of

the reinforcement, stiffness of the ground, etc. on the over all response.

Shahgholi, M. et al. presented Horizontal Slice Method (HSM) as a new limit

equilibrium method. Comparative analyses using HSM and established computer

program showed good agreement, and confirmed the advantages of method. By the way,

same approach was developed as inclined slice method for nailing (Gutierrez, V. and

Tatsuoka, F., 1988).

2.4.4. New applications and approach as the embankment reinforcement

Li, G. X. et al. assess performance and stability of fiber reinforced cohesive soil by

centrifuge model tests (45 – 120G) compared with un-reinforced case. By reinforcing

fill material, acceleration at failure occurred and critical height and depth of tension

zone were increased.

Tatta, N. et al. considered earth flow prevention embankment reinforced with

geosynthetics as flexible structure. Horizontal loading were conducted for two types of

model (H = 1.5 m, 6m, vertical facing) assumed as pseudo-static load of earth flow

force. Shear resistance of the structure was increased by reinforcements. It was

concludes that additional treatment by pre-stressing would improve the structure as an

integration body.

Deformation analysis of preload and pre-stressed (PLPS) reinforced soil method

for railway bridge pier, constructed in 1996 during construction and in service was

introduced by Uchimura, T. et al. To develop a methodology for predicting the time

dependent behavior of such structure, New Isotach Model, three component theology

models, was used to analyze the observed behavior of PLPS pier during preloading

procedures.

Kubo, T. et al. successfully constructed an arching structure by using large-sized

sand bag that perform restriction effect of geosynthetics as bag material. The purpose

was to evaluate its constructability and result of its observation. From the success of

construction, the authors concluded the proposed method was able to construct with

regarding some deformation of soil bags.


21

2.4.5. Stability of the two Test facilities from the deformation characteristics of soft

foundation

Marolo C. Alfaro indicated that the stability of reinforced soil structures are usually

checked from the integrity of the reinforced soil mass (internal stability) and its global

stability (external stability). As indicated by Jones and Edwards (1980) and

demonstrated by the results of the filed tests and finite element analyses in this paper,

the response of the soft foundation to behavior of the reinforced soil mass has

significant influence on the global stability. Consequently, the safety against instability

of reinforced soil wall-embankment systems on soft clay foundation can be assessed the

premise that the soft foundation essentially controls the stability of the structure. The

method described by Matsuo and Kawamura (1977) for the failure prediction of

embankments on soft ground was employed. It makes use of the observed lateral

movements (δ) and settlements (S) of the foundation under the embankment loading

which are plotted inδ/S-S coordinate as shown in Fig.2.2 The degree of safety against

instability is estimated from a plot of curves corresponding to the values of q/qf, where q

is the load at any stage and qf is the load at failure. These curves are defined by

regression equations derived from results of a large number of actual failure case

histories. Any observed deformation path in the δ/S-S relation which approaches the

q/qf = 1.0 curve signifies a high probability of failure. As can be seen in Fig. 2.2., the

observed deformation paths of the two test facilities for almost a year of observation are

well below the critical boundary curves. It should be noted that the process of

consolidation of soft clay foundation results to the increase of its shear strength, which

then reduce the rate of lateral movements and settlements. Thus, it is expected that the

observed deformation path in the long term condition would diverge from the critical

curves and would eventually tend towards stability. Moreover, based on visual

inspection of the two test facilities, the facing units remained intact even with large

lateral deformation and settlement owing to their flexible characteristics. Special

attention has to be paid however when rigid facing units will be used. The chosen type

of facing units will put some limitations on the amount of settlement that can be

tolerated by the wall. If settlement is still of concern, technically viable and

cost-effective ground improvement techniques such as granular piles, lime or cement

columns and vertical drains with preloading may be used (refer Bergado et al 1994).


22

Figure 2.2: The Construction controlling diagram by Matsuo et.al.

2.5. Foundation for building

2.5.1. Pile-Raft system

Poulos and Davis (1980) point out that in the design of the foundation for a large

building on a deep deposit of clay, it may be found that a raft foundation would have an

adequate factor of safety against ultimate bearing-capacity failure, but that the

settlements would be excessive. Traditional practice (assuming the additional of

basements to produce a floating foundation is unacceptable) would then be to pile the

foundation (Fig. 2.3), and to choose a number of piles to give an adequate factor of

safety against individual pile failure, assuming the piles take all the load, However, it is

clearly illogical to design the piles on an ultimate-load basis when they have only been

introduced in order to reduce the settlement of an otherwise satisfactory raft.

2.5.2. Stabilized box foundation (SBF) method

Koumoto et all 1992 point out that the stabilized box foundation (SBF) method, a

box shaped wall is formed by soil-cement columns installed in ground with an improved

surface layer connected together to form a box foundation (Fig. 2.4.). This kind of

foundation has following functions.

C

maδ o S

L

( )⎪⎭

⎪⎬⎫

⎪⎩

⎪⎨⎧

⎟⎟

⎠

⎞

⎜⎜

⎝

⎛−+⎟⎟

⎠

⎞⎜⎜⎝

⎛=

SSs

oo

oδδ max4.328.1exp93.5

2

max

( )⎪⎭

⎪⎬⎫

⎪⎩

⎪⎨⎧

⎟⎟

⎠

⎞

⎜⎜

⎝

⎛−+⎟⎟

⎠

⎞⎜⎜⎝

⎛=

SSs

oo

oδδ max49.240.0exp80.2

2

max

fqq /

0.1

9.0

S/δ

0.5

1

1.5

2

2.5

3

3.5

4

0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9

s(m

)


23

1) The vertical wall has confines the soil reduces the lateral displacements, and

consequently, the settlement of the ground.

2) Load is transmitted to a deeper location and a larger bearing area. The thickness of

the compressible soft layer is thus reduced resulting in a smaller settlement.

3) Bearing capacity is increased due to the side resistance between the box wall and

the soil.

4) The box and the soil inside act together forming a deep embedded foundation,

which increases the bearing capacity of the ground.

Figure 2.3: Simplifield representation of a piled-raft unit (Poulos, 2001).

Figure 2.4: Concept of stabilized box foundation method.(Koumoto, 1992)


24

2.6. Finite element analysis on earth reinforcement

A. Yashima point out that, the statistics on current state of the numerical analysis is

summarized based on the survey of technical papers related to the numerical analysis on

earth reinforcement, which were published in Journals and Proceedings for last 12 years

(1985 – 1996). The details of this survey are introduced in another report written by the

Japanese Local Task Force Committee of Asian Technical Committee for Earth

Reinforcement in this proceeding volume.

More than 160 technical papers in which there is a keyword “analysis” were

selected and reviewed from the following journals and proceedings;

- ASCE Journal of Geotechnical Engineering

- Geotechnique

- Canadian Geotechnical Journal

- Soils and Foundations

- Geotextiles and Geomembranes

- Computers and Geotechnics

- Int. J. Num. Anal. Math. In Geomechnics.

- Proc. JSCE

- Proc. IS-Kyushu (’88, ’92 and ’96)

Half of the selected papers treated the deformation problem and another half

treated the stability problem (Fig. 2.5.(a)). It is found from Fig. 2.5.(b) that the finite

element method (FEM) was used for the prediction of the deformation of the ground

with earth reinforcement. On the other hand, the limit equilibrium method (LEM), slip

line method, upper/lower bound theorem, and RBSM, etc. were used for the assessment

of the stability of the ground with earth reinforcement. Figure 2.5.(c) summarizes the

type of structure treated in the technical papers. Deformation/stability problems on

embankment and foundation subsoil were mainly investigated by the numerical

analysis. From this statistic, the finite element analysis on embankment and foundation

subsoil with earth reinforcement is found to have been studied intensively for last 12

years. In this report, therefore, the finite element analysis will be discussed as a main

subject.


25

(a) (b)

(c)

Figure 2.5: Static on numerical analysis for earth reinforcement (after A. Yashima,

1997).

2.6.1. Contrast between LEM and FEM

For the analysis by the limit equilibrium method (LEM) which is commonly used

for the assessment of the stability of the ground with earth reinforcement, the design

strength parameters are only needed. The construction process should be reflected in

determining design strength parameters. We have to choose untrained, partially drained

or fully drained strength of subsoil materials based on the construction process and

permeability of subsoil. The compressive, extensive or simple shear strength should be

carefully used based on the stress condition of the subsoil near the failure.

For the analysis by the finite element method (FEM) which is used for the

prediction of the deformation of the ground with earth reinforcement, on the other hand,

we have to predict the elastic behavior, plastic (viscoplastic) behavior and failure

(ultimate limit state) based on a correct initial condition (Fig. 2.6). For this purpose, the

correct initial condition, constitutive model and stable numerical algorithm must be

prepared before the analysis.


26

Figure 2.6: Contrast between LEM and FEM (after A. Yashima, 1997).

2.6.2. Cornerstones of analysis

R.A.Jewell point out that evaluation of the effects of forces on a deformable body

is carried out with respect to three conditions: equilibrium, compatibility, and

stress-strain relations. In soil mechanics, this evaluation is often based on plasticity

theory, with sufficient strain being assumed to have been mobilized at every point

within the zone of plastic deformation in the soil to allow an ultimate strength to be

developed. This simplifies the design problem to one of solving for equilibrium only,

using an appropriate soil strength and relevant imposed loads.

The same approach is used for the ultimate limit state analysis of reinforced soil,

where a limiting design strength for the reinforcement, and limiting interaction

coefficients, must also be selected

The more detailed analysis for reinforced soil could be described as follows

(Figure 2.7.). Starting with gravity forces (at the top of Figure 2.7.), a distribution of soil

stresses in the structure which are in equilibrium may be found. Accompanying these

soil stresses would be soil strains, some of which will be in extension adjacent to

reinforcement elements.

Compatibility requires a shear interaction if there is to be any relative displacement


27

between the soil and the reinforcement, and this causes tension in the reinforcement and

reduces the magnitude of tensile strain in the soil. The reinforcement now exerts body

forces in the soil, and a new equilibrium must be sought allowing for these

reinforcement forces, the loading from gravity and any external boundary forces. The

cycle of analysis is repeated until a satisfactory equilibrium is found.

The cycle shown in Figure 2.7. provides a useful reference frame within which the

factors influencing the behavior of reinforced soil may be identified. For example, the

role of boundary forces in reinforced soil walls or the role of the interaction between the

soil and the reinforcement, are already included in Figure 2.7. Likewise, the material

properties of the soil and the reinforcement dominate one diagonal (stress-strain

force-displacement), while the behavior of the reinforced soil system governs the other

diagonal (equilibrium/compatibility).

Figure 2.7: Four cornerstones of analysis (after Bolton, 1991).

As mentioned earlier, ultimate limit state analysis in soil mechanics is usually

reduced to a problem of equilibrium alone. A design strength for the soil and the

reinforcement is selected (no consideration of stress-strain properties) and the limiting

force in anyone reinforcement layer is set to the lesser of the design strength or the bond

force, (no consideration of compatibility or force and displacement). But note that the

influence of deformation is usually tacitly considered in the choice of the design


28

strength for the soil, especially where the possibility of stain-softening to the critical

state strength has been anticipated.

The analysis for serviceability should consider properly all the main aspects of

behavior governing the equilibrium. However, by limiting the deformation in the

reinforcement to a maximum allowable elongation (which is associated empirically with

satisfactory behavior), enables a maximum allowable force to be derived, and the

analysis can be simplified once more to one of equilibrium alone.

2.7. General model aspects of Plaxis version 8

Plaxis is a finite element package that has been developed specifically for the

analysis of deformation and stability in geotechnical engineering projects. The simple

graphical input procedures enable a quick generation of complex finite element models,

and the enchanced output facilitys provide a detailed presentation of computational

results. The calculation itself is fully automated and based on robust numerical

procedures.

During the generation of the mesh, clusters are divided into triangular elements. A

choice can be made between 15-node elements and 6-node elements (Fig. 2.8). The

powerful 15-node elements provides an accurate calculation of stresses and failure loads,

in addition, 6-node triangles are available for a quick calculation of serviceability states.

Figure 2.8: Typical of element, node and stress point during mesh generation a)15-node

element b) 6-node element (Plaxis manual).


29

2.8. Summary

In this chapter, the concepts and applications related to Raft & Pile method have

been reviewed. Previous applications of reinforced embankment on soft ground have

also been reviewed. The main points covered include the following:

1) Most of the methods are using the artificial geosynthetics and/or chemical admixture

as the reinforcement and/or improvement geomaterial. The use of the artificial material

may cause environmental problems in a long term.

2) Since the raft and pile are made of the surplus trees, which are free except for the

transportation fee, which is an economic way. Moreover, since the surplus trees are

removed, it is environmental friendly way to the mountains. Since the timbers are

natural materials, the raft timbers will not give pollute to environment comparing to

other method.

References

1) Al Gore, An inconvenient truth (2007)

2) Atsushi Numata and Sgiakiraos: As for the possibility of the wood use for global

warming, it is the 14th global environment symposium, American Society of Civil

Engineers, and 2006.8.

3) Atsushi Numata: The seventh time environmental ground large amount of money

symposium of outline, American Society of Civil Engineers, and 2007.8 of wooden

pile investigations gathered in Asuwa river

4) Atsushi Numata: About the symposium of the research of the 42nd time

geoengineering of the investigation of the document of the wood use and construction

and the preservation technology of the DS-3 historical ground structure, the material.

Japan Cement Association: Version in cement handbook fiscal year 2005 and p.7,

2005.6.

5) Bolton, M.D. (1992) Design Methods. Wronth Memorial Symposium, Cambridge,

July, Published by Thomas Telford.

6) Bell, J. R., D. R. Greenway & W. Vischer 1977. Construction and analysis of a fabric

reinforced low embankment. Comptes Rendus du Colloque International Sur

l’Empoli des Textiles en Geotechnique, 20-22 Mai, Paris, vol. 1, p.71-75.

7) Hayashi, S., Miura, N., Koumoto, T., Fujikawa, K., Chai, J. C. and Li, X.(1997),


30

Recent developments of geotechnology on soft ground in Kyushu, Proceedings of

the Japan – chaina joint symposium on Recent development of theory & practice in

geotechnology, October 29-30 1997, Shanghai.

8) Holtz, R. D.1990. Design and construction of geosynthetically reinforced

embankment on very soft soils. Proceedings of the international reinfoced soil

conference, Performance of the reinforced soil structures, Glasgow-UK, 10-12 Sep.,

pp. 391-402.

9) Kagoshima Prefecture: Ishibashi memorial exhibition manual」

10) Kyoto Protocal (2007)

11) Lester R. Brown: Plan B 2.0（2006）, W.W. Norton & Company, Ltd, pp. 61-62

12) Jewell, R.A., Milligan, G.W.E., Sarsby, R.W. and DuBois, D.D. (1984) Interactions

between soils and grids. Polymer Grid Reinforcement in Civil Engineering, Thomas

Telford, 18-30.

13) Jewell, R.A., Milligan, G.W.E., (1989) Deformation calculations for reinforced soil

walls. Proc 12th Int Conf Soil Mechanics and Foundation Engineering, Rio de

Janeiro, Vol.2, 1257-62.

14) Jewell, R.A., (1990) Strength and deformation in reinforced soil design. Proc 4th Int

Conference on Geoteciles, Geomembranes and Related Rpducts, The Hague.

15) Jewell, R.A., Brud, H.J. and Milligan, G.W.E. (1992) Predicting the effect of

boundary forces on the behaviour of reinforced soil walls. Wroth Memborial

Symposium, Cambridge, July, Published by Thomas Telford.

16) Jewell, R. A.(1993), Key note lecture: Links between the testing, modelling and

design of reinforced soil . Earth Reinforcement Practice, Ochiai et al (eds) © 1993,

Balkema, Rotterdam.

17) Ladd, C.C. (1991) Stability evaluation during staged construction. ASCE

Geotechnical Journal, Vol. 117, No. 4, April.

18) Leroueil, S. , Magnan, J. P, Tavenas, F. and Wood, D. M. (1990), Embankments on

soft clays, Ellis Horwood.

19) Madhav, M. R., FuKuda, N.(1997), Technical report - embankment . Landmarks in

earth reinforcement Earth Reinforcement Practice, Proc. Int. Symposium, Eds.

Ochiai et al (eds) © 2003, Balkema, Rotterdam.


31

20) Matsuo, M. and Kawamur, K. (1977). “Diagram for construction control of

embankment on soft ground” Soils and Foundations, Vol. 17, No.3, pp. 37-52.

21) Marolo C. A. (1996) “Reinforced soil wall-embankment system on soft foundation

using inextensible and extensible grid reinforcements” , Doctoral thesis,

Department of Civil Engineering, Saga University, Japan.

22) Penta – Ocean Construction 1896-1996

23) Rowe, R.K. and Ho, S.K. (1992) A review of the behaviour of reinforced soil walls.

Earth Reinforcement Practice, Proc. Int. Symposium, Eds. Ochiai, Hayashi and

Otani, Balkema, Vol.2.

24) Tatsuoka, F., K Okahara, M., Tanaka, T., Tani, K., Morimoto, T. and Siddiquee,

M.S.A. (1991) Progressive failure and particle sized effect in bearing capacity of a

footing on sand, Proc Geotechnical Engineering Congress, Boulder, ASTM Special

Publication No. 27, 788-802.

25) Tatsuoka, F., Murata, O. and Tateyama, M.(1992)Permanent geosynthetic-reinforced

soil retaining walls used for railway embankments in Japan. Geosynthetic-reinforced

Soil Retaining Walls, Wu (Ed.), Balkema, 101-130.

26) Yashima, A.(1997) Finite element analysis on earth reinforcement current and

future . Earth Reinforcement Practice, Proc. Int. Symposium, Eds. Ochiai, Yasufuku

and Omine , Balkema, Rotterdam.

List of figures

32

CHAPTER Ⅲ


3.1. General

A new test apparatus was developed to examine the deformation behavior of

embankment reinforced with Raft & Pile foundation during embankment loading and

evaluate their effectiveness (see in Figure 3.1.). Simple index was evaluate

effectiveness of Raft & Pile is purpose.

The laboratory model tests are conducted, in which width of raft and length of pile

and the connecting manner with tie rod for pile line.

3.2. Laboratory test program

The laboratory investigation was conducted on a model ground, which was a

reconstituted Ariake clay. The width of raft and length of pile were varied to investigate

the overall behavior of raft-pile-foundation soil system with varying raft and pile

dimensions.

CHAPTER III SMALL-SCALE MODEL TEST OF RAFT & PILE FOUNDATION

33

3.2.1 Test set up

Two model boxes with dimensions of 900 mm long, 200 mm wide, and 300 mm

high were used in the model tests. The frame and bottom of the boxes are made of steel

and the side wall is made of acryl. In order to minimize side wall friction, latex rubber

membrane coated with grease was used. The latex rubber membrane can be stretched

and rolled to the desired position using a wooden bar. The schematic view of the model

test set up is illustrated in Figure 3.2. The load was applied using a beloframe system

and increased through an air compressor. Figure 3.3 shows the photograph of grid and

circle marked on acryl and latex rubber membrane used for measuring displacements.

Embankment

Raft

Pile

Tie rod

Soft ground

(a) Cross section

Soft Ground

Raft & Pile Foundation

Embankment

Side reapers RaftTie rod

Pile

Soft Ground

Soft Ground

Raft & Pile FoundationRaft & Pile Foundation

Embankment

Side reapersSide reapers RaftRaftTie rodTie rod

PilePile

(b) Isometric view

Figure 3.1: Schematic diagram of embankment on very soft clay supported by Raft &

Piles as foundation (a) cross section (b) isometric view.


34

60mm70mm70mm 70mm

hard glass plate (800mm*250mm)

<<soil box>>

rulers

steel plate185mm* 300mm * 0.6 mm

Sand mat 40 mmdial gauge

handle for rolling up the latex rubber memb rane

draining pipes

300 mm160 mm

900 mm

300m

m

180mm450mm

Loading system

circles marked

grid marked

Figure 3.2: Test set up used in Raft and Pile model test.

CLCLCLCLCLCLCLCLCLCLCLCL

Circle was marked on hard glass plate (fixed)

Grid was marked on latex rubber membrane

Figure 3.3: Photo showing grid and circles marked on the hard glass plate and latex

rubber membrane respectively for measuring displacements.


35

3.2.2. Model ground

The model ground was made through consolidating remolded Ariake clays in the

test box under one-dimensional consolidation. The remolded Ariake clay was taken

from Kawazoe town, Saga, Japan at a depth of 4 m. At this depth, the clay is in a very

soft state with dark gray color. The Ariake clay sample has 46% clay, 49% silt, and 5%

sand. The physical properties are: specific gravity Gs=2.66, natural water content

wn=110%, liquid limit, wL =87.5%, and plasticity index Ip=47.5%. The disturbed clay

sample brought from the site was mixed completely into a slurry state using an electric

mixer. Then, the clay slurry was poured into the test box to a thickness of 260 mm. The

clay was then consolidated in the test set up under a consolidation pressure of 2.8 kPa

for five weeks under a two-way drainage condition until the primary consolidation was

almost complete. In order to reduce the side friction of the box during consolidation,

grease was used on the side walls of the box. After consolidation, the variation of water

content with depth was measured. The results are shown in Figure 3.4(a). It may be seen

that the water content average about 99.5% within a depth of 225 mm

( )125.1200225

===WidthDepth

BD

. Then, by using the cone penetration method the strength

of the model ground was determined and the test results are presented in Figure 3.4(b):

It can be seen that qc has increased from about 25 kPa to about 34 kPa within a depth

of 225 mm ( )125.1200225

==BD .


36

3.2.3. Test procedure

After the model ground was prepared, model embankment was built on the model

ground with sand material for a thickness of about 40 mm. Then, a stainless steel plate

with the thickness of 0.6 mm was placed on the embankment. Load was applied on the

steel plate through a piston cylinder, as shown in Figure 3.2 corresponding to multi

stage of embankment constructions in field. Totally six different tests were conducted:

two cases on the consolidated clay without any improvement; two tests with a raft only

on the top of the clay and two tests on a raft with piles. Table 3.1 presents the

experimental programme. For the two unsupported cases, the width of the foundation of

embankment varied, 360 and 300 mm. For these two cases the rate of loading and the

maximum pressure were also varied (Table 3.1). Similarly, the details of testing for the

raft supported cases and the raft plus piles supported cases are shown in Table 3.1. The

raft consists of square shaped Hinoki Cypress timber reapers of size 5 mm×5 mm×200

mm length, 40 nos making a raft of size of 200 mm × 200 mm. The reapers are

placed side by side to a width of 200mm to form a flexible raft. The piles are 5 mm × 5

mm square shaped Hinoki Cypress timber with two different piles of lengths, 80 mm

and 110 mm pushed to a depth of 70 and 100 mm respectively. The piles were kept in

position using 2 nos of side reapers. Stainless wire with a diameter of 0.45 mm is used

as the tie member to keep the piles and raft in position. Fig 3.5 shows the isometric view

of the whole set up of the raft (made of reapers) and the piles. The load was applied step

by step to simulate multi-staged construction of a road embankment in field conditions.

Figure 3.6 shows the load-time relations for the test cases. For the fast tests, each step of

loading was 1day until the maximum load 30kPa. For the slow tests, each step of

loading was 1 week until the maximum load 30kPa. The vertical displacements at the

ground surface during the loading stage were measured by dial gauges with a sensitivity

of 1/100mm; while the deformations within the clay were monitored by means of the

deformed meshes plotted from frequent photo shots on the latex rubber membrane that

had been marked with grid points (see Figs. 3.2 and 3.3).


37

-250

-200

-150

-100

-50

0

95 96 97 98 99 100 101 102 103 104

(a)

Water content(%)

Elev

atio

ns(m

m)

(a) Variation of water content with depth.

-250

-200

-150

-100

-50

0

5 10 15 20 25 30 35

test No.1 test No.2

qc (kN/m2)

Dep

th(m

m)

(b) Variation of qc with depth (cone penetration test) content with depth

was measured.

Figure 3.4: (a) Variation of water content with depth (b) Variation of qc with depth

(cone penetration test) content with depth was measured.


38

200mm

200mm 70 / 100mm

Side reaper 2 Nosof 5mm×5mm

5mm×5mm×200mm, 40 Nos reapers

5mm×5mm Hinoki Cypress timber pile, of length 70/100 mm

Tie member of 0.4mm

Figure 3.5: Isometric view of raft (made of reapers), piles, side reapers and tie members.

0

5

10

15

20

25

30

0 10 20 30 40 50

SLOWFAST

MT-0, MT-R1 MT-RP1, MT-RP2 MT-0s, MT-R2

Elapsed time (days)

Pres

sure

(kPa

)

Figure 3.6: Embankment loading with time.


39

Table 3.1: Detail for test cases.

Test case Foundation

type

MT-0 without

any support

MT-0s without

any support

MT-R1 with raft

MT-R2 with raft

MT-RP1 with raft & pile

MT-RP2with raft & pile

Width of top of sand mat(mm)

160 120 160 120 160 160

Width of bottom of sand

mat(mm)

360 300 360 300 300 300

Cross section Dimension of reapers (mm)

- - 5 x 5 5 x 5 5 x 5 5 x 5

Thickness of raft (mm)

- - 5 5 5 5

Width of rafts (mm)

- - 200 200 200 200

Dimension of piles (mm)

- - - - 5 x 5 5 x 5

Spacing of pile (mm)

- - - - 5 5

Length of piles (mm)

- - - - 80 110

Installed depth of piles (mm)

- - - - 70 100

Tie rod method - - - - top top

Maximum of embankment loading (kPa)

10 15 30 30 30 30

Load pattern fast slow fast slow fast fast Final state collapsed collapsed Did not

collapseDid not collapse

Did not collapse

Did not collapse


40

3.3. Test results and discussions

The deformation behavior of the ground with and without support was studied to

evaluate the effectiveness of the raft and piles method. The behavior based on vertical

and lateral displacements of the ground are discussed below.

3.3.1. Settlements at embankment center

Figure 3.7 shows the variation of settlements at the embankment center with the

applied pressures. As shown in the figure, the two cases without any raft or/and pile

collapsed at a very low applied pressures of 10 to 15 kPa. For the collapsed cases, large

settlements were noticed in the last stage. For the non-collapsed cases only around 20

mm of settlement ( 10.Width

SettlementB

≅=δ ) occurred under the pressure of 30 kPa. The

raft treatment cases (MT-R1 and MT-R2) showed that even at a loading of 30 kPa, no

sign of collapse was observed. The fast loaded case (MT-R1) resulted in a lower

settlement, being less than 15 mm, at the applied pressure 30 kPa. This is primarily

because no sufficient time was allowed for the pore water pressure to dissipate, i.e. an

undrained condition. For the raft and piles cases, test MT-RP1 (pile length of 70 mm)

resulted in a settlement of 27.2mm, whereas test MT-RP2 (pile length of 100 mm)

resulted in a maximum settlement of 20.32 mm at the maximum applied pressure 30

kPa. When the pile length is longer, the settlement at center of embankment was less as

expected. The above results have demonstrated the effectiveness of the raft – pile

foundation in reducing settlements.

Figure 3.8(a), 3.8(b), and 3.8(c) show the settlement patterns for the fast loaded

test of MT-0 (no raft), MT-R1 (raft only) and MT-RP2 (raft and piles), respectively.

The settlement pattern is shown at different loading stages. Test MT-0 (no raft) in

Figure 3.8(a) shows that the settlement is maximum at center, being 88mm

(Bδ = 44.0

20088

= ) at the end of the test with the maximum applied pressure 15 kPa.

Beyond the embankment toe, the heave is noticed and the heave slowly reduced to zero

at a distance of 400 mm from the center line. For pressures less than 15 kPa no

appreciable settlement was noticed. Test MT-R1 (raft only) in Figure 3.8(b) shows a

significantly reduced settlement at the embankment center line, being 9 mm at the

applied pressure 15 kPa. Most importantly, the embankment did not fail, even at the


41

maximum applied pressure 30 kPa. In addition, there is a reduction in the heaving of the

ground surface beyond the embankment toe, comparing Figure 3.8(b) (raft only) to

Figure 3.8(a) (no raft). Test MT-RP2 (raft and piles) in Figure 3.8(c) shows that no

heave was observed beyond the embankment toe. The settlements were more

concentrated at the embankment center line. Test MT-RP2 measured a maximum

settlement of 20 mm, which was greater than the 15 mm measured in the test MT-R1

(raft only) at the maximum applied pressure 30 kPa. In other words, the settlement is

contained beneath the embankment only and not affecting the adjoining ground surface.

3.3.2. Subsoil movement

The above discussed phenomena can be confirmed from the displacement vectors,

as illustrated in Figure 3.9(a), 3.9(b) and 3.9(c). Figure 3.9(a) shows the subsoil

movement vectors for the case without any support (test MT-0). In this case, the soil

movement is at first downward and then outward towards the embankment toe direction.

Outside the embankment toe, both upward and outward movements were observed. This

result confirms the heaving noticed outside the toe in Fig 3.8(a). For the case with raft

(Fig. 3.9(b)), the subsoil movement is also downward and outward towards the

embankment toe direction. However, the outward movement is much smaller than that

observed in the no support case (Figure 3.9(a)).For the case with the raft-pile, test

MT-RP2 (Figure 3.9(c)), the downward movement is large. However, the outward

movement is very small compared to other cases. These results confirm that the raft-pile

foundation can transfer the embankment loading into deeper into subsoil. Thus, the

raft-pile foundation not only improves the stability of the embankment on the soft

subsoil but also reduces the displacement of the subsoil significantly. In addition, no

ground movement is observed beyond the embankment toe. Longer piles are the most

effective in reducing differential deformation beneath embankment and beyond

embankment toe. This is an important advantage.


42

-- 2.5 5 10 15 20 25 30-0.40

-0.35

-0.30

-0.25

-0.20 MT-0s(slow) MT-R2(slow) MT-RP2(fast) MT-RP1(fast) MT-R1(fast) MT-0F(fast)

collapsed

collapsed

Pressure(kPa)

-0.10

-0.05

0.00

Surfa

ce se

ttlem

ent a

t cen

ter l

ine

(mm

)

Thi

ckne

ss o

f ini

tial g

roun

d (m

m)

Figure 3.7: Settlement at center of embankment.


43

0 10 20 30 40 50-100-80-60-40-20

020406080

100LCLC

end of the test 10kPa, collapsed 5kPa 2.5kPa Initial

Dis

plac

emen

t(mm

)

Distance from center line(mm)

(a) without any support MT-0.

0 100 200 300 400 50020

15

10

5

0

5

10

15

20LC

30kPa 25kPa 20kPa 10kPa 5kPa 2.5kPa initial

Disp

lace

men

t(mm

)


raft

LC

raft

LC

0 100 200 300 400 50020

15

10

5

0

5

10

15

20LC

30kPa 25kPa 20kPa 10kPa 5kPa 2.5kPa initial

Disp

lace

men

t(mm

)


raft

LC

raft

LC

(b) support with raft MT-R1.

0 100 200 300 400 500

-20

-10

0

10

20

pileraft

LCLC

30kPa 25kPa 20kPa 15kPa 10kPa 5kPa initial

Disp

lace

men

t(mm

)


(c) support with raft-pile MT-RP2

Figure 3.8: Surface vertical displacement for case (a) without any support MT-0 (b)

support with raft MT-R1(c) support with raft-pile MT-RP2.


44

0 100 200 400 mm

200

0

50

100

150

250

0 100 200 300 400 mm

300

Note:displacementvector

initial

end

oΔ


initial

end

oΔ

(a)

0 100 200 400 mm

200

0

50

100

150

250

0 100 200 300 400 mm

300


initial

end

oΔ


initial

end

oΔ

0 100 200 400 mm

200

0

50

100

150

250

0 100 200 300 400 mm

300


initial

end

oΔ


initial

end

oΔ

(a)

(a) without any support MT-0.

50

0 100 200 300 400（mm）

0 100 200 300

200

0

100

150

250

400 (mm)

50

0 100 200 300 400（mm）

0 100 200 300

200

0

100

150

250

400 (mm)


initial

end

oΔ


initial

end

oΔ

(b)

50

0 100 200 300 400（mm）

0 100 200 300

200

0

100

150

250

400 (mm)

50

0 100 200 300 400（mm）

0 100 200 300

200

0

100

150

250

400 (mm)


initial

end

oΔ


initial

end

oΔ

50

0 100 200 300 400（mm）

0 100 200 300

200

0

100

150

250

400 (mm)

50

0 100 200 300 400（mm）

0 100 200 300

200

0

100

150

250

400 (mm)


initial

end

oΔ


initial

end

oΔ

(b)

(b) support with raft MT-R1.

0 100 200 300

200

0

100

150

250

0 100 200 300 400 (mm)

50

400 (mm)0 100 200 300

200

0

100

150

250

0 100 200 300 400 (mm)

50

400 (mm)

(c)


initial

end

oΔ


initial

end

oΔ

0 100 200 300

200

0

100

150

250

0 100 200 300 400 (mm)

50

400 (mm)0 100 200 300

200

0

100

150

250

0 100 200 300 400 (mm)

50

400 (mm)

(c)


initial

end

oΔ


initial

end

oΔ

(b) support with raft-pile MT-RP2.

Figure 3.9: Displacement in ground (a) without any support MT-0, collapsed under 10

kPa (b) support with raft MT-R1, under 30 kPa (c) support with raft-pile

MT-RP2, under 30 kPa.


45

-250

-200

-150

-100

-50

0

0 2 4 6 8 10 12

MT-R2 MT-RP2 MT-RP1 MT-R1

Lateral displacment from point A(mm)

Dep

th (m

m)

Figure 3.10: Lateral displacements at the end of final loading of 30 kPa for the raft cases

and the raft + piles cases at the edge of raft location.

3.3.3. Lateral displacements

Figure 3.10 shows the lateral displacement of the ground with depth at the end of

final loading of 30 kPa for the raft cases and the raft + piles cases. The maximum lateral

displacement of 8 mm occurs for raft foundation (MT-R1) at a depth of 35 mm and

steadily decreases to zero at a depth of 85 mm. For the raft + piles cases, test MT-RP1

shows the maximum lateral displacement of 7mm occurs at a depth of 45 mm and

steadily reduces to zero at a depth of 160 mm. For test MT-RP2 the maximum lateral

displacement of 3mm occurs at a depth of 70 mm and steadily reduces to zero at a depth

of 160 mm. For the raft case of MT-R2, the lateral displacement, increases with depth

and becomes constant at 4 mm of lateral displacement up to a depth of 150 mm, below

which the displacement steadily decreases to zero at a depth of 230 mm. It has been

observed that the raft + piles cases have considerably reduced the lateral displacement

beneath.

Figure 3.11 shows the vertical displacement at the center of the embankment for

the two test cases MT-RP1 (pile length of 70 mm) and MT RP-2 (pile length of 100

mm), i.e. for two different depths of embedment. The test with longer piles settles


46

slightly more than the test with shorter piles. This is because the lateral movements of

the soil are significantly reduced when the longer piles are used. However, the

maximum vertical settlement at the surface for tests MT-RP2 was 20.32 mm which is

less than the 27.2mm measured in the test MT-RP1. It can be seen from Figure 3.8(c)

that the surface settlement beyond the embankment toe for test MT-RP2 is almost zero

without any heave, which is different from the cases of MT- 0 (Figure 3.8(a)) and

MT-R1 (Figure 3.8(b)). A combined study of Fig 3.11 and 3.8 reveals that raft + piles

treatment is effective. It is noteworthy that there were no cracks on the top of the soft

clay for the test case of MT-RP2. However under an embankment loading of 25 kPa in

the test MT-RP1 (pile length 70 mm), tension cracks appeared near the toe of the sand

mat at top for embedded depth of 70 mm. It is clear from the overall study that longer

piles are more effective in controlling deformations.

-250

-200

-150

-100

-50

0

0 1 2 3 4 5 6 7 8 9

Pile embedded length = 100mmPile embedded length = 70mm

Vertical displacement(mm)

Dep

th (m

m)

Figure 3.11: Vertical displacement at center of embankment for raft + pile.

For construction of highway embankments around the Ariake sea gulf, the

important design considerations include allowable embankment settlement and ground

deformation. Matsuo and Kawamura (1977) proposed a ratio of vertical displacement to

horizontal displacement for control of embankment construction on soft ground.

Comparing the case of raft and piles, it is clear that piles can significantly reduced


47

lateral movement in the Ariake clay ground, concentrating the lateral displacement

within an area bounded by the piles. For the test carried out, the displacement ratio of

vertical displacement at the center to lateral displacement of the raft edge is a good

measure of the overall performance of the test cases. The displacement ratio, (δh,max/

δv,max) i.e. ratio of maximum horizontal displacement , δh,max to maximum vertical

displacement , δv,max at the maximum applied pressure for the three cases of MT-R1,

MT-RP1 and MT-RP2 are presented in Fig 3.12. It can be seen that the least ratio of less

than 0.2 is observed for MT-RP2, and it is most effective. However, the other two cases

MT-RP1 and MT-R1 are slightly less effective as MT-RP2.

Figure 3.13 compares to the ratio of δh,max to (δv,max- δheave,max) for two cases of MT-

RP1 and MT-RP2 at a loading pressure of 30 kPa. It is clear that MT-RP2 performs

much better.

0.0

0.2

0.4

0.6

0.8

1.0

max,hδmax,vδ max,hδmax,vδ max,hδmax,vδ

MT-R1

30kPa

Fast test

B=360mm

MT-RP1

30kPa

Fast test

B/L=200/70

MT-RP2

30kPa

Fast test

B/L=200/100

-

+ Eff

ectiv

enes

s

Figure 3.12: Effectiveness of Raft & Pile Foundation (L = installed depth of piles).


48

0.0

0.1

0.2

0.3

B/L32

max,hδmax,vδ

max,heaveδ

max,hδmax,vδ

max,heaveδ

MT-RP2

30kPa

Fast test

B/L=200/100

MT-RP1

30kPa

Fast test

B/L=200/70

-

+ Eff

ectiv

enes

s

Figure 3.13: Effectiveness of longer pile.

3.4. Summary

Based on the detailed laboratory investigations the following conclusions can be

drawn:

1) In the model test, an embankment loading on the soft Ariake clay without

support resulted in failure at very low pressures of 10 to 15kPa. With the

addition of a raft or a raft plus piles, the failure was not observed even an

applied pressure of 30 kPa.

2) With the addition of a flexible raft, both vertical settlements and lateral

deformations have reduced significantly. No failure was observed with an

increase in the applied pressure up to 2 to 3 times the failure pressure in the

unsupported case.

3) The stability of the raft foundation could be further improved by the addition of

piles at the end of the raft. This additional support removes almost completely

the lateral heave. The vertical deformations have also reduced significantly. The

raft plus piles supports can transfer the embankment loading deeper into the

subsoil. The lateral movement at the embankment toe is also reduced to almost

negligible levels.


49

4) The length of piles has significant influence on the vertical and lateral

deformations of the ground. The raft with its edges supported with piles is found

to be very effective in reducing differential deformations both beneath the

embankment and beyond the embankment toe. The experimental results clearly

demonstrate that even in very soft soils, locally available timber at minimal cost,

could be effectively utilized to safely carry embankment loading common in the

Ariake gulf area. The results of the model test need to be validated by full-scale

tests in the field.

References

1) Chai, J.C., Miura, N., and Shen, S.L. (2002). Performance of embankments with and

without reinforcement on soft subsoil. Canadian Geotechnical Journal, 234-435.

2) Karl.T, Ralph B.P. and Gholamreza M. (1967). “Soil mechanics in engineering

practice” John Wiley and Sons, New York.

3) Hayashi, S., Du, Y.J. (2005). Geotechnical analysis of Mizuki embankment remains.

Soils and Foundations, 45(6), 43-53

4) Matsuo, M. and Kawamura, K. (1977), Diagram for construction control of

embankment on soft ground Soils and Foundations, 17(3), 37-52.

5) Ochiai, H., Hayashi, S., Umezaki, T., and Otani, J. (1991). Model test on sheet-pile

countermeasures for clay foundation under embankment. Developments in

Geotechnical Aspects of Embankment, Excavations and Buried Structures,

Netherlands, 277-291.

6) Shen, S.L., and Miura, N. (2001). A technique for reducing settlement difference of

road on soft clay. In Computer Methods and Advances in Geomechanics, Proc. 10

IACMAG, 2, 1391-1394.

7) Toyosato, E., Takizawa, Y., Tokutake, H. (2003). Treatment for the settlement of the

embankment beside bridge abutment of Hokurikku Expressway by using EPS. The

Foundation Engineering & Equipment, 31(6), 59-64 (in Japanese).

8) T.S. Nagaraj and N. Miura. (2001). “Soft clay behaviour: Analysis and Assessment”

A.A. Balkema, Rotterdam, pp.1-281.

9) Wood, D. M. (2004), “Geotechnical modelling” Applied Geotechnics Volume 1,

Taylor & Francis Group, New York, USA.

50

CHAPTER IV


4.1. General

In this study scale model of a new pattern of raft that consisted of circular timber

of diameter 6mm and binded together by using screws, in order to mimick the true

physical flexural behavior of an actual raft.

Totally twenty test cases (Table. 1) of tree-point bending test on composites Raft

were conducted (Fig. 4.1) to investigated firstly, the effects of the number layers and the

patterns of raft element. Secondly, the tests were used to developed effective patterns

for raft composites applicable to design and construction. Thirdly, to link the results

between short timber raft composites (field condition) and long timber raft composites

(simulated FEM analysis).

CHAPTER IV INVESTIGATION ON RAFT COMPOSITES FOR CONSTRUCTION

51

4.2. Bending test for raft composites

4.2.1 Test set up

The test set up shown in Fig. 4.1 monitored displacements at the center of the

model test embankment by using displacement transducers (one for each specimen).

The applied loads using standard load cells of odometer tests start from 0.1kg, 0.2 kg,

0.4 kg, 0.8 kg and 1.6 kg. At each loading level data were collected.

Table 4.1: Detail for test cases

Examination

13n13n1nB-2( 7.8, 6.6, 3.0)1R15cm, 11n-13L⑮

13n13n1nB-3( 7.8, 6.6, 3.0)1R18cm,11n-13L⑯

13n13n1nB-4( 7.8, 6.6, 3.0)1R21cm,11n-13L⑰

11n9n3nB-1( 7.8, 6.6, 4.2)3R12cm,11n -9L⑱

3n

6n

2n

2n

2n

2n

13n

13n

13n

13n

13n

2n

2n

2n

2n

2n

No. axial member

2n3nA-1( 15, 6.6, 3.0)3RL,11n –2L①

2n2.5nA-1( 15, 6.6, 3.0)2.5RL,11n –2L②

2n2nA-1( 15, 6.6, 2.4)2RL,11n –2L③

2n1.5nA-1( 15, 6.6, 2.4)1.5RL,11n –2L④

2n1nA-1( 15, 6.6, 1.8)1RL,11n –2L⑤

11n3nB-1( 7.8, 6.6, 4.2)3R12cm,11n -3L⑳

B-1

A-5

A-4

A-3

A-2

B-1

B-1

B-1

B-1

B-1

Pattern of Raft

( 7.8, 6.6, 4.2)

( 15, 9.0, 1.8)

( 15, 8.4, 1.8)

( 15, 7.8, 1.8)

( 15, 7.2, 1.8)

( 7.8, 6.6, 3.0)

( 7.8, 6.6, 3.6)

( 7.8, 6.6, 3.6)

( 7.8, 6.6, 4.2)

( 7.8, 6.6, 4.2)

Dimension of 1 Unit model (cm)

( X, Y, t )

Model chartSide view

15n1n1RL,15n -2L⑭

11n3n3R12cm,11n-6L⑲

14n1n1RL,14n -2L⑬

12n1n1RL,12n -2L⑪

13n1n1RL,13n -2L⑫

⑩

⑨

⑧

⑦

⑥

Number

13n1n1R12cm,11n -13L

1.5n

2n

2.5n

3n

No. layer of Raft

13n

13n

13n

13n

No. bending member

Model chartTop view

1.5R12cm,11n -13L

2R12cm,11n -13L

2.5R12cm,11n -13L

3R12cm,11n -13L

Model chartFront viewCaseExamination

13n13n1nB-2( 7.8, 6.6, 3.0)1R15cm, 11n-13L⑮

13n13n1nB-3( 7.8, 6.6, 3.0)1R18cm,11n-13L⑯

13n13n1nB-4( 7.8, 6.6, 3.0)1R21cm,11n-13L⑰

11n9n3nB-1( 7.8, 6.6, 4.2)3R12cm,11n -9L⑱

3n

6n

2n

2n

2n

2n

13n

13n

13n

13n

13n

2n

2n

2n

2n

2n

No. axial member

2n3nA-1( 15, 6.6, 3.0)3RL,11n –2L①

2n2.5nA-1( 15, 6.6, 3.0)2.5RL,11n –2L②

2n2nA-1( 15, 6.6, 2.4)2RL,11n –2L③

2n1.5nA-1( 15, 6.6, 2.4)1.5RL,11n –2L④

2n1nA-1( 15, 6.6, 1.8)1RL,11n –2L⑤

11n3nB-1( 7.8, 6.6, 4.2)3R12cm,11n -3L⑳

B-1

A-5

A-4

A-3

A-2

B-1

B-1

B-1

B-1

B-1

Pattern of Raft

( 7.8, 6.6, 4.2)

( 15, 9.0, 1.8)

( 15, 8.4, 1.8)

( 15, 7.8, 1.8)

( 15, 7.2, 1.8)

( 7.8, 6.6, 3.0)

( 7.8, 6.6, 3.6)

( 7.8, 6.6, 3.6)

( 7.8, 6.6, 4.2)

( 7.8, 6.6, 4.2)

Dimension of 1 Unit model (cm)

( X, Y, t )

Model chartSide view

15n1n1RL,15n -2L⑭

11n3n3R12cm,11n-6L⑲

14n1n1RL,14n -2L⑬

12n1n1RL,12n -2L⑪

13n1n1RL,13n -2L⑫

⑩

⑨

⑧

⑦

⑥

Number

13n1n1R12cm,11n -13L

1.5n

2n

2.5n

3n

No. layer of Raft

13n

13n

13n

13n

No. bending member

Model chartTop view

1.5R12cm,11n -13L

2R12cm,11n -13L

2.5R12cm,11n -13L

3R12cm,11n -13L

Model chartFront viewCase

3R-13L

3R-9L

3R-6L

3R-3L

3R-13L

3R-9L

3R-6L

3R-3L

2R-13L3R-13L

Eff

ect o

fL

( No.

axi

al m

embe

r)E

ffec

t of R

(No.

laye

r of

Raf

t)

1R-13L3R-13L

1R-13L3R-13L

2R-13L3R-13L

3R-13L

2.5R-13L

1.5R-13L

2R-13L

1R-13L

3R-13L

3R-13L

3R-13L

3R-13L 1R-13L1R-13L

3R-13L

3R-13L

1R-13L1R-13L

1R-13L1R-13L

1R

Eff

ect o

f R

aft p

atte

rnE

ffec

t of n

(N

o. o

f ben

ding

M

embe

r)

*aRc,dn-bL a; No. of layers Raft (R)b; No. of axial element (L)c; pattern of Raft member

d; N0. of bending membern; timber diameter 0.6cm


52

(a)

(b) (c)

Figure 4.1: Test set up (a) Dimensions of test set-up (b) Overview of test set-up (c)

end of testing.

Detail of raft layer

Detail of scale model

(Model, Prototype)

a) Pattern A-1 (91 cm, 31m)

(6.6

cm

, 2.2

m)

d) Pattern A-4

e) Pattern A-5

(8.4

cm

, 2.8

m)

(9.0

cm

, 3m

)

(91 cm, 31m)

(91 cm, 31m)

(91cm, 31m)

(6.6

cm

, 2.2

m)

(91cm, 31m)(6.6

cm

, 2.2

m)

(91cm, 31m)(6.6

cm

, 2.2

m)

(91cm, 31m)(6.6

cm

, 2.2

m)

(91cm, 31m)

(6.6

cm

, 2.2

m)

(91cm, 31m)(6.6

cm

, 2.2

m)

(91cm, 31m)(6.6

cm

, 2.2

m)

(91cm, 31m)(6.6

cm

, 2.2

m)

f) Pattern B-1

g) Pattern B-2

h) Pattern B-3

i) Pattern B-4

(21cm, 7m)(18cm, 6m)

(15cm, 5m)(12cm, 4m)

(91cm, 31m)

b) Pattern A-2 (91 cm, 31m)

(7.2

cm

, 2.4

m)

c) Pattern A-3 (91 cm, 31m)

(7.8

cm

, 2.6

m)

Figure 4.2: Detail of scale model pattern raft.

Weight

90cm


53

Cross section of embankment

raft

Width of raft (91cm, 31m)

Cross section of embankment

raft

Width of raft (91cm, 31m)

Dimension of raft

Circle shape diameter (0.6cm, 20cm)

Dimension of raft

Circle shape diameter (0.6cm, 20cm)

Top View Front View Side View

Ｘ

Ｙｔｔ

ＸＹ

Detail of 1 unit of Raft composites

Top View Front View Side View

Ｘ

Ｙｔｔ

ＸＹ

Detail of 1 unit of Raft composites

Figure 4.3: Detail of scale model.

4.3 Test result and discussions

4.3.1 General Observations

The flexural response of composites raft during the test for each test case is shown

in detail from Fig. 4.4 to Fig. 4.12.

Figure 4.4: Composites Raft case 1R12cm,11n-13L (side view).


54

Figure 4.5: Composites Raft case 1.5R12cm,11n-13L (side view).



55

Figure 4.7: Composites Raft case 2.5R12cm,11n-13L (side view).



56

Figure 4.9: Composites Raft case 3R12cm,11n-13L (Top view).

Figure 4.10: Composites Raft case 3R12cm,11n-9L (Top view).


57

Figure 4.11: Composites Raft case 3R12cm,11n -6L (Top view).

Figure 4.12: Composites Raft case 3R12cm,11n 3L (Top view).


58

4.3.2 Effect of Raft layer

Figure 4.13 shows the effect of the number of raft layer, pattern A (long timber)

which controlled the same axial raft member. The result indicated that for case 3RL,11n

-13L(No.①) has the highest rigidity, the deflection of which is about 7.48mm under the

loading of 3.25 kg. For case 1RL,11n -13L(No.⑤) has the lowest rigidity with deflection

of around 31.59mm under 3.15 kg loading.

Figure 4.14 shows the effect of the number of raft layer pattern B (short timber)

which controlled the same axial raft member. The case 3R12cm,11n -13L(No.⑥) resulted to

the highest rigidity with deflection of about 8.78mm under loading 3.25 kg. For case

1R12cm,11n -13L(No.⑩) has the lowest rigidity, deflection of about 17.48mm under the

same loading level.

4.3.3 Effect of bending members

Figure 4.15 shown the effect of the number of bending members of raft pattern A

(long timber) which has the same one layer of raft. The result clearly shown that 1RL,15n

-2L(No.⑭) has the highest rigidity with deflection of about 20.26mm under loading

3.25 kg while 1RL,11n－2L（No.⑤）has the lowest rigidity with deflection of about

31.59mm.

0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5048

12162024283236 No.5(case 1RL,11n-2L)

No.4(case 1.5RL,11n-2L)No.3(case 2RL,11n-2L)No.2(case 2.5RL,11n-2L)No.1(case 3RL,11n-2L)

Def

lect

ion(

mm

)

Load(kg)

Figure 4.13: Effect of No. of Raft layer pattern A (long timber).


59

0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5048

12162024283236 No.10(case 3RL,11n-2L)

No.9(case 2.5RL,11n-2L)No.8(case 2RL,11n-2L)No.7(case 1.5RL,11n-2L)No.6(case 1RL,11n-2L)

Def

lect

ion(

mm

)

Load(kg)

Figure 4.14: Effect of No. of Raft layer pattern B(short timber).

0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5048

12162024283236 No.14(case 1RL,15n)

No.13(case 1RL,14n) No.12(case 1RL,13n) No.11(case 1RL,12n) No.5(case 1RL,11n)

Def

lect

ion(

mm

)

Load(kg)

Figure 4.15: Effect of bending member raft pattern A.

4.3.4 Effect of Raft pattern

Figure 4.16 shown the effect of raft pattern A (short timber) which the same one

layer of raft composites. ZThe case of 1R12cm,11n－2L（No.⑩) resulted to highest rigidity,

deflection of about 17.48mm while 1R21cm,11n－2L（No.⑰） has the lowest rigidity,

deflection of about 24.41mm under loading level of 3.25 kg.


60

4.3.5 Effect of axial member

Figure 4.17 shown the effect of the axial member raft pattern A (short timber)

which has three layers of raft composites. The result showed that the most rigid case is

3R12cm,11n－13L（No.⑥） with the deflection of about 8.78mm while 3R12cm,11n－3L

（No.⑳） resulted to lowest rigidity with deflection of about 15.59mm under loading

level of 3.25 kg.

0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5048

12162024283236 No.17(case 1R21cm,11n-13L)

No.16(case 1R18cm,11n-13L) No.15(case 1R15cm,11n-13L) No.10(case 1R12cm,11n-13L)

Def

lect

ion(

mm

)

Load(kg)

Figure 4.16: Effect of Raft pattern.

0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5048

12162024283236 No.20(case 3R12cm,11n-3L)

No.19(case 3R12cm,11n-6L) No.18(case 3R12cm,11n-9L) No.6(case 3R12cm,11n-13L)

Defle

ction

(mm

)

Load(kg)

Figure 4.17: Effect of No. axial member for raft composites pattern B (short timber).


61

Combined studies to evaluate the effective of raft composites in terms of the ratio

of deflection(mm) to load(kg), show that for case 3RL,11n－2L（No.①） is the highest

rigidity, deflection of about 7.48mm under loading 3.25 kg as shown in Fig. 4.18.. For

case 3R12cm,11n－13L（No.⑥） the deflection is about 8.78mm.

No.1(case 3RL,11n -2L)No.2(case 2.5RL,11n -2L)

No.3(case 2RL,11n -2L)No.4(case 1.5RL,11n -2L)

No.5(case 1RL,11n -2L)No.6(case 3R12cm,11n -13L)

No.7(case 2.5R12cm,11n -13L)No.8(case 2R12cm,11n -13L)

No.9(case 1.5R12cm,11n -13L)No.10(case 1R12cm,11n -13L)

No.11(case 1RL,12n -2L)No.12(case 1RL,13n -2L)No.13(case 1RL,14n -2L)No.14(case 1RL,15n -2L)

No.15(case 1R15cm,11n -13L)No.16(case 1R18cm,11n -13L)No.17(case 1R21cm,11n -13L)

No.18(case 3R12cm,11n -9L)No.19(case 3R12cm,11n -6L)No.20(case 3R12cm,11n -3L)

0 5 10Deflection (mm) / weight (kg)

Test case

Effectiveness+ -

Figure 4.18: Evaluate the effect of pattern of Raft composites for constructions.


62

4.4 Links between the bending tests, FEM analysis, and design of Raft composites

The results of laboratory model testing and numerical modeling were used to

examine in order to develop design and construction guidelines. The Links between the

composites short timber Raft pattern A that simulated field conditions and composites

long timber Raft pattern B that simulated FEM 2D analysis in chapter V (Fig 4.19 and

Fig. 4.20) are first examined. The delection of beam is given by the equation below

EIPl

16=δ ………………….. (1)

where δ = deflection (m), P=load (kg), l = span length (m), EI=bending stiffness.

Rearranging Equation 1, the rigidity can be calculated:

δ16

PlEI = …………………..(2)

The results of the calculations using the above equation and from bending test results is

shown in Fig. 4.19

0 1 2 3 40

1

2

3

4

5

6

7

8

compoites long timber Raft compoites short timber Raft

y=-1.37x+7.42 ;R=-0.99226

y=-2.258x+9.133 ;R=-0.97955

Def

lect

ion(

mm

)/Loa

d(kg

)

No. of layer Raft0 1 2 3 4

0

1

2

3

4

5

6

7

8

compoites long timber Raft compoites short timber Raft

y=-1.37x+7.42 ;R=-0.99226

y=-2.258x+9.133 ;R=-0.97955

Def

lect

ion(

mm

)/Loa

d(kg

)

No. of layer Raft

Figure 4.19: Considered effect of short timber and long timber for composites Raft.


63

n=0.2m5n=1.0m

Timber Raft Φ 0.20m

Width of Raft 32m

Figure 4.20: Detail for calculation EI of Raft (2D FEM analysis).

and then;

⎟⎟⎠

⎞⎜⎜⎝

⎛=⎟⎟

⎠

⎞⎜⎜⎝

⎛

timberlong

timbershort

FEM

design

EIEI

EIEI

………………….. (3)

given example, Figure 4.19 for 2 layers composites Raft 1=⎟⎟⎠

⎞⎜⎜⎝

⎛

timberlong

timbershort

EIEI

so from

equation (3), we may say FEMdesign EIEI = .

Next, the relation of EI between design and analysis in term of layer Raft as shown in

Table 4.2

Table 4.2: Purpose relation of EI between design and analysis in term of layer Raft

EIdesign = 0.8EIFEM3

EIdesign = 0.8EIFEM2.5



Purpose relation of EINo. of layer Raft





Purpose relation of EINo. of layer Raft


64

4.5 Summary

This chapter presented the results of twenty test cases of 3-point bending tests on

composites raft. Based on the test results the following conclusions can be drawn:

1) Composites raft very flexural response that can be use as flexible foundation

support road embankment.

2) Both the number of raft layers and the number of axial raft elements were

significant on the rigidity of raft composites.

3) Links between short timber Rafts (simulated field condition) and long timber

Raft (simulated FEM analysis) have been established. They provide

relationships between the designed rigidity (EI) and FEM analysis in terms of

the number of layers of Raft.

References

1) Fam, A. Z., and Rizkalla, S. H. (2002), Flexural behavior of concrete- filled

fiber-reinforced polymer circular tubes. J. Compos. Constr.m., 6(2), 123-132.

2) Teng, J. G., Chen, J. F., Smith, S. T, and Lam, L.(2002). FPR strengthened RC

structure, John Wiley & Sons Ltd., New York.

3) Dieterich, J. S.(1977), Scale models in engineering, Calspan Corporation, Buffalo.

New York.

4) Wood, D. M. (2004), “Geotechnical modelling” Applied Geotechnics Volume 1,


65

CHAPTER V

FINITE ELEMENT MODELING OF RAFT & PILE FOUNDATION ON SOFT GROUND

5.1 General

In order to verify the effectiveness of Raft & Pile foundation with various analysis

conditions, finite element analysis was done. FEM simulation embankment constructed

over soft Ariake clay ground using a computer code PlaxisV.8. Estimated the residual

settlement of the road surface for an embankment heights of 5.5m. The design and

construction guidelines of using Timber Raft & Pile foundation to support embankment

on soft clay are also given.

5.2 Numerically examining an embankment on soft ground support with Raft

&Pile foundation

5.2.1 Problem considered and numerical model

A very soft and thick alluvial clay layer deposit in Saga plain. This clay has the

properties of high compressibility and high sensitivity. The depth of the layer ranges 10

to 20 meters (Miura et al 2006).

Figure 5.1(a) shows a location of boring test for road embankment in Saga

Prefecture. Fig. 5.1(b) shows a cross section of the deposits at boring test site and Fig.

CHAPTER V FINITE ELEMENT MODELING OF RAFT & PILE FOUNDATION ON SOFT GROUND

66

5.1(c) shows soil profile of Kaseminami and Kubota. For this case soft ground layer

thickness varies from 9 to 11 m.


settlement and lateral deformation will occur. These deformations will be transferred to

the surrounding ground surface and subsoil and will damage the adjoining buildings and

agricultural lands along the highways.

(a)

Kubota

Kaseminami

(b)

KubotaKaseminami

Soft ground

(c)

KubotaKaseminami

Soft ground

Figure 5.1: At test site in Saga Prefecture, Japan (a) Boring test locations (b) Cross

section (c) Soil profile of Kaseminami and Kubota.

Referring to Fig.5.1, in this present study examines the construction of highway

embankment support with Raft &Pile foundation on a 11m deep soft clay ground.

Figure 5.8 shows a cross section of road embankment. The embankment is 12m wide


67

top and 5.5m high. The slopes have a slope of 1:1.8. The embankment itself is

composed of loose sandy soil. The subsoil consists of 11.0m of soft soil as mentioned

before. The upper 1.0m of this soft soil layer is modelled as a weather crust and the

lower 10m as clay. The phreatic level concides with the original ground surface. Under

the soft soil layers there is a sand layer, which is included in the model.

5.2.2 Boundary conditions and model parameters

Plane strain condition was assumed. The model has dimensions of 30 m deep from

ground surface and 80 m wide. The soft layer is about 11 m, the top weather crust(B) is

about 1.0m, followed by the soft clay layer, Ac2, with thickness of 10 m. A sand layer,

As2 below Ac2, with thickness of 5m underlying a thick and dense sand layer(Ds). The

displacement boundary conditions were as follows: at bottom, both vertical and

horizontal displacements were fixed, and for left and right vertical boundaries, the

horizontal displacement was fixed while the vertical displacement was free (see Fig.5.2).

The adopted drainage boundary conditions were as follow: the ground surface and

bottom line (sand layer) were drained. The left and right boundaries were drained. The

mechanical behavior of the clay layers was represented by soft soil model and the sand

layers were assume to be elastic. The model parameters for subsoil based on the work

by Chai et al (1999) and from other literature are listed in Table 5.1. Timber for Raft &

Pile foundations were also assumed to be elastic materials (K.J.Kim,

WWW.Ihitasca.co.th). The groundwater level was the same with ground surface level.

The mechanical property of the fill material was represented by Mohr-Coulomb. The

detail of the numerical simulation for embankment are summarized in Table 5.2

( 8cases,H=5.5ｍ). Deformation behavior of the ground was studied to evaluate the

effectiveness of the raft and pile method of the ground improvement. The vertical and

lateral displacement behavior of the ground is the focus of the following discussions.


68

Soft clay groundthickness11m

Construction fill thickness 5.5m

12m

1:1.8

32m

Sand

CL

Dense Sand

Weather crust 1.0m

Road embankment

Soft clay

RaftPile Pile

Figure 5.2: Situation of road embankment on soft soil.

Embankment

With 2R6P

(5.5m)

CL

(m)

(m)

0.00 10.00 20.00 30.00 40.00 50.00 60.00 70.00 80.00

0.00

10.00

-10.00

-20.00

-30.00

CLCL

(m)

(m)

0.00 10.00 20.00 30.00 40.00 50.00 60.00 70.00 80.00

0.00

10.00

-10.00

-20.00

-30.00

Embankmentloading rate

0

5.5m

Loadfill amount

(m) 27days

Weathered crust (B), 0m-1m

Ariake clay (Ac2), 1m-11m

Sand (As2), 11m-16m

Dense Sand (Ds), 16m-30m

Figure 5.3: Typical finite element mesh for embankment height 5.5m.


69

Tabl

e 5.

1 M

ater

ial P

rope

rties

-1

0.25

056

0.25

056

0.00

228

0.00

989

[m/d

ay]

k x

-1

0.25

056

0.25

056

0.00

152

0.00

657

[m/d

ay]

k y

-30

--

2525

[ °]

ϕ

--

--

1.2

4[-

]O

CR

-1

--

55

[kN

/m²]

c

0.33

0.30

0.20

0.20

0.3

0.25

[-]

ν

1.0E

088.

0E03

3.0E

041.

5E04

--

[kN

/m²]

E ref

-1

11

2.5

2.0

[-]

e init

--

--

0.02

50.

0083

[-]

κ∗

--

--

0.25

0.08

3[-

]λ∗

4.70

16.0

19.0

015

.50

14.5

015

.0[k

N/m

³]γ u

nsat

Line

arEl

astic

Moh

r C

oulo

mb

Line

arEl

astic

Line

arEl

astic

Soft-

Soil

Soft-

Soil

Mod

el ty

pe

timbe

rFi

llD

ense

Sand

(D

s)

16m

-30m

Sand

(As2

)

11m

-16m

Aria

kecl

ay(A

c2)

1m-1

1m

Wea

ther

ed c

rust

(B)

0m-1

m

Mat

eria

ls(1

)

-1

0.25

056

0.25

056

0.00

228

0.00

989

[m/d

ay]

k x

-1

0.25

056

0.25

056

0.00

152

0.00

657

[m/d

ay]

k y

-30

--

2525

[ °]

ϕ

--

--

1.2

4[-

]O

CR

-1

--

55

[kN

/m²]

c

0.33

0.30

0.20

0.20

0.3

0.25

[-]

ν

1.0E

088.

0E03

3.0E

041.

5E04

--

[kN

/m²]

E ref

-1

11

2.5

2.0

[-]

e init

--

--

0.02

50.

0083

[-]

κ∗

--

--

0.25

0.08

3[-

]λ∗

4.70

16.0

19.0

015

.50

14.5

015

.0[k

N/m

³]γ u

nsat

Line

arEl

astic

Moh

r C

oulo

mb

Line

arEl

astic

Line

arEl

astic

Soft-

Soil

Soft-

Soil

Mod

el ty

pe

timbe

rFi

llD

ense

Sand

(D

s)

16m

-30m

Sand

(As2

)

11m

-16m

Aria

kecl

ay(A

c2)

1m-1

1m

Wea

ther

ed c

rust

(B)

0m-1

m

Mat

eria

ls(1

)

Not

e: (1

) Cha

iet a

l 199

9, e

xcep

t for

Fill

(Pla

xis)

and

tim

ber (

WW

W.h

itasc

a.co

m)


70

Tabl

e 5.

2:D

etai

l for

num

eric

al si

mul

atio

n of

em

bank

men

t (8c

ases

, H=5

.5m

)

----

(0.2

m, 0

)

0.2

(1)

32m

Circ

le sh

ape

diam

eter

(0.2

m)

With

Raf

t(1

R)

Emba

nkm

ent h

eigh

t 5.5

m

---

(0.1

7m,

spac

ing0

.17m

)

(0.2

m, n

o sp

acin

g)

0.37 (2)

32m

Circ

le sh

ape

diam

eter

(0.1

7m,

0.2m

)

With

Raf

t(1

.5R

)

---

(0.1

7m, n

o sp

acin

g)

(0.2

m, n

o sp

acin

g)

0.37 (2)

32m

Circ

le sh

ape

diam

eter

(0.1

7m,

0.2m

)

With

Raf

t(2

R)

7m

no sp

acin

g

Circ

le sh

ape

diam

eter

0.2

m

-

(0.2

m, n

o sp

acin

g)

0.2

(1)

32m

Circ

le sh

ape

diam

eter

(0.2

m)

With

Raf

t & P

ile(1

R7P

)

9m

no sp

acin

g

Circ

le sh

ape

diam

eter

0.2m

-

(0.2

m, n

o sp

acin

g)

0.2

(1)

32m

Circ

le sh

ape

diam

eter

(0.2

m)

With

Raf

t & P

ile(1

R9P

)

6m

spac

ing

0.2m

Circ

le sh

ape

diam

eter

0.2m

(0.1

7m,

spac

ing0

.17m

)

(0.2

m, n

o sp

acin

g)

0.37 (2)

32m

Circ

le sh

ape

diam

eter

(0.1

7m,

0.2m

)

With

Raf

t & P

ile(1

.5R

, 6-0

.5P)

4m

no sp

acin

g

Circ

le sh

ape

diam

eter

0.2

m

(0.1

7m, n

o sp

acin

g)

(0.2

m, n

o sp

acin

g)

0.37 (2)

32m

Circ

le sh

ape

diam

eter

(0.1

7m,

0.2m

)

With

Raf

t & P

ile(2

R4P

)

6m

no sp

acin

g

Circ

le sh

ape

diam

eter

0.2

m

(0.1

7m, n

o sp

acin

g)

(0.2

m, n

o sp

acin

g)

0.37 (2)

32m

Circ

le sh

ape

diam

eter

(0.1

7m,

0.2m

)

With

Raf

t & P

ile(2

R6P

)

Wid

th o

f raf

ts

first

laye

r of R

aft

(thic

knes

s, Sp

acin

g)

Tota

l thi

ckne

ss o

f raf

t(N

o. la

yer o

f Raf

t)

Inst

alle

d de

pth

of p

iles

Spac

ing

Dim

ensi

on o

f pile

s

seco

nd la

yer o

f Raf

t (th

ickn

ess,

Spac

ing)

Dim

ensi

on o

f raf

ts

Foun

datio

n ty

pe

----

(0.2

m, 0

)

0.2

(1)

32m

Circ

le sh

ape

diam

eter

(0.2

m)

With

Raf

t(1

R)

Emba

nkm

ent h

eigh

t 5.5

m

---

(0.1

7m,

spac

ing0

.17m

)

(0.2

m, n

o sp

acin

g)

0.37 (2)

32m

Circ

le sh

ape

diam

eter

(0.1

7m,

0.2m

)

With

Raf

t(1

.5R

)

---

(0.1

7m, n

o sp

acin

g)

(0.2

m, n

o sp

acin

g)

0.37 (2)

32m

Circ

le sh

ape

diam

eter

(0.1

7m,

0.2m

)

With

Raf

t(2

R)

7m

no sp

acin

g

Circ

le sh

ape

diam

eter

0.2

m

-

(0.2

m, n

o sp

acin

g)

0.2

(1)

32m

Circ

le sh

ape

diam

eter

(0.2

m)

With

Raf

t & P

ile(1

R7P

)

9m

no sp

acin

g

Circ

le sh

ape

diam

eter

0.2m

-

(0.2

m, n

o sp

acin

g)

0.2

(1)

32m

Circ

le sh

ape

diam

eter

(0.2

m)

With

Raf

t & P

ile(1

R9P

)

6m

spac

ing

0.2m

Circ

le sh

ape

diam

eter

0.2m

(0.1

7m,

spac

ing0

.17m

)

(0.2

m, n

o sp

acin

g)

0.37 (2)

32m

Circ

le sh

ape

diam

eter

(0.1

7m,

0.2m

)

With

Raf

t & P

ile(1

.5R

, 6-0

.5P)

4m

no sp

acin

g

Circ

le sh

ape

diam

eter

0.2

m

(0.1

7m, n

o sp

acin

g)

(0.2

m, n

o sp

acin

g)

0.37 (2)

32m

Circ

le sh

ape

diam

eter

(0.1

7m,

0.2m

)

With

Raf

t & P

ile(2

R4P

)

6m

no sp

acin

g

Circ

le sh

ape

diam

eter

0.2

m

(0.1

7m, n

o sp

acin

g)

(0.2

m, n

o sp

acin

g)

0.37 (2)

32m

Circ

le sh

ape

diam

eter

(0.1

7m,

0.2m

)

With

Raf

t & P

ile(2

R6P

)

Wid

th o

f raf

ts

first

laye

r of R

aft

(thic

knes

s, Sp

acin

g)

Tota

l thi

ckne

ss o

f raf

t(N

o. la

yer o

f Raf

t)

Inst

alle

d de

pth

of p

iles

Spac

ing

Dim

ensi

on o

f pile

s

seco

nd la

yer o

f Raf

t (th

ickn

ess,

Spac

ing)

Dim

ensi

on o

f raf

ts

Foun

datio

n ty

pe


71

5.2.3 Settlement at the center of embankment

The important aspect of design of Raft & File foundation is how to estimate

residual settlement of the embankment road surface correctly after it is opened to traffic

loading.

For embankment height 5.5m, Figure 5.4 presents the settlement at the

embankment center at various support arrangements, the result shown that the case

2R6P is the most effective in reducing settlement at embankment center, resulted to a

settlement of only 1.33m after 3 years. Therefore, Raft & Pile foundation is the most

effective type of support. The exact value of the results for all cases are summarize in

Table 5.3.

5.2.4 Displacement vectors in subsoil

When an area of soil is loaded, the vertical stresses within the soil mass will be

increased. The increase is greatest directly under the loaded area, but tends to extend

indefinitely in all directions. Unbalanced forces exist which tend to cause the soil to

move from high points to low points. The most important of these forces are the force of

gravity and the force of seeping water which induce shearing stresses in soil. Unless the

resultant shearing resistance on every plan within the soil mass is greater than the

shearing forces, failure will occur in the from of movement of large mass of soil along a

more or less definite surface.

Figure 5.5(H=5.5m) show that displacement vectors for case with Raft support,

downward movements extending to deeper locations are shown with outward

movements near the toe. Raft & Pile supports resulted to more uniform downward

movements that were confined mostly beneath the embankment, depending on the

length of the pile.


72

Embankment (5.5m)

LCLC

0 500 1000 1500-3.0

-2.5

-2.0

-1.5

-1.0

-0.5

0.0

0.5

1.0Se

ttlem

ent a

t em

bank

men

t cen

ter(

m)

2R6P 1.5R,6-0.5P 2R4P 2R 1.5R 1R9P 1R7P 1R

Time(days)

0 1 2 3 4 5 Time(years)

0 10 20 30 40 50 60 Time(Months)

Figure 5.4: Settlement at embankment center (embankment height 5.5m).

Table 5.3. : Summarize the results of embankment heigh 5.5m

-1.721.5R

-2.081R

-1.971R7P

-1.821R9P

-1.692R

-1.451.5R,6-0.5P

-1.512R4P

-1.332R6P

Maximum settlement at the center of embankment(m)Cases

-1.721.5R

-2.081R

-1.971R7P

-1.821R9P

-1.692R

-1.451.5R,6-0.5P

-1.512R4P

-1.332R6P

Maximum settlement at the center of embankment(m)Cases

Emba

nkm

ent h

eigh

t 5.5

m


73

With 2R

Embankment(5.5m)

CL

(m)

(m)

0.00 10.00 20.00 30.00 40.00 50.00 60.00 70.00 80.00

0.00

10.00

-10.00

-20.00

-30.00

CLCL

(m)

(m)

0.00 10.00 20.00 30.00 40.00 50.00 60.00 70.00 80.00

0.00

10.00

-10.00

-20.00

-30.00

(a) with 2R

With 2R6P

Embankment(5.5m)

CL

(m)

(m)

0.00 10.00 20.00 30.00 40.00 50.00 60.00 70.00 80.00

0.00

10.00

-10.00

-20.00

-30.00

CLCL

(m)

(m)

0.00 10.00 20.00 30.00 40.00 50.00 60.00 70.00 80.00

0.00

10.00

-10.00

-20.00

-30.00

(b) with 2R6P

Figure 5.5: Displacement for embankment height 5.5m (a) with 2R (b) with 2R6P.

5.2.5 Effects of Raft

Number layer of Raft can be an important factor affecting subsoil deformatoions.

Figure 5.6 show the effect of number layer of Raft, the result clearly investigated the

posible dimension of necessary Raft layer that can be support embankment height 5.5m.

For case 2R is the highest rigidity can be made uniform displacement under

embankment, maximum vertical displacement of about 1.67m and maximum horizontal

displacement of about 0.24m. For case 1R is lowest rigidity, slope belong embankment

that will not be permit the construction of a 5.5m embankment on soft ground thickness


74

11m. For cases 1.5R and 2R, using Raft alone is reasonably enough to minimize reduce

differential settlements beneath embankment.

Due to excess pore pressure that corresponding to displacement vector in subsoil,

Figure 5.7. show the effect of number layer of Raft inreducing excess pore pressure, the

result of excess pore pressure express by contour. For case with 2R uniform beneath

embankment, extreme excess pore pressure 59.37kPa slowly to zero at a distance of 20

m from the centerline.

5.2.6 Effects of Pile

Figure 5.8. show the effect of Raft support reduces surface vertical displacement

beneath embankment more effectively, but with optimum pile length, the differential

settlements at transition zones and heave beyond toe of embankment have also been

reduced.

Figure 5.9. show the effect of longer pile(2R6P), extreme excess pore pressure

reduced to 42.97kPa. The effectiveness of longer Pile is evident.

Based on mention above, 2R6P will be permit the construction of a higher

embankment.

5.2.7 Effects of Raft & Pile

Based on laboratory model test results(chapter3), with the addition of Raft & Pile

Foundation, vertical settlements have reduced significantly. Lateral movement of toe

embankment also gets reduced to almost negligible levels. The lateral movements at the

embankment toe are also reduced to almost negligible level. Pile length has significant

influence on the vertical and lateral deformations of the embankment.

FE analysis, combined study with Fig. 5.10 and Fig. 5.11, the result clearly

investigated that for composities Raft & Pile foundation case 1.5R,6-0.5P. Composities

Raft & Pile with necessary of Raft layer (1.5) and Pile length 6m with spacing 0.2m is

the more effective in reducing deformation in ground compared to case 1R7P.


75

CLCL

With 1R

Embankment, 5.5m

0.0 0.5 1.0-12

-10

-8

-6

-4

-2

0

Horizontal displacement at toe of embankment(m)

3years 1years 6months 3months 27d

Dep

th(m

)

0 5 10 15 20 25 30 35

-3

-2

-1

0

1

27d 3months 6months 1years 3years

Surf

ace

verti

cal s

ettle

men

t(m)

Distance from center line(m)

Embankment(5.5m)With 1R

LCLC

(a) displacement for case 1R

CLCL

With 1.5R

Embankment, 5.5m

-12

-10

-8

-6

-4

-2

0

0.0 0.5 1.0

Dep

th(m

)



0 5 10 15 20 25 30 35

-3

-2

-1

0

1


Surf

ace

verti

cal d

ispl

acem

ent(m

)


With 1.5R

LCLC

Embankment(5.5m)

(b) displacement for case 1.5R

CLCL

With 2R

Embankment, 5.5m

-12

-10

-8

-6

-4

-2

0

0.0 0.5 1.0

Dep

th(m

)



0 5 10 15 20 25 30 35

-3

-2

-1

0

1


Surf

ace

verti

cal d

ispl

acem

ent(m

)


With 2R

LCLC

Embankment(5.5m)

(c) displacement for case 2R

Figure 5.6: Effects of Raft (a) displacement for case 1R (b) displacement for case 1.5R

(c) displacement for case 2R.


76

Embankment(5.5m)

CLCL

(m)

(m)

0.00 10.00 20.00 30.00 40.00 50.00 60.00

0.00

10.00

-10.00

-20.00

-30.00

With 1R

Extreme excess pore pressure – 69.11 kN/m2

(Pressure=Negative)

0

-10

Exce

ss p

ore

pres

sure

(kPa

)

-20

-30

-40

-50

-60

-70

-80

(a) excess pore pressure for case 1R

With 1.5R

Embankment(5.5m)

CL

(m)

(m)

0.00 10.00 20.00 30.00 40.00 50.00 60.00 70.00 80.00

0.00

10.00

-10.00

-20.00

-30.00

CLCL

(m)

(m)

0.00 10.00 20.00 30.00 40.00 50.00 60.00 70.00 80.00

0.00

10.00

-10.00

-20.00

-30.00


(Pressure=Negative)

(b) excess pore pressure for case 1.5R

With 2R

Embankment(5.5m)

CL

(m)

(m)

0.00 10.00 20.00 30.00 40.00 50.00 60.00 70.00 80.00

0.00

10.00

-10.00

-20.00

-30.00

CLCL

(m)

(m)

0.00 10.00 20.00 30.00 40.00 50.00 60.00 70.00 80.00

0.00

10.00

-10.00

-20.00

-30.00


(Pressure=Negative)

(c) excess pore pressure for case 2R

Figure 5.7: Effects of Raft on excess pore pressure (a) excess pore pressure for case

1R (b) excess pore pressure for case 1.5R (e) excess pore pressure for

case 2R


77

CLCL

With 2R

Embankment, 5.5m

-12

-10

-8

-6

-4

-2

0

0.0 0.5 1.0

Dep

th(m

)



0 5 10 15 20 25 30 35

-3

-2

-1

0

1


Surf

ace

verti

cal d

ispl

acem

ent(m

)


With 2R

LCLC

Embankment(5.5m)

(a) displacement for case 2R

CLCL

With 2R4P

Embankment, 5.5m

-12

-10

-8

-6

-4

-2

0

0.0 0.5 1.0

Dep

th(m

)



With 2R4P

LCLC

Embankment(5.5m)

0 5 10 15 20 25 30 35-4

-3

-2

-1

0

1


Surf

ace

verti

cal d

ispl

acem

ent(m

)


(b) displacement for case2R4P

CLCL

With 2R6P

Embankment, 5.5m

-12

-10

-8

-6

-4

-2

0

0.0 0.5 1.0

Dep

th(m

)



0 5 10 15 20 25 30 35

-3

-2

-1

0

1


Surf

ace

verti

cal d

ispl

acem

ent(m

)


With 2R6P

LCLC

Embankment(5.5m)

(c) displacement for case 2R6P

Figure 5.8: Effects of Pile (a)displacement for case 2R (b)displacement for case2R4P

(c)displacement for case2R6P.


78

Embankment(5.5m)

CLCL

(m)

(m)

0.00 10.00 20.00 30.00 40.00 50.00 60.00

0.00

10.00

-10.00

-20.00

-30.00

With 2R


(Pressure=Negative)

0

-10

Exce

ss p

ore

pres

sure

(kPa

)

-20

-30

-40

-50

-60

-70

-80

(a) excess pore pressure for case 2R

With 2R4P

Embankment(5.5m)

CL

(m)

(m)

0.00 10.00 20.00 30.00 40.00 50.00 60.00 70.00 80.00

0.00

10.00

-10.00

-20.00

-30.00

CLCL

(m)

(m)

0.00 10.00 20.00 30.00 40.00 50.00 60.00 70.00 80.00

0.00

10.00

-10.00

-20.00

-30.00


(Pressure=Negative)

(b) excess pore pressure for 2R4P

With 2R6P

Embankment(5.5m)

CL

(m)

(m)

0.00 10.00 20.00 30.00 40.00 50.00 60.00 70.00 80.00

0.00

10.00

-10.00

-20.00

-30.00

CLCL

(m)

(m)

0.00 10.00 20.00 30.00 40.00 50.00 60.00 70.00 80.00

0.00

10.00

-10.00

-20.00

-30.00


(Pressure=Negative)

(c) excess pore pressure for 2R6P

Figure 5.9: Effects of Pile on excess pore pressure (a) excess pore pressure for case

2R (b) excess pore pressure for case 2R4P (c) excess pore pressure for

case 2R6P.


79

CLCL

With 1R7P

Embankment, 5.5m

-12

-10

-8

-6

-4

-2

0

0.0 0.5 1.0


Dep

th(m

)


0 5 10 15 20 25 30 35

-3

-2

-1

0

1


Surf

ace

verti

cal s

ettle

men

t(m)


With 1R7P

LCLC

Embankment(5.5m)

(a) displacement for case 1R7P

CLCL

With 1.5R, 6-0.5P

Embankment, 5.5m

-12

-10

-8

-6

-4

-2

0

0.0 0.5 1.0

Dep

th(m

)



With 1.5R, 6-0.5P

LCLC

Embankment(5.5m)

0 5 10 15 20 25 30 35

-3

-2

-1

0

1


Surf

ace

verti

cal d

ispl

acem

ent(m

)


(b) displacement for case 1.5R,6-0.5P

CLCL

With 2R6P

Embankment, 5.5m

-12

-10

-8

-6

-4

-2

0

0.0 0.5 1.0

Dep

th(m

)



0 5 10 15 20 25 30 35

-3

-2

-1

0

1


Surf

ace

verti

cal d

ispl

acem

ent(m

)


With 2R6P

LCLC

Embankment(5.5m)

(c) displacement for case 2R6P

Figure 5.10: Effects of Raft & Pile (a) displacement for case 1R7P (b) displacement

for case 1.5R, 6-0.5P (c) displacement for case 2R6P.


80

Embankment(5.5m)

CLCL

(m)

(m)

0.00 10.00 20.00 30.00 40.00 50.00 60.00

0.00

10.00

-10.00

-20.00

-30.00

With 1R7P


(Pressure=Negative)

0

-10

Exce

ss p

ore

pres

sure

(kPa

)

-20

-30

-40

-50

-60

-70

-80

(a) excess pore pressure for case 1R7P

With 1.5R,6-0.5P

Embankment(5.5m)

CL

(m)

(m)

0.00 10.00 20.00 30.00 40.00 50.00 60.00 70.00 80.00

0.00

10.00

-10.00

-20.00

-30.00

CLCL

(m)

(m)

0.00 10.00 20.00 30.00 40.00 50.00 60.00 70.00 80.00

0.00

10.00

-10.00

-20.00

-30.00


(Pressure=Negative)

(b) excess pore pressure for case 1.5R,6-0.5P

With 2R6P

Embankment(5.5m)

CL

(m)

(m)

0.00 10.00 20.00 30.00 40.00 50.00 60.00 70.00 80.00

0.00

10.00

-10.00

-20.00

-30.00

CLCL

(m)

(m)

0.00 10.00 20.00 30.00 40.00 50.00 60.00 70.00 80.00

0.00

10.00

-10.00

-20.00

-30.00


(Pressure=Negative)

(c) excess pore pressure for case 2R6P

Figure 5.11: Effects of Raft & Pile on excess pore pressure (a) excess pore pressure

for case 1R7P (b) excess pore pressure for case 1.5R, 6-0.5P (c) excess

pore pressure for case 2R6P.


81

5.2.8 Surface vertical displacement with time response

For embankment height 5.5m (see Figure 5.12), the case of 2R6P is found to be the

most effective in reducing surface vertical displacement with time resulting to

maximum surface displacement at center of 1.33m after 3 years. Beyond the toe, the

heave was negligible, subsiding slowly to zero at a distance of 25 m from the centerline.

The effectiveness of Raft & Pile foundation is again evident.

0 5 10 15 20 25 30 35

-3

-2

-1

0

1

Surfa

ce v

ertic

al d

ispl

acem

ent(m

)


2R6P 1.5R,6-0.5P 2R4P 2R 1.5R 1R9P 1R 1R7P

LCLC

Embankment(5.5m)

After 27days

0 5 10 15 20 25 30 35

-3

-2

-1

0

1

Surfa

ce v

ertic

al d

ispla

cem

ent(m

)


2R6P 1.5R,6-0.5P 2R4P 2R 1.5R 1R9P 1R 1R7P

LCLC

Embankment(5.5m)

After 3 months

(a) after 27 days (b) after 3 months

0 5 10 15 20 25 30 35

-3

-2

-1

0

1

Surfa

ce v

ertic

al d

ispla

cem

ent(m

)


2R6P 1.5R,6-0.5P 2R4P 2R 1.5R 1R9P 1R 1R7P

After 6months

LCLC

Embankment(5.5m)

LCLC

Embankment(5.5m)

After 1 yr

0 5 10 15 20 25 30 35

-3

-2

-1

0

1

Surfa

ce v

ertic

al d

ispla

cem

ent(m

)


2R6P 1.5R,6-0.5P 2R4P 2R 1.5R 1R9P 1R 1R7P

(c) after 6 months (d) after 1year

0 5 10 15 20 25 30 35

-3

-2

-1

0

1

Surfa

ce v

ertic

al d

ispla

cem

ent(m

)


2R6P 1.5R,6-0.5P 2R4P 2R 1.5R 1R9P 1R7P 1R

LCLC

Embankment(5.5m)

After 3yr

(e) after 3 years

Figure 5.12: Surface vertical displacement with time for embankment height 5.5m (a)

after 27 days (b) after 3 months (c) after 6 months (d) after 1year (e) after 3 years.


82

5.2.9 Lateral displacement with time response

FEM analysis has shown that for embankment height 5.5m (see Figure 5.13). The

case of 2R6P is found to be the most effective in reducing lateral displacement with

time resulting, result shown that the lateral displacement, increase with depth and

maximum at 0.17m of lateral displacement up to a depth of 8.5m and steadily decreases

to zero at a depth of 11m. The effectiveness of Raft & Pile foundation is again evident.

CLCL

Embankment, 5.5mAfter 27 days

-12

-10

-8

-6

-4

-2

0

0.0 0.5 1.0

Dep

th(m

)


2R6P 1.5R,6-0.5P 2R4P 2R 1.5R 1R9P 1R 1R7P

CLCL

Embankment, 5.5mAfter 3months

-12

-10

-8

-6

-4

-2

0

0.0 0.5 1.0D

epth

(m)


2R6P 1.5R,6-0.5P 2R4P 2R 1.5R 1R9P 1R 1R7P

(a) after 27 days (b) after 3 months

CLCL

Embankment, 5.5mAfter 6 months

-12

-10

-8

-6

-4

-2

0

0.0 0.5 1.0

Dep

th(m

)


2R6P 1.5R,6-0.5P 2R4P 2R 1.5R 1R9P 1R 1R7P

CLCL

Embankment, 5.5mAfter 1 yr

-12

-10

-8

-6

-4

-2

0

0.0 0.5 1.0

Dep

th(m

)


2R6P 1.5R,6-0.5P 2R4P 2R 1.5R 1R9P 1R 1R7P

(c)

after 6 months (d) after 1year

CLCL

Embankment, 5.5mAfter 3 yr

-12

-10

-8

-6

-4

-2

0

0.00 0.25 0.50 0.75 1.00

Dep

th(m

)


2R6P 1.5R,6-0.5P 2R4P 2R 1.5R 1R9P 1R 1R7P

(e) after 3 years

Figure 5.13: Lateral displacement at toe with time for embankment height 5.5m


83

5.3 Evaluation of the effectiveness of Raft & Pile Foundation with subsidence

correction

Figure 5.14 shows the results to evaluate the effectiveness of Raft & Pile

Foundation with subsidence correction for embankment height of 5.5m. It can be seen

that with Raft & Pile test case 2R6P the subsidence beneath embankment is only about

19.9 m3/m and the subsidence beyond the toe of embankment 5.5m is about 3.38 m3/m.

2R6P

2R4P

2R

1.5R,6-0.5P

1.5R

0 5 10 15 20 25

19.9

3.38

22.5

3.35

25.5

3.53

21.9

3.78

3.38

25.9

Subsidence(m3/m)

Case

s

Subsidence beyond toe of embankment Subsidence beneath of embankment

LCLC

Embankment(5.5m)

After 1 yr

0 5 10 15 20 25 30 35

-3

-2

-1

0

1Su

rface

ver

tical

dis

plac

emen

t(m)


Figure 5.14: Effectiveness of Raft & Pile foundation.


84

5.4. Performance of road embankment design height 5.5 m above the ground level

with Raft & Pile foundation.

Scott (1981) pointed out that design is the most complicated process that an

engineer is called upon to perform. Design is an intricate amalgam of experience,

judgment, measurement, and analysis, usually pursued on a trial-and-error basis. One of

the most important design aspects for Raft & Pile Foundation is how to estimate

residual settlements of road embankments road under operational conditions. Based on

the results in Chapters 3, 4 and 5, the geometries of Raft and Pile have been investigated

further in terms of determining the optimum design.

The model dimensions are 30 m deep from ground surface and 80 m wide from the

centerline of the embankment. The displacement boundary conditions were as follows:

at bottom, both vertical and horizontal displacements were fixed, and for left and right

vertical boundaries, the horizontal displacement was fixed. The adopted drainage

boundary conditions were as follow: the ground surface and bottom line (sand layer)

were drained. The left and right boundaries were drained. Figure 5.15 shows the finite

element mesh for the cross section of embankment while the loading history is also

shown in Fig. 5.21. The mechanical behavior of the clay was represented by soft soil

model(combined Mohr-Coulomb and Modified Cam Clay) and the sand layers were

assume to be elastic. The determined model parameters for subsoil (Chai et al 1999) and

material properties are listed in Table 5.1. Timber for Raft & Pile foundations were

assumes to be elastic (K.J.Kim, WWW.hitasca.co.th). The ground –water level was the

same with ground surface. The mechanical property of the fill material was represented

by Mohr-Coulomb model (PlaxisV.8). The conditions of the model tests are summarized

in Table 5.4.


85

Embankment

Without 2R6P

(7.3m)

CL

(m)

(m)

0.00 10.00 20.00 30.00 40.00 50.00 60.00 70.00 80.00

0.00

10.00

-10.00

-20.00

-30.00

CLCL

(m)

(m)

0.00 10.00 20.00 30.00 40.00 50.00 60.00 70.00 80.00

0.00

10.00

-10.00

-20.00

-30.00


0

7.3m

Loadfill amount

(m) 25days

Weathered crust (B), 0m-1m

Ariake clay (Ac2), 1m-11m

Sand (As2), 11m-16m

Dense Sand (Ds), 16m-30m

Figure 5.15: Typical Finite element mesh.

Table 5.4: Detail for numerical simulation of embankment.

6No spacing0.2320.3722R8P(with Raft & Pile)

Installed depth of pile (m)

Spacing of pile (m)

Diameter of raft (m)

Width of raft (m)

Thickness of raft (m)

No. layer of rafts

Test case

6No spacing0.2320.3722R8P(with Raft & Pile)

Installed depth of pile (m)

Spacing of pile (m)

Diameter of raft (m)

Width of raft (m)

Thickness of raft (m)

No. layer of rafts

Test case

Note: Table 4.2 for 2 layers of Raft purpose relation of EI, EIdesign= EIFEM.

Figure 5.16 presents the settlement at the embankment center with the applied

pressure. As shown in the figure, the settlement was reduced to 1.8m with the use of

Raft & Pile.


86

Figure 5.16: Settlement at embankment center.

Figure 5.17 (a) shows that displacements for the case with Raft & Pile are more

uniform that tend to move downwards and confined beneath the embankment with the

pile length of 6m. Figure 5.17 (b) shows that the surface heave beyond the toe and the

settlement at center reduced significantly. This is a very significant advantage as there is

negligible heave beyond the toe. Figure 5.17 (c) shows that the lateral displacements

increase with depth and the maximum value of 0.2m located at a depth of 8.5m and

subsequently decrease to zero at a depth of 11m. Based on this pattern of deformation,

Raft & Pile foundation functions satisfactorily in terms of reducing lateral

displacements.

The excess pore pressures are shown in Fig. 5.18. The excess pore pressure with

Raft & Pile amounted only to -77.44kPa.


0

7.3m

Loadfill amount

(m)25days

0 500 1000 1500-3.0

-2.5

-2.0

-1.5

-1.0

-0.5

0.0

0.5

1.0

1.5

2R6P

Settl

emen

t at e

mba

nkm

ent c

ente

r(m

)

Time(days)

0 1 2 3 4 5 Time(years)

0 10 20 30 40 50 60 Time(Months)


87

(a)

Embankment

With 2R6P

(7.3m)

CL

(m)

(m)

0.00 10.00 20.00 30.00 40.00

0.00

10.00

-10.00

-20.00

-30.00

CLCL

(m)

(m)

0.00 10.00 20.00 30.00 40.00

0.00

10.00

-10.00

-20.00

-30.00

0 5 10 15 20 25 30 35

-3

-2

-1

0

1


Surf

ace

verti

cal d

ispl

acem

ent(m

)


Embankment(7.3m)With 2R6P

LCLC

(b)

(c)

CLCL

Embankment 7.3m

2R6P

-12

-10

-8

-6

-4

-2

0

0.0 0.5 1.0

Dep

th(m

)



Figure 5.17: Displacement for case with Raft & Pile (a) Total displacement after end of

construction (25days) (b) Surface vertical displacement (c) Lateral

displacement at toe of embankment.

Embankment(7.3m)

CLCL

(m)

(m)

0.00 10.00 20.00 30.00 40.00 50.00 60.00

0.00

10.00

-10.00

-20.00

-30.00

With 2R6P

0

-10

Exce

ss p

ore

pres

sure

(kPa

)

-20

-30

-40

-50

-60

-70

-80


(Pressure=Negative)

Figure 5.18: : Excess pore pressures for embankment design height 5.5m


88

For the case of road embankments on Ariake clay foundation, a road

embankment of 12m wide and 5.5m of height and slope of 1V:1.8H can reasonably

work when the soft foundation is 11m thick. The probable dimensions of Raft & Pile

foundation can be Raft 2 layer, 32m wide and thickness 0.37m, pile diameter 0.2m and

length 6m, subsidence correction thickness 1.8m as show in Figure 5.19.

Subsidence correction thickness 1.8m1.43m

Thickness of Raft(2R: 0.37m)

Soft Ariake clay groundthickness: 11m

Construction fill thickness 7.3m

Design embankmentheight 5.5m

12m

1:1.8

32m

Sand

6m

CL

Figure 5.19: Cross section of embankment.

5.5 Summary

Based on results of the finite element analysis, the following conclusions can be

drawn.1) Raft & Pile foundation shows promise in reducing both settlements at

embankment center and heave beyond the toe of embankment. They are also effective in

reducing the generation of excess pore pressures and its rate of dissipation.

2) Embankment load was carried by Raft, the contribution of the piles was found to

decrease at large displacements.

3) Using Raft alone is reasonably enough to minimize reduce differential settlements

beneath embankment, but with piles the differential settlements at transition zone and

heave beyond toe of embankment was further reduced.

4) The numerical analyses provided optimum design in terms of geometries and layouts

of Raft & Pile foundation.

5) A 12m wide and 5.5m of high and side slopes of 1V:1.8H can be designed on soft

Ariake clay foundation 11 m thick. The probable dimensions of Raft & Pile foundation


89

includes the following: Raft 2 layers, 32m wide and thickness 0.37m, pile diameter

0.2m and length 6m, subsidence correction thickness 1.8m.

References

1) Bergado, D.T., Teerawattanasuk, C., 2007. 2D and 3D numerical simulations of

reinforce embankment on soft ground. Geotextiles and Geomembrane, : 1 - 17.

2) Chai, J.C., Miura, N.(1999), Investigation of factors affecting vertical drain behavior

soil. Journal of Geotechnical and Geoenvironmental Engineering, pp. 216-226.

3) Chai, J.C., Miura, N., and Shen, S.L. (2002). Performance of embankments with and

without reinforcement on soft subsoil. Canadian Geotechnical Journal, 234-435.

4) Hayashi, S., Poungchompu, P, Du, Y.J, D. Suetsugu and S. L. Shen(2006).

Experimental study on performance of timber Raft & Pile foundation of

embankment on soft ground. Proceedings of the International Symposium on

Lowland Technology, September 14-16, 2006, pp.207-212.

5) Hayashi, S., Du, Y.J. (2005). Geotechnical analysis of Mizuki embankment remains.

Soils and Foundations Vol. 45, No. 6, 43-53, Dec. 2005, pp.43-53.

6) Kim, K.J. of the north Carolina department of Transportation on the design of

foamed concrete embankment supported by timber piles for the U.S. 64 widening

project in Tyrrell Country, North Carolina, WWW.hcitasca.co .th.

7) Ochiai, H., Hayashi, S., Umezaki, T., and Otani, J. (1991). Model test on sheet-pile

countermeasures for clay foundation under embankment. Developments in

Geotechnical Aspects of Embankment, Excavations and Buried Structures,

Netherlands, pp.277-291.

8) Plaxis V.8 Manual (2008).

9) Poulos, H.G. (1991). “Analysis of piled strip foundations” Computer Methods and

Advances in Geomechanics, Beer, Booker & Carter (eds), Balkema, Rotterdam, pp.

183-191.

10) Ronald F. Scott. (1981). “Foundation Analysis” Civil Engineering and Engineering

Mechanics Series, USA, pp. 1-493.

11)Robert W. Brown. (1996), “Practical Foundation Engineering Handbook”

McGraw-Hill, USA.

12) Shen, S.L., and Miura, N. (2001). A technique for reducing settlement difference of


90

road on soft clay. In Computer Methods and Advances in Geomechanics, Proc. 10

IACMAG, Vol.2, Edited by C.S. Desai et al., A.A. Balkema, pp.1391-1394.

13) Wood, D. M.(1990), “Soil behaviour and critical state soil mechanics” The press

syndicates of the University of Cambridge, New York, USA, pp.1-459.

14) Wood, D. M.(2004). “Geotechnical modelling” Applied Geotechnics Volume 1,


91

CHAPTER VI

CONCLUSIONS AND RECOMMENDATIONS

6.1. Conclusions

Based on laboratory model test results, the following conclusions can be drawn:

1) An embankment loading without reinforcement can only support very low

pressures of 10 to 15kPa. With the addition of a raft or raft plus piles, failure was

not observed even under applied pressures of 30 kPa.

2) The addition of a flexible raft resulted to a significant reduction of both vertical

settlements and lateral deformations. No failure was observed with increase in

applied pressure up to 2 to 3 times the failure pressure in the case without

support.

3) The stability of the raft foundation could be further improved by the addition of

piles at both sides of the raft. This additional support removes almost completely

the heave beyond the embankment toe. The vertical deformations have also

reduced significantly. The raft plus piles transfers better the embankment

loading deeper into the subsoil. The lateral movements at the embankment toe

are also reduced to almost negligible level.

4) Pile length has significant influence on the vertical and lateral deformations of

CHAPTER VI CONCLUSIONS AND RECOMMENDATIONS

92

the embankment. The raft with its edges supported with piles is found to be very

effective in reducing differential deformation beneath and beyond the toe of

embankment. The experimental results demonstrate that even in very soft soils,

locally available and cheap timber can be effectively utilized to carry the

embankment load safety. The results of the model test need to be validated by

full-scale tests in the field.

5) The flexural characteristics of raft composites can be treated as flexible

reinforcements in the design.

6) The number of raft layers and axial raft elements were significant on the rigidity

of raft composites. Links between short timber Rafts (simulated field condition)

and long timber Raft (simulated FEM analysis) have been established. They

provide relationships between the designed rigidity (EI) and FEM analysis in

terms of the number of layers of Raft.

The results of finite element analysis have the following conclusions:

1) FEM of analysis has shown that the advantages of using timber Raft & Pile are

reducing both settlements at embankment center and heave beyond toe of

embankment.

2) FEM analysis has shown that timber Raft & Pile also reduce and the generation of

excess pore pressures in the subsoil and its rate of dissipation.

3) Embankment load was carried by raft, the contribution of the piles was found to

decreases at large displacements.

4) Raft support reduces differential settlements beneath embankment more

effectively, but with pile, the differential settlements at transition zones and heave

beyond toe of embankment have also been reduced.

5) A design example has been produced to illustrate different design parameters such

as number of layers, optimum pile length, and thickness of corrected filled

material and differential settlements inside and outside the toe of embankment.

An embankment of12m wide and 5.5m of high, and side slopes of 1V:1.8H works

well for soft Ariake clay 11m thick. Raft & Pile geometries are as follows: 2

layers of Raft with 32m wide and thickness 0.37m, pile diameter of 0.2m and 8m

long, and subsidence correction thickness of 2m.

CHAPTER VI CONCLUSIONS AND RECOMMENDATIONS

93

6) Use timber as construction material in highway construction can absorb CO２

resulting to an environmental-friendly material. Besides, timber is durable when

they are placed below the groundwater table.

6.2. Recommendations for future research

1) Development of bending test and bending test supported by soil for large scale

composites Raft for construction.

2) Investigate the interaction between embankments and structures, which can be

difficult to control in design. Most embankments on compressible soils are crossed by

box culverts, which can pose another complexity.

3) Study on the interaction between old and new embankments. Differential

settlements can be an issue and solutions should be pursued.

4) Investigation in to performance of Raft & Pile method with PVD on thick clay

ground.

5) FEM 3D modeling of timber Raft & Pile Foundation of embankment on soft

ground

94

PUBLICATIONS

Refereed Papers

1) S. Hayashi, P. Poungchompu, D. Suetsugu., and Y. J. Du (2008), “Investigation into

performance of Raft & Pile supported embankment on soft ground”, Geotechnical

Engineering Journal. (Accepted)

2) P. Poungchompu, S. Hayashi, D. Suetsugu, Y. J. Du. And M. C. Alfaro（2008),

“Analysis of Raft & Pile foundation on soft Ariake clay ground subjected to

embankment loading”, Lowland Technology International Journal. (under review)

3) S. Hayashi, P. Poungchompu, Y. J. Du, and D. Suetsugu. (2006), “Experimental

study on performance of timber Raft & Pile foundation of embankment on soft ground”,

International Symposium on Lowland Technology, Saga University, Japan, pp. 207-212.

4) P. Poungchompu, S. Hayashi, Y. J. Du, and D. Suetsugu. (2007), “Laboratory study

on timber Raft & Pile foundation on soft ground”, International 60th Canadian

Geotechnical Conference & 8th Joint CGS/IAH-CNC Ground water Conference, Ottawa,

Ontario, Canada, pp. 1809-1815.

5) P. Poungchompu, S. Hayashi, D. Suetsugu, Y. J. Du. And M. C. Alfaro（2008),

“Finite element of Raft & Pile foundation on soft Ariake clay ground subjected to

embankment loading”, International Symposium on Lowland Technology, Busan,

Korea.

PUBLICATIONS

95

Domestic conferences

1) P. Poungchompu, S. Hayashi, Y. J. Du, D. Suetsugu and Y. Miyoshi (2006),

“Development of Laboratory test apparatus for studying performance of Raft & Pile

foundation in soft ground subjected to embankment loading”, Proceedings of 41thAnual

Meeting of the Japan Geotechnical Society, Kagoshima, Japan.

2) P. Poungchompu, S. Hayashi, Y. J. Du, D. Suetsugu and Y. Miyoshi (2006),

“Preliminary study on behavior of model Raft & Piled foundations in soft ground

subjected to road embankment loading”, Proceedings of the Anual Conference of the

Japan Society of Civil Engineers, Kyushu Branch, Miyasaki, Japan.

3) P. Poungchompu, S. Hayashi, D. Suetsugu, Y. J. Du, Y. Miyoshi and T. Umeda.

(2008), “Finite element modeling of Raft & Pile foundation on soft Ariake clay ground

subjected to embankment loading”, Proceedings of the Anual Conference of the Japan

Society of Civil Engineers, Kyushu Branch, Nagasagi, Japan.

PUBLICATIONS

96

DEVELOPMENT OF A TIMBER RAFT & PILE...

Documents

Transcript of DEVELOPMENT OF A TIMBER RAFT & PILE...