SWELLING PERFORMANCE OF SOME EXPANSIVE SOIL TREATMENT TECHNIQUES BY

Ain Shams University Faculty of Engineering

Department of Structural Engineering

SWELLING PERFORMANCE OF SOME

EXPANSIVE SOIL TREATMENT

TECHNIQUES BY

Mohamed Youssef Abd El-Latif B.Sc, Civil Engineering "Strucural" (2005)

Ain Shams University - Faculty of Engineering

THESIS

Submitted in partial fulfillment of the requirements for Degree of Master of Science in Civil Engineering

Structural Engineering Department (Geotechnical Engineering)

Supervised By

Prof. Dr. Ali Abd El-Fattah Ali Ahmed Professor of Geotechnical Engineering

Structural Engineering Department Ain Shams University - Faculty of Engineering

Dr. Hoda Abd El-Hady Ibrahim Lecturer of Geotechnical Engineering

Structural Engineering Department Ain Shams University - Faculty of Engineering

2008

i

Ain Shams University Faculty of Engineering

Structural Engineering Department

Abstract of the M.Sc. Thesis submitted by:

Mohamed Youssef Abdel-Latif Title of Thesis:

Swelling performance of some expansive soil treatment techniques

Supervisors: Prof. Dr. Ali Abdel-Fattah Ali Ahmed

Dr. Eng. Hoda Abdel-Hady Ibrahim

Registration date : 10/ 10/ 2005 Examination date:

Abstract

Swelling behavior of shallow foundations resting on treated expansive soil is generally affected by

different factors. Some of these factors are related to the boundary conditions controlling the site

deposition and the mode of water migration, while the others are related to the employed treatment

technique for damping the heave potential of the expansive soil. Although the utilization of some

treatment techniques were considered as an adequate techniques for reducing both swelling and

swelling pressure of expansive soils, during the last decades, many problems occurred for shallow

foundation buildings constructed on treated expansive soil.

The main objective of the present thesis is to study the effectiveness of some widely used treatment

techniques to eliminate or damp the swelling behavior of expansive soil and also to study the factors

that causes differential movement between footings. A case study of an inclined super structure at the

middle Mokattam plateau was studied. Although the expansive soil layer were treated with sand

replacement but it failed in preventing the harmful effects due to the differential heaving of soil, and

this can be attributed to the presence of water seeping from a nearby drainage utility below the

replacement layer. The stratigraphy of the middle Mokattam plateau was presented and the

geotechnical properties of the soil at the site area were investigated. The movement of the building

was simulated depending on six survey readings during a time period reaches about four years nearly

after the building inclination. Also the horizontal displacement components for the building were

obtaind (using two semi empirical methods) and compared with the measured values to decide the

best and more suitable method which can be used in predicting heave of footings.

ii

A laboratory testing program was designed to determine the swelling behavior and differential heave

effect on shallow footings resting on treated expansive soil employing different treatment techniques

using a large laboratory model. The geotechnical properties were investigated and classified using

direct and indirect methods. The expansive soil was treated using sand replacement, sand

replacement with 5% lime, expansive soil with 5% lime stabilization and sand replacement with

horizontal plastic barriers. Footings heave, moisture distribution, and differential heave between

footings were measured and predicted using empirical and semi empirical equations. Loading

conditions, level of horizontal barrier and water leakage spacing from footings were taken in

consideration.

All the used treatment techniques succeeded in decreasing the final heave with nearly the same

value. The use of plastic horizontal barrier at the top of replacement layer succeeded in distributing

the moisture uniformly over the expansive soil layer in case of water seeping from the sheet, and this

result in uniform heave of footings with time. While in case of water seeping from the plastic sheet

used at bottom of sand cushion, this concentrated the leakage at definite point which results in high

differential movement between footings. This research project has helped to identify the expansive

soils in this area and some of its associated problems and to increase the effectiveness of the

proposed treatment techniques, which can help in mitigating the structural damages originating from

the behavior of expansive soils.

Keywords: Expansive soil; Geotechnical properties; Treatment techniques; Sand replacement;

Heave prediction; Middle Mokattam plateau.

iii

APPROVAL SHEET

Name of Thesis: Mohamed Youssef Abd El-Latif

Title of Thesis: "Swelling performance of some expansive soil

treatment techniques"

Degree : Master of Science in Civil Engineering

(Structural Engineering)

EXAMINERS COMMITTEE

Name, Title & Affiliation: Signature

Prof. Dr. Khadiga I. Abdel-Ghany

Prof. of Geotechnical Engineering

Housing & Building National Research center Cairo, Egypt -----------------

Prof. Dr. Abdel-Monem Moussa


Faculty of Engineering, Ain Shams University -----------------

Prof. Dr. Ali Abd El-Fattah Ali


Faculty of Engineering, Ain Shams University -----------------

Date: / 3 / 2008

iv

STATEMENT

This dissertation is submitted to Ain Shams University for the

degree of M. Sc. in Civil Engineering.

The work included in this thesis was carried out by the author in

the Department of structural Engineering, Ain Shams University

from 2006 to 2008.

No part of this thesis has been submitted for a degree or for

qualification at any other university or situation.

Name : Mohamed Youssef Abd El Latif

Signature :

Date : / / 2008

v

ACKNOWLEDGMENT

The present work was conducted out at the Department of Structural Engineering,

Faculty of Engineering, Ain Shams University.

It was completed under the supervision of Prof. Dr. Ali Abdel Fattah, Prof. of

Geotechnical Engineering and Dr. Eng. Hoda Abd El-Hady, Lecturer of Geotechnical

Engineering, whom I have the pleasure of working under their supervision. I express

sincere appreciation for their helpful, generous advice and guidance throughout the

period of this research.

I thank my supervisors who have been very instrumental in enriching my thesis.

Appreciably I thank them so much for accepting me to be their student and for

providing me with the guiding hand of Great Spirit in carrying out this research.

I would like to thank the soil mechanics laboratory staff for their valuable helps during

testing period. A debt of gratitude is to all people who in one way or another

contributed ideas directly or indirectly. Because it would end up into long list to

mention all the people I am indebted to, I gratefully thank all of them collectively.

Last, I would like to express my deep feelings towards each member of my family to

whom I owe every success in my life. My cordial thanks spread out to my mother for

her love, support and guidance throughout my life and for inculcating in me the

passion for knowledge.

Mohamed Youssef

Mars, 2008


vi

TABLE OF CONTENTS

Chapter 1. INTRODUCTION Page

1.1 General 1

1.2 Objectives of This Research 3

1.3 Organization of the Present Work 4

Chapter 2. LITERATURE REVIEW

2.1 General 6

2.2 Swelling Mechanism 6

2.3 Classifications and Identifications 8

2.3.1 Indirect techniques 10

2.3.1.1 Atterberg Limits Tests: 12

2.3.1.2 Colloid Content Test: 15

2.3.1.3 Activity Method 18

2.3.1.4 Free Swell Test 22

2.3.1.5 Soil Expansion Potential (ASTM D-4829) 22

2.3.1.6 Cation Exchange Capacity: 23

2.3.1.7 Cation Exchange Activity: 24

2.3.1.8 Coefficient of Linear Extensibility (COLE) 27

2.3.1.9 Soil water characteristic curve 29


vii

2.3.2 Direct techniques 34

2.3.2.1 Constant volume test 34

2.3.2.2 Double oedometer test 34

2.3.2.3 Simplified oedometer test 35

2.3.2.4 Pre-swell sample method 39

2.3.3 Combination techniques 41

2.4 Heave prediction 41

2.4.1 Semi-Empirical Methods 41

2.4.1.1 Texas Method No. 1 41

2.4.1.2 Double oedometer test 42

2.4.1.3 Simplified oedometer test 43

2.4.1.4 Constant volume test 44

2.4.1.5 Stress change method "Closed Form Heave Equation" 49

2.4.1.6 Suction change method 52

2.4.2 Empirical Methods 59

2.5 Treatment of expansive soils 69

2.5.1 Miscellaneous Treatments 71

2.5.1.1 Replacement Fill 71

2.5.1.2 Remolding and compaction 86

2.5.1.3 Pre-Wetting 87

2.5.1.4 Sub-Drainage 87


viii

2.5.2 Hydraulic Barriers 89

2.5.2.1 Horizontal moisture barriers 89

a) Horizontal Membranes 89

b) Rigid Barriers 91

2.5.2.2 Deep Vertical Moisture Barriers (DVMB) 91

2.5.3 Chemical Soil Treatments 98

2.5.3.1 Calcium-Based Stabilizers 99

a) Lime stabilization 99

b) Coal fly ash stabilization 107

i) Stabilization with Non-Self-Cementing Coal Fly Ash 110

ii) Stabilization with Self-Cementing Fly Ash 111

c) Portland cement stabilization 114

d) Cement By-Pass Dust (CBPD) Stabilization 118

e) Slag Stabilization 118

2.5.3.2 Non-Calcium Based Chemical Stabilizers 120

2.5.4 Mechanical Treatments 123

2.5.4.1 Surcharge Stress 123

2.5.4.2 Fiber Reinforcement 123

2.5.4.3 Geogrids 124


ix

Chapter 3. SUPPER STRUCTURE RESTING ON TREATED EXPANSIVE SOIL “CASE STUDY”

3.1 General 125

3.2 Description of building and cracks 125

3.3 Geological formation 133

3.4 Geotechnical Properties 135

3.4.1 Indirect measurements 141

3.4.2 Direct Measurements 143

3.5 Heave prediction 144

3.5.1 Using semi-empirical heave equations 144

3.5.2 Using empirical heave equations 146

3.6 Measured building movements 148

3.7 Comparison between the measured and predicted horizontal displacements components 154

Chapter 4. SWELLING PREFORMANCE OF SOME EXPANSIVE

SOIL TREATMENT TECHNIQUES

4.1 General 157

4.2 Properties of expansive soil used 157

4.3 Indirect measurement tests results 158

4.4 Direct measurement tests results 162

4.5 Laboratory model test 163

4.6 Test procedure 163


x

4.7 Laboratory test program 165

4.8 Test results and discussions 169

4.8.1 Measured heave results 169

4.8.1.1 Untreated expansive soil 169

4.8.1.2 Treated expansive soil using various treatment techniques 172

4.8.1.3 Treated expansive soil using various horizontal barrier locations 179

4.8.1.4 Treated expansive soil using various leakage spacing 185

4.8.2 Predicted heave results 191 Chapter 5. SUMMARY, CONCLUSIONS AND

RECOMMENDATIONS FOR FURTHER STUDIES

5.1 Summary 196

5.2 Conclusions 197

5.2.1 Case study 197

5.2.2 Experimental results 199

5.2.2.1 Measured results 199

5.2.2.2 Predicted heave 200

5.3 Recommendations 201

5.4 Recommended Future Studies 201

REFERENCES 202

Appendix A 222

Appendix B 226


xi

LIST OF FIGURES

Figure No. Page

Figure (2.1) Schematic representation of the structure of Smectite minerals

"Montmorillonites" (Colmenares, 2002). 7

Figure (2.2) Swell potential as a function of soil plasticity index (Seed, et al., 1962b)

12

Figure (2.3) Plot of clay minerals on Casagrande’s chart (Lucian, 2006). 16

Figure (2.4): Soil classification chart (Skempton, 1953). 16

Figure (2.5) Swell potential as function of colloids content and Activity (Seed, et

al.,1960) 20

Figure (2.6a) Soil swell potential based on size fraction and activity (Seed ,1962a) 20

Figure (2.6b) Potential severity of volume change for clay soils (Van Der

Merwe,1964). 21

Figure (2.7) Clay mineralogy as a function of Activity and Cation Exchange Capacity

(Holt, 1969) 25

Figure (2.8) Mineralogical classification from Pearring (1963). 26

Figure (2.9) Expansion potential as a function of CEAc and Ac from Nelson and

Miller (1992) 26

Figure (2.10) Swell potential as a function of colloids content and COLE (McKeen

and Hamberg, 1981) 31


xii

Figure (2.11) Soil expansiveness and COLE regions as a function of Activity and

Cation Exchange Capacity (McKeen and Hamberg, 1981) 31

Figure (2.12) Example of the relationship between soil suction and water content

(McKeen, 1992) 32

Figure (2.13) Example of the relationship between volume strain and soil suction

(McKeen, 1992) 32

Figure (2.14) Relation of soil water characteristic curves, soil plasticity and percent

fines (Zapata, et al., 2000) 33

Figure (2.15) Relation of suction compression index, Ch, to the slope of the soil water

characteristic curve (McKeen, 1992). 33

Figure (2.16) Typical constant volume swell test results (After Porter and

Nelson,1980). 37

Figure (2.17) Double oedometer test results, Initially moist sample pair (Jennings and

Kerrich,1962) 37

Figure (2.18) Free swell under load in the oedometer (after Fredlund, 1983) 38

Figure (2.19) Simplified oedometer test analysis (After Jennings et al., 1973) 38

Figure (2.20) Typical plot of consolidation - swell test results (After Jennings et

al.,1973) 39

Figure (2.21) Laboratory relationship between void ratio and effective pressure(After

Richard et al., 1969). 46

Figure (2.22) Idealized three dimensional loading surface for unsaturated soils in

terms of void ratio versus indepented stress (After Fredlund, 1983). 46


xiii

Figure (2.23) Correction of constant volume swell test data for sample disturbance

(After Fredlund, 1983). 47

Figure (2.24) Idealized and actual versus analysis stress path for prediction based on

constant volume (After Fredlund, 1983). 47

Figure (2.25) Void ratio versus water content (After Hamberg, 1985), 55

Figure (2.26) Idealized moisture boundary profile for the Pierre Shale, fort collins

(Hamberg, 1985). 55

Figure (2.27) Relationships for determining (a) Plasticity index (P.I) and (b) Reduction

factor (P) for Van Der Merwe’s empirical heave prediction methods

(After Van Der Merwe, 1964). 61

Figure (2.28) Laboratory model test set up (after Awad 2005). 75

Figure (2.29) Swelling- log stress of expansive soil model and oedometer ( clay

content 100%) (after Awad 2005). 75

Figure (2.30) Swelling potential of treated soil-lime content in oedometer and model

(after Awad 2005). 76

Figure (2.31) Swelling potential of treated soil-lime content in oedometer and model

(after Awad 2005). 76

Figure (2.32) Laboratory model test set up (after Awad and Abdel-Hady, 2005). 78

Figure (2.33) Predicted ground surface heave versus thickness of replacement using

Stress Change method 84

Figure (2.34) Reduction of heave versus thickness of replacement using Stress Change

Method 85


xiv

Figure (2.35) embedded drains around structures 88

Figure (2.36) Application of a horizontal membrane (Tm–Army, 1983).

91

Figure (2.37) Deep vertical moisture barrier, DVMB (Snethen, 1979) 93

Figure (2.38) vertical and horizontal moisture barriers (Tm–Army, 1983). 93

Figure (2.39) Lateral view of laboratory barrier model (after Rojas et al. 2006). 94

Figure (2.40) Finite-element grid for vertical moisture barrier model (after Rojas et al.

2006). 95

Figure (2.41) Isovalue–suction curves in meters of water (after Rojas et al. 2006). 96

Figure (2. 42) Theoretical and experimental result comparisons for surface heave and

different barrier depths 97

Figure (2.43)The Visual Effect of Lime Addition (Wibawa, 2003). 104

Figure (3.1) A satellite photo for the housing project in the Mokattam middle plateau.

128

Figure (3.2) The utilized foundation system. 128

Figure (3.3a) Photo for defects and cracks on front 1 129

Figure (3.3b) Photo for defects and cracks on front 2 129

Figure (3.4) Locations of vertical lines of inclination-marks and settlement-marks for

the defected building. 131

Figure (3.5) Composite stratigraphic section of G. Mokattam (after El-Nahhas et al.,

1990). 134


xv

Figure (3.6) Schematic layout for the described building and boreholes, the defected

building is the shaded one. 138

Figure (3.7a) Photo for the expansive soil layer bearing on calcareous sandstone layer

139

Figure (3.7b) Photo showing the shrink – swell effect for expansive soil layer. 139

Figure (3.7e) Undisturbed expansive soil sample before preparation

140

Figure (3.7f) Samples from the expansive soil layer showing the apparent composition

of this layer 140

Figure (3.8): Classification of expansive soil samples according to Van de Merwe,

1964. 141

Figure (3.9): Classification of expansive soil according to Seed et al. (1960) and

modified by Carter and Bentley, 1991. 141

Figure (3.10): Results of expansive soil samples using simple modified oedometer test

according to Jenning et al. (1973).

143

Figure (3.11): Comparison between the predicted heave according to Hamberg, 1985

and Rama et al., 1988. 145

Figure (3.12): Vertical displacement for the building front (1) versus measuring date.

149


149


xvi


150


150

Figure (3.16): Vertical displacement for the building bench marks versus measuring

date. 151

Figure (3.17): Relative vertical displacement for the building facades versus

measuring date. 151

Figure (3.18): Horizontal displacement at roof in the longitudinal direction relative to

the first reading versus measuring date. 152

Figure (3.19): Horizontal displacement at roof in the transverse direction relative to

the first reading versus measuring date. 152

Figure (3.20): Movement of the apparent defected building away from the adjacent

one. 154

Figure (3.21): Building rotation between bore (3) and bore (2). 155

Figure (3.22): Directions of horizontal displacement components.

155

Figure (4.1): Classification of expansive soil mineral using Casagrande’s chart. 159


1964. 160


modified by Carter and Bentley, 1991. 160


xvii


according to Jenning et al. (1973). 162

Figure (4.5) Testing mould details. 164

Figure (4.6) A schematic cross-section in the used mould 167

Figure (4.7) Footings heave versus logarithmic time for variable footing stress 170

Figure (4.8) Differential heave between footings versus log time.

170

Figure (4.9a) Footing heave versus log time for different treatment techniques 173

Figure (4.9b) Footing heave versus log time for different treatment techniques 173

Figure (4.9c) Footing heave versus log time for different treatment techniques 174

Figure (4.9d) Footing heave versus log time for different treatment techniques 174

Figure (4.10a) Differential heave between loaded footings versus log time for

different treatment techniques, q=0.5 Kg/cm2 175

Figure (4.10b) Differential heave between unloaded footings versus log time for

different treatment techniques, q=0.0 175

Figure (4.11a) Footing heave versus log Time for various plastic horizontal barrier

locations 181

Figure (4.11b) Footing heave versus log Time for various plastic horizontal barrier

locations 181

Figure (4.11c) Footing heave versus log Time for various plastic horizontal barrier

locations 182


xviii

Figure (4.11d) Footing heave versus log Time for various plastic horizontal barrier

locations 182

Figure (4.12a) Differential heave between loaded footings versus log Time for various

plastic horizontal barrier locations, q=0.5 Kg/cm2 183

Figure (4.12b) Differential heave between unloaded footings versus log Time for

various plastic horizontal barrier locations, q=0.0 183

Figure (4.13a) Heave of footing 1 versus logarithmic time for various leakage spaces

186

Figure (4.13b) Heave of footing 4 versus logarithmic time for various leakage spaces

187

Figure (4.13c) Heave of footing 2 versus logarithmic time for various leakage spaces

187

Figure (4.13d) Heave of footing 3 versus logarithmic time for various leakage spaces

188

Figure (4.14a) Differential heave between loaded footings versus log Time for various

leakage spacing, q=0.5 Kg/cm2 188

Figure (4.14b) Differential heave between unloaded footings versus log Time for

various leakage spacing, q=0.0 189

Figure (A.1) Footings heave versus logarithmic time for variable footing stress Group

1 and Group 2 222

Figure (A.2) Footings heave versus logarithmic time, Group 2 222



xix


Figure (A.5) Footings heave versus log time leakage below F1, Group 3 224

Figure (A.6) Footings heave versus log time leakage below F1, Group 3 and Group 4

224

Figure (A.7) Footings heave versus log time leakage 10cm from F1, Group 4 225

Figure (A.8) Footings heave versus log time leakage 20cm from F1, Group 4 225

Figure (B.1) Wc (%) versus depth for variable footing stress Group 1 and Group 2 226

Figure (B.2) Wc (%) versus depth, Group 2 226



Figure (B.5) Wc (%) versus depth leakage below F1 , Group 3 228

Figure (B.6) Wc (%) versus depth leakage below F1 , Group 3 and Group 4 228

Figure (B.7) Wc (%) versus depth leakage 10cm from F1 , Group 4 229

Figure (B.8) Wc (%) versus depth leakage 10cm from F1 , Group 4 229


xx

LIST OF TABLES

Table No. Page

Table (2.1): Characteristics of some clay minerals. (After Mitchell, 1976) 8

Table (2.2): Indirect Techniques for Identification and Classification of Expansive

Soils (after Snethen et al., 1975) 11

Table (2.3): Expansive Soil Classification based on shrinkage limit or linear shrinkage

after Altmeyer (1955) 13

Table (2.4): Swelling potential after Ranganatham and Satyanarayana (1965) 14

Table (2.5): Swelling potential after Sowers and Sowers (1970). 14

Table (2.6): Swelling potential after Snethen (1980) 14

Table (2.7): Swelling potential after Chen,(1987) 14

Table (2.8): Swelling potential after Seed et al. (1962b) 15

Table (2.9): Data for making estimates of probable volume changes for expansive

soils. (After Holtz and Gibbs, 1956) 17

Table (2.10): Swelling potential after the Bekkouche et al. (2001) 17

Table (2.11): Swelling potential after Chen (1988) 17

Table (2.12): Identification of potential swell based on plasticity (Carter and Benley,

1991). 18

Table (2.13): Typical values of activities for the three principal clay mineral groups 19

Table (2.14): Expansion potential from Expansion Index after Uniform Building Code

(1968). 23


xxi

Table (2.15): Estimation of clay mineralogy using cole: 28

Table (2.16): Ranges of COLE to determine soil swell-shrink potential (Thomas et al.,

2000). 28

Table (2.17): Soil expansiveness classification based on soil properties related to soil

suction (McKeen, 1992) 29

Table (2.18): Expansive Soil Classification based on Atterberg Limits and in situ

suction after Snethen, 1984. 30

Table (2.19): Direct Techniques for Quantitatively Measuring Volume Change of

Expansive Soils (after Snethen et al., 1975). 40

Table (2.20): Definitions of Volume Change Indices with Respect to Suction Changes

(after Hamberg, 1985) 57

Table (2.21): Comparison between measured and predicted footing heave percentage

(after Abdel-Moaty, 1999) 58

Table (2.22): Prediction of swell percent (After Schneider and Poor, 1974). 62

Table (2.23): Correlations for swelling soils. (After El-Sohby et al, 1995). 68

Table (2.24): Properties of untreated expansive soils (after Awad, 2005) 74

Table (2.25): Properties of untreated and treated expansive soil using sand-lime

cushion (after Abdel-Hady, 2007). 81

Table (2.26): Relation of required surcharge fill thickness to soil plasticity (Chen,

1988) 82

Table(2.27): Predicted ground surface heave for treated and untreated expansive soil

using soil replacement. 84


xxii

Table (2.28 ): reduction of final footing heave in case of treated expansive soil using

soil replacement: 85

Table (2.29): Suggested Lime Contents (Ingles and Metcalf, 1972) 103

Table (2.30) Properties of untreated and treated expansive soil-lime mix 105

Table (2.31): Typical Chemical Compositions of Class F and Class C Fly Ashes

(expressed as percent by weight) (TFHRC, 2003). 109

Table (2.32): Non-calcium based chemical soil stabilizers (after Hardcastle, 2003). 122

Table (3.1): Vertical displacement in centimeters for the specified points. 131

Table (3.2): The detected structure horizontal displacement in centimeters: 132

Table (3.3): Relative compaction of replacement soil for building (101) 135

Table (3.4): The soil profile for the boreholes around the defected building: 136

Table (3.5): Geotechnical properties of the existing clay: 137

Table (3.6): Classification of swelling potential of expansive soil samples as assessed

using empirical equations 142

Table (3.7): Swelling & swelling pressure of expansive soil samples*: 143

Table (3.8): Predicted heave according to Hamberg, 1985 and Rama et al., 1988. 144

Table (3.9): Comparison between the swelling potential according to Hamberg, 1985

(Spo) and the predicted swelling potential (Sp) for other different

equations: (Hexp= 2.00m) 146

Table (3.10): Comparison between the swelling pressure using oedometer test and

predicted swelling pressure for different equations: 147

Table (3.11): Calculated tilting angle using measured and predicted values: 156


xxiii

Table (4.1): Physical properties of the expansive soil used 158

Table (4.2): Classification of the expansive soil used 159

Table (4.3): Swelling potential of expansive soil used as assessed by different

empirical equations 161

Table (4.4): Properties of medium sand-lime mixes 165

Table (4.5): Laboratory modeling test program 168

Table (4.6): Final measured footings heave (∆Hf ) untreated expansive soil: 171

Table (4.7): Moisture content beneath footings for untreated expansive soil: 171

Table (4.8): Differential heave between footings (∆) for untreated expansive soil: 171

Table (4.9): Final measured footings heave (∆Hf) results: 176

Table (4.10): Moisture content beneath footings for various treatment techniques: 177

Table (4.11): Differential footings heave (∆) for various treatment techniques: 178


Table (4.13): Moisture content beneath footings for various horizontal barrier

locations: 184

Table (4.14): Differential footings heave (∆) for various horizontal barrier locations:

185


Table (4.16): Moisture content beneath footings for various leakage spacing: 190

Table (4.17): Differential footings heave (∆) for various leakage spacing: 190

Table (4.18): The final predicted footings heave for untreated expansive soil: 192

Table (4.19): The final predicted footings heave for various treatment techniques: 193


xxiv

Table (4.20): The final predicted footings heave for various horizontal barrier

locations: 194

Table (4.21): The final predicted footings heave various leakage spacing: 195


xxv

Notations and Symbols

Roman Letters

Ac Activity

C Colloids (or clay) content

COLE Coefficient of linear extensibility on a whole-soil base in cm cm-1

d Diameter of particle in mm

Dr Relative density

E Young’s modulus

e Void ratio

emin Void ratio of the soil at its densest possible state

emax Void ration of the soil at its loosest possible state

FS Free swell (%)

h c Matric suction

h o Osmotic suction

WLL Liquid limit

WPI Plasticity index

WPL Plastic limit

Ps Swelling pressure of the soil

SL Shrinkage limit

V Total volume

ΔV Volume change

wi Initial moisture content (%)

wn Natural water content

Wc Moisture content


1

Chapter 1

INTRODUCTION

1.1 General

Swelling behavior of shallow foundations rested on treated expansive soil is generally

affected by different factors. Some of these factors are related to the boundary

conditions controlling the` site deposition and the mode of water migration, while the

others are related to the employed treatment technique for damping the heave potential

of the expansive soil.

Several research projects have been conducted in Egypt in recent years to investigate the

fundamental mechanisms of heaving of treated and untreated expansive soils and the

influence of geological and climatic conditions on swelling potential.

Numerous works have been done to study the swelling properties of treated expansive

soil in oedometer using sand replacement (e.g. Satyanaryana, 1969; Moussa et al.,

1985; Marie et al., 2000). The behavior of footing resting on expansive treated

expansive soil using sand replacement has been investigated by (Abouleid and Reyad,

1985; Abdel-Moaty, 1999; Awad and Abdel-Hady, 2005; Awad, 2005). Many

alternatives have been proposed to mitigate the effects of expansive soils on civil

infrastructure (Rojas, 2006). Sand replacement is considered to be one of the most

effective and practical technique that can be utilized for construction of safe

foundations (Awad, 2005). Marei et al. , 2000 studied the improvement of expansive

soil using three types of cushions in oedometer i) sand+ 5% lime, ii) sand cushion and

iii) clayey silt cushion of clay content equal 20%.

Many of the heave prediction methods {such as the direct method, Jennings and

Knight (1975) “Double oedometer method,” and Sullivan and McClelland method

(1969)} are based on the oedometer test. These methods are simple to use; yet they

take into account soil in situ condition (soil structure, dry density, moisture content,


2

etc.) and are widely used by practicing engineers. Abdullah, 2002 found that for a

finite loaded area (such as plane strain condition), however, predicted footing heave

value using such methods is always overestimated. Little attention has been paid to

measure the relative movement between footings in case of untreated and treated

expansive soil (Awad and Abdel-Hady, 2005).

The main objective of the present thesis is to study the effectiveness of some widely

used treatment techniques to eliminate or damp the heave behavior of expansive soil

and also to study the factors that causes differential movement between footings. A

case study of an inclined super structure was studied; this building is bearing on a

treated expansive soil using sand replacement on the middle Mokattam plateau. A

laboratory testing program was designed to determine the swelling behavior of

shallow footings resting on treated expansive soil employing different treatment

techniques using a large laboratory model. The expansive soil was treated using sand

replacement, sand replacement with 5% lime, expansive soil +5% lime stabilization

and sand replacement with horizontal plastic barriers. Footings heave, moisture

distribution and differential heave between footings were measured and predicted

using empirical and semi empirical equations. Loading conditions, level of horizontal

barrier and water leakage spacing from footings were taken in to consideration.

Although the horizontal barrier technique is used as a preserving technique to prevent

water leakage to soil, no attention has been given to study the effect of horizontal

barrier location and the effect of water leakage space in case of deterioration of this

horizontal barrier. Accordingly the main aim of using horizontal barrier in this

research is to measure the differential heave between footings in case of deterioration

of this plastic sheet.


3

1.2 Objectives of This Research: The main objectives of this research:

1) Studying the performance of a severely cracked reinforced concrete (RC)

building constructed on treated expansive using medium sand replacement at site

in middle Mokattam plateau. Discussing the stratigraphy of the middle Mokattam

plateau and investigate the geotechnical properties of soil at the site area using

direct and indirect measurement tests. Simulate the building movement in the

vertical, longitudinal and transverse directions to recognize the reason of building

deterioration depending on six survey readings during a time period reaches

about four years nearly after the building inclination. Predicting the horizontal

displacement for the building using two semi-empirical heave equations, Stress

Changes method (according to Rama et al., 1988) and Suction Changes method

(according to Hamberg, 1985). Comparing between the predicted and measured

horizontal displacement for the building.

2) Setting a laboratory testing program to investigate the effectiveness of some

treatment techniques (medium sand, medium sand +5% hydrated lime , expansive

soil +5% hydrated lime and horizontal plastic barrier) in eliminating or damping

the swelling behavior of expansive soil using large scale laboratory model.

3) Measuring the footings heave for untreated and treated expansive soil and

evaluating the effect of footing stress, horizontal plastic barrier locations and

water leakage spacing from footings on the induced heave.

4) Predict the vertical movement of the laboratory footings models using empirical

and semi empirical formulas and comparing the predicted heave with that

measured from the laboratory model.


4

1.3 Organization of the Present Work:

The thesis is organized in five chapters as follows:

Chapter 2 contains a review for all the available literature related to the topic

classification and identification of expansive soil and employed treatment techniques

used for damping the swelling phenomenon of expansive soil. Also, the methods used

to evaluate the heave associated with wetting and swelling characteristics before and

after treatment.

Chapter 3 presents a case study of a damaged reinforced concrete (RC) building

constructed on the middle Mokattam plateau summarizes the geological formation,

geotechnical properties of the site area, Discuss the causes of cracking and deviations

that occurred to the building. Outline graphical simulation for the measured readings

of the building in the three directions. Predict the horizontal displacement of the

building using empirical and semi empirical formulas. Compare between measured

and predicted values.

Chapter 4 evaluates the swelling behavior of some of the employed treatment

techniques for expansive soil using a large scale laboratory model. The used treatment

techniques were medium sand, medium sand + 5% lime, clay+5% lime stabilization

and plastic horizontal barrier, Taking in consideration the effect of footings stresses,

plastic sheet location and leakage space from footings. Predict the vertical movement

of the shallow foundations using empirical and semi empirical formulas. Compare

between measured and predicted values.

Chapter 5 contains summary, conclusions and recommendations for further studies.

Appendix: contains Tables and Figures.

A summary of the thesis is outlined in the following flow chart.


5


6

Chapter 2

LITERATURE REVIEW

2.1 General

This chapter concerns the topic foundation on expansive soil and employed treatment

techniques used for damping the heave movements. The classification and

identification of expansive soil are thoroughly discussed. In addition, it presents the

heave prediction of foundation using empirical and semi-empirical equations.

2.2 Swelling Mechanism

Most soil classification systems arbitrarily define clay particles as having an effective

diameter of 2 microns (0.002 mm) or less. Particle size alone does not determine clay

mineral. Probably the most important grain property of fine-grained soil is the

mineralogical composition (Peck et al. 1974). For small size particles, the electrical

forces acting on the surface of the particles are much greater than the gravitational

force. These particles are said to be in the colloidal state. The colloidal particle

consists primarily of clay minerals that were derived from parent rock by weathering.

Methods of engineering classification of soils into swelling potential classes have been

to a large extent based on recognizing the presence of these minerals through their

physical/chemical properties (Olson et al., 2000).

Three most important groups of clay minerals arranged in decreasing order of

potential volume change are; i) Smectite, ii) Illite, and iii) Kaolinite.

The swelling is caused by the chemical attraction of water where water molecules are

incorporated in the clay structure in between the clay plates separating and

destabilizing the mineral structure (Figure 2.1). The magnitude of expansion depends

upon the kind and amount of clay minerals present, cation exchange capacity, clay


7

particle size, soil density, soil moisture content, soil structure and organic matter

content among other minor contributing factors.

Figure (2.1): Schematic representation of the structure of Smectite minerals

"Montmorillonites" (Colmenares, 2002).

Where Smectite is the generic name given to all expanding 2:1 phyllosilicates with a

layer charge of 0.4 to 1.2 (Bailey, 1980). The basic structure of a 2:1 mineral is an

octahedral sheet that shares oxygen atoms between two tetrahedral sheets. In

smectites, cation substitution occurs in either octahedral sheet, tetrahedral sheet, or

both. These substitutions determine the properties and chemical composition of the

smectites. Identification of the type and amount of expanding lattice clay minerals in a

soil is fundamental to its swell potential evaluation especially for smectite group,

which includes montmorillonites, those are considered as the highly expansive and

most trouble clay mineral. Particle features and engineering properties of the

important clay minerals are summarized, from Mitchell, 1976 in Table (2.1).


8

Table (2.1): Characteristics of some clay minerals. (After Mitchell, 1976)

Mineral Group Basal

Spacing (A)

Particle Features Interlayer Bonding

Specific (m2/g)

Surface

Atterberg Limits Activity=PI / clay content

WLL (%)

WPL (%)

WSL (%)

Kaolinites 14.4 Thick, stiff 6-

sided flacks 0.1 to 4 x0.05 to 2um

Strong hydrogen bonds 10-20 30-100 25-40 25-29 0.38

Illites 10 Thin stacked plates 0.003 to 0.lxl.0 to 10um

Strong potassium bonds 65-100 60-120 35-60 15-17 0.9

Montmorillonites 9.6 Thin, filmy,

flakes >10A x1.0 to10 um

Very weak Vander Waals bonds 700-840 100-900 50-100 8.5-

15 7.2

2.3 Classifications and Identifications

Identification of potential swelling or shrinking subsoil problems is an important tool

for selection of appropriate foundation (Hamilton, 1977 and Van Der Merwe, 1964).

Many tests and methods have been developed or modified for estimating shrink-swell

potential. These include both indirect and direct measurements. Indirect methods

involve the use of soil properties and classification schemes to estimate shrink-swell

potential. Direct methods provide actual physical measurements of swelling.

An indication of the potentially expansive nature of earth materials may be deduced in

the field by examination of exposures of the material and by simple field tests. The

accurate identification and study of clay minerals and their expandable properties

should be accomplished in the laboratory. There are many correlations that are useful

for identifying potentially expansive soils. It is also possible to identify them visually.

Visual indications may be (Wayne et al., 1984);

1. Wide and deep shrinkage cracks occurring during dry periods,

2. Soil is rock-hard when dry, but very stiff and sticky when wet,

3. Damages on the surrounding structures due to expansion of soil.


9

Snethen et al., 1975 summarized the purpose of identification and testing of expansive

soils is to qualitatively and quantitatively describe the volume change behavior of the

soils. The obvious need for qualitative identification is to forewarn the engineer during

the planning stages of the potential for volume change and to generally classify the

potential with regard to the probable severity. Quantitative testing is necessary to

obtain measurable properties for predicting or estimating the magnitude of volume

change the material will experience in order to ascertain approximate treatment and/or

design alternatives. With this in mind, a threefold categorization of identification and

testing techniques is possible.

1. Indirect techniques in which one or more of the related intrinsic properties are

measured and complemented with experience to provide indicators of potential

volume change. These may be grouped according to soil composition;

physicochemical, physical, and index properties; and currently used soil

classification systems.

2. Direct techniques which involve actual measurement of volume change in an

odometer-type testing apparatus. These are generally grouped into swell or

swell pressure tests depending on the need for deformation or stress related

data.

3. Combination techniques in which data from the indirect and direct techniques

are correlated either directly or by statistical reduction to develop general

classifications with regard to probable severity.

Mitchell, 1993 states that methods used to identify clay minerals in soils include X-ray

diffraction, electron microscopy, differential thermal analysis and wet chemical

analysis, this can be considered from the mineralogical point. Many private and

governmental organizations have the personnel and equipment to perform these

identification analyses. Probably the most important technique is X-ray diffraction

(XRD). This method is relatively fast, uses small amounts of material, permits

accurate identification, and may provide a semi quantitative estimation of the amount

of expandable clay minerals present.


10

Using combinations of the various methods, the different type of clay minerals present

in a given soil can be evaluated quantitatively. But because of the requirements for

special, often expensive apparatus and skills, these mineralogical methods are not

routinely used in civil engineering practice. The following discussions are an attempt

to define the techniques published in the literature with regard to the categories

previously described. As would be expected, the available techniques are quite varied

and numerous, and in some cases categorical delineation may be subjective.

2.3.1 Indirect techniques

Expansive soil as known has a large number of intrinsic properties and ambient

conditions which influence its volume change. Hence, the variety of indirect

techniques for qualifying potential volume change is just as numerous and varied. The

common techniques used are relatively routine. Table (2.2) defines and describes a

majority of the published techniques.

The indirect methods to assess soil expansiveness in terms of swell potential utilize

empirical relationships among easily measured index properties and the one-

dimensional swelling response of soils after they have been brought to some specific

initial state and then inundated. Initial states for swell tests have been specified in

terms of dry unit weight, water content and method of compaction. The swell potential

of a soil is usually described qualitatively using such terms as low, medium, high and

very high corresponding to the amount of one-dimensional volume change occurring

after the test specimen is inundated (Hardcastle, 2003).

The literature contains a considerable number of empirical techniques for assessing the

swelling potential of soils. The indirect measurements can include, atterberg limits,

colloid content, activity method, cation exchange capacity and cation exchange

activity.


11

Table (2.2): Indirect Techniques for Identification and Classification of Expansive Soils (after Snethen et al., 1975)

Indicator Group

Property and/or method

Description

Soil composition

Clay mineralogy by X-ray diffraction

Measure of diffraction characteristics of clay minerals when exposed to x-radiation. Procedure permits qualitative, and semi quantitative identification of clay mineral components based on structural differences between the clay minerals. Salvation techniques identify expansive clay minerals

Clay mineralogy by differential thermal analysis (DTA)

Identification is based upon exothermic and/or endothermic reactions which occur at particular temperatures. The type of reaction and temperature are functions of mineralogy. Heating rates, grain size, and sample size influence results. Multi component samples we difficult to analyze

Clay mineralogy by infrared radiation

Measure of selective absorption of infrared radiation by hydroxyls in clay minerals. Fair indicator, but not conclusive

Clay mineralogy by dye adsorption

Qualitative indicator based on selective adsorption of different types of dyes by different clay minerals. Accuracy decreases if more than one mineral is present

Clay mineralogy by dielectric dispersion

Measure of the radiofrequency electric properties of clay water systems. Dispersion is the measure of the dielectric constant at two frequencies. Good indicator of type and amount of clay minerals. Some problems evolve when mixtures of different expandable minerals are present in the soil

Physico-chemical

Cation exchange capacity

Measure of the ion adsorption properties of clay minerals. CEC increases from a minimum for kaolinite to a maximum for montmorillonite. Good indicator of hydration properties of clay minerals

Exchangeable cations Measure of the type of cations adsorbed on the clay minera1s. Does not directly relate to swell potential but rather to the expected degree of swell from ion hydration

Physical Colloidal content from hydrometer analysis

Measure of percent by dry weight basis of particles less than 1 micron in size. Indicator of amount of clay but no reference to type of mineral. Not conclusive

Specific surface area of clay particles

Measure of available clay mineral surface area for hydration. Fair indicator of amount of clay mineral and to some extent the type, since montmorillonite minerals are vet-y fine and result in large specific surface areas for given samples

Soil fabric by electron microscopy

No direct measure of swell potential. Primarily used for studies of the influence of soil fabric on volume change

Structure by X radiography

Good for determining the extent of cracks and fractures of undisturbed materials which will influence moisture movement. NO direct measure of swell potential

Index properties

Atterberg limits Measures of the plasticity and shrinkage characteristics of cohesive soils. Liquid limit (LL) and plastic index (PI) correlate reasonably well with swell potential primarily because there are good correlations between them and the type and amount of clay minerals present. For shrinkage limit and shrinkage index (LL-SL) the property of volume reduction is correlated with swell potential because of similarities between the phenomena. Some of the published classifications based on Atterberg limits are :

Raman (1967)

Ranganatham et al. (1965)

Ladd et al.(1961)

Degree of Expansion PI Shrinkage

Index Shrinkage

Index LL

Low <l2 <l5 <20 20-35 Medium 12-23 15-30 20-30 35-50 High 23-32 30-60 30-60 50-70 Very high >32 >60 >60 70-90 Extra high - - - - - - >90

Linear shrinkage Measure of shrinkage from a given moisture content. Reasonably good indication of swell potential

Soil classification system

AASHO A-6 and A-7 and borderline soils to A-4, A-6, and A-7 generally have high swell potentials

SCS Pedo-logical classification system in which the vertisol order is by expansive soils


12

2.3.1.1 Atterberg Limits Tests:

The use of Atterberg limits as predictors of soil behavior has been common since their

development. The Testing is relatively inexpensive, re-producible, and fast compared

to many other tests. Plasticity index "the difference between liquid limit and plastic

limit" is the most commonly used indicator of soil expansive behavior. The Atterberg

limits, which include liquid limit, plastic limit, and plasticity index, define moisture

content boundaries between states of consistency in soils (Casagrande, 1948).

Holtz and Gibbs, 1956 demonstrated that plasticity index and liquid limit are useful

indices for determining the swelling characteristics of most clays. Seed et al., 1962b

have demonstrated that the plasticity index alone can be used as a preliminary

indication of swelling characteristics of most clays. Soil plasticity limits used to

identify expansive soils include the shrinkage, plastic and liquid limits and the

plasticity index. Examples of early and still widely used plasticity criteria are given in

(Figure 2.2).

Figure (2.2): Swell potential as a function of soil plasticity index (Seed et al., 1962b)

Range Percent Swell= (0.00216) PI2.44

Swel

l, %

Plasticity index


13

The swell potential is defined as the percentage swell soil sample which has soaked

under a surcharge of 1 pound per square inch after being compacted to maximum

density at optimum moisture content according to AASHO test.

Relation between swelling potential of clays and plasticity index was presented by

Chen (1988) as follows;

Swelling Potential Plasticity index Low 0- 15 Medium 10-35 High 20-55

Very high 35 and above

Some authors consider that this potential can be linked to a single parameter. Thus, as

shown in Tables (2.3), (2.4), (2.5) and (2.6) , Altmeyer (1955), Ranganatham and

Satyanarayana (1965), Sowers and Sowers (1970), and Snethen (1980) have proposed

classifications which respectively give the swelling potential as a function of the

shrinkage limit SL, the shrinkage index SI and the plasticity index PI. The shrinkage

index is defined as the difference between the liquid limit LL and the shrinkage limit

SL.

Another classification by Chen (1987) gives the swelling potential as a function of the

plasticity index, the liquid limit and the shrinkage limit, Table (2.7).

Table (2.3): Expansive Soil Classification based on shrinkage limit or linear

shrinkage after Altmeyer (1955)

linear shrinkage Shrinkage limit (%) Probable swell Swelling potential

<5 >12 <0.5 Noncritical

5-8 10-12 0.5-1.5 Marginal

>8 <10 >1.5 Critical


14

Table (2.4): Swelling potential after Ranganatham and Satyanarayana (1965)

SI (%) Swelling potential

0-20 Low

20-30 Moderate

30-60 High

>60 Very high

Table (2.5): Swelling potential after Sowers and Sowers (1970).

Swelling potential PI (%) SL (%)

Low <15 >12

Moderate 15-30 10-12

High >30 <10

Table (2.6): Swelling potential after Snethen (1980)

PI (%) Swelling potential

<18 Low

22-32 Moderate

22-48 High

>35 Very high

Table (2.7): Swelling potential after Chen,(1987)

Vijayvergiya and Ghazzaly (1973) have proposed a classification on the basis of the

plasticity limit PL and the plasticity index. The Casagrande plasticity chart is divided

into two zones by the line A (Figure 2.3), with expansive soils above the line and non-

expansive soils below it.

Swelling potential PI (%) SL (%) LL (%)

Low <18 >15 20-35

Moderate 15-25 10-15 35-50

High 25-35 7-12 50-70


15

2.3.1.2 Colloid Content Test:

Colloids (or clay) content is defined as the percent by weight of soil particles smaller

than 0.002 mm. Colloids content is usually determined in hydrometer tests. The

influence of colloids content on soil expansion potential is through its indirect

relationship to the density of positive charge deficiency of clay minerals, the finer the

particles in a soil, the greater the likelihood that the soil contains charged particles.

Colloid content as well as Atterberg limits should be included in the routine laboratory

investigation on expansive soils. Figure (2.4) represented the relation of volume

change to colloid content, plasticity index, and shrinkage limit. The author has over

the past 15 years performed many thousands of tests on potential swell and index

properties. For soils with clay content of between 8 and 65%, Seed et al. (1962b)

propose the classification given in Table (2.8) which relates the swelling potential to

the plasticity index.

The swelling potential (S) is defined as the percentage swelling of a clay sample that

has been compacted to the optimum Proctor water content and subjected to a load of 7

kPa. It is given by the following relationship:

S = 1.10-5 Ip 2.24 (2.1)

Table (2.8): Swelling potential after Seed et al. (1962b)

Swelling potential S (%) I p

Low 0-1.5 0-10

Moderate 1.5-5 10-20

High 5-25 20-35

Swelling p

Fig

Plas

ticity

inde

x (%

)

erformance

gure (2.3):

Figu

Plas

ticity

inde

x (%

)

e of some e

: Plot of cl

ure (2.4):

Ac

expansive s

lay minera

Soil class

ctive clay

C

soil treatme

16

als on Cas

sification c

Clay Cont

ent techniq

sagrande’

chart (Ske

In

tent (%)

ques

’s chart (L

empton, 19

Non Act clay

nactive cla

Lucian, 20

953).

tive

ay

06).


17

Table (2.9): Data for making estimates of probable volume changes for expansive

soils. (After Holtz and Gibbs, 1956)

Data from index tests* Probable expansion, (% total volume

change)

Degree of expansion Colloid content, (%

minus 0.0001mm) Plasticity

index Shrinkage

limit >28 >35 <11 >30 Very high

20-13 25-41 7-12 20-30 High

13-23 15-28 10-16 10-30 Medium

>15 <18 >15 <10 Low

• Based on vertical loading of 1.0 psi.

Other classifications proposed by Bekkouche et al., 2001 and by Chen, 1988 which are

given in Tables (2.10) and (2.11). The first of these is based on the plasticity index and

the percentage of clayey particles (those with a diameter of less than 2 um). The

second is based on the liquid limit and the percentage of particles with a diameter of

less than 74 micrometres.

Table (2.10): Swelling potential after the Bekkouche et al. (2001)

PI(%) % < 2 µm Swelling potential >35 >95 Very high

22-35 60-95 High 18-22 30-60 Moderate <18 <30 Low

Table (2.11): Swelling potential after Chen (1988)

% < 74 µm WLL (%) Swelling pressure (6 years) (MPa)

Swelling potential

>95 >60 1 Very high 60-95 40-60 0.25 - 0.25 High 30-60 30-40 0.15-0.25 Moderate <30 <30 <0.05 Low


18

Carter and Bentley, 1991 proposed an empirical equation to calculate the potential

swell (Table 2.12) as follows:

Swell (%) = 60 K ( PI )2.44 (2.2)

Where PI is the plasticity index and K is the constant, equal to 3.6 x 10-5

Table (2.12): Identification of potential swell based on plasticity (Carter and Benley,

1991).

Classification of potential swell Plasticity index (%) Plasticity index (%)

Low (0-1.5%) 0-15 0-15

Medium (1.5-5%) 10-30 15-24

High (5-25%) 20-55 25-46

Very high (25+%) >40 >46

2.3.1.3 Activity Method

Plasticity characteristics and volume change behavior of soils are theorized to be

directly related to the amount of colloidal particles (< 1 μm) in soils (Anderson et al.,

1973; Nelson and Miller, 1992).

For this reason, Atterberg limits and clay content have been combined into a single

parameter called the Activity Ratio (A) developed by Skempton (1953). The activity

ratio, sometimes called the activity index, is defined as follows:

% (2.3)


19

Skempton suggested three classes of clays according to activity:

• inactive for activities less than 0.75

• normal for activities between 0.75 and 1.25

• active for activities greater than 1.25

Seed, et al., 1960, 1962(a) and Van Der Merwe, 1964 used the colloids content and

Activity to assess swell potential as shown in Figures (2.5) , (2.6a) and (2.6b). Active

clays provide the most potential for expansion. Typical values of activities for the

three principal clay mineral groups are as shown in Table (2.13).

Table (2.13): Typical values of activities for the three principal clay mineral groups

Mineral Exchangeable Ion WLL (%) WPL (%) WPI (%) WSL (%) Activity

Montmorillonite

Na+1 710 54 656 9.9 7.2

K+1 660 98 562 9.3 -

Ca+2 510 81 429 10.5 1.5

Illite

Na+1 120 53 67 15.4 0.9

K+1 120 60 42 17.5 -

Ca+2 100 45 55 16.8 -

Kaolinite

Na+1 53 32 21 26.8 0.33-0.46

K+1 49 29 20 - -

Ca+2 38 27 11 24.5 -


20

Figure (2.5): Swell potential as function of colloids content and Activity (Seed et al.,

1960)

Figure (2.6a): Soil swell potential based on size fraction and activity (Seed, 1962a)


21

Figure (2.6b): Potential severity of volume change for clay soils (Van Der Merwe,

1964).

However, as with most soil systems, the activity classification scheme does not

accurately estimate shrink-swell potential in mixed mineralogy soils. Parker et al. ,

1977 found that the activity index was too imprecise for both mixed and

montmorillonitic mineralogy soils to be useful. However Schreiner, 1988 observed

consistent trends in soil and bentonite/sand mixtures using activity index as an

indicator of shrink-swell potential (Thomas, 1998).

Another way of identifying the expansive soil is to use the activity method quoted by

Carter and Bentley, 1991. The proposed classification chart is shown in Figure (2.6a).

The activity term in the Figure is defined as follows:

(2.4)

Where PI is plasticity index and C is colloids (or clay) content.


22

2.3.1.4 Free Swell Test

Free swell, in percent, is defined as the ratio of the wet bulk volume to the dry bulk

volume. A small sample (10cm) of dry soil passing the No. 40 sieve is added to a

graduated cylinder and filled with water. Holtz and Gibbs, 1956 stated that soils with

free swells greater than 100 percent can cause considerable damage to lightly loaded

structures. Whereas soils with free swell values below 50% probably do not exhibit

appreciable volume changes. However Dawson, 1953 reported free swell values of

about 50% of several Texas clays showed extensive expansion and this is due to

extreme climatic conditions in combination with the expansion characters of the soil.

2.3.1.5 Soil Expansion Potential (ASTM D-4829)

The expansion index (EI) test is used in California to evaluate building sites (Nelson

and Miller, 1992). This test was developed in Orange County, California in the mid-

1960s and introduced in the Uniform Building Code as UBC Test Standard 29-2. It

was re-designated as UBC Test Standard 18-1 in the 1994 code. This standard was

adopted by ASTM in 1988. Soil material is disaggregated and passed through the #4

sieve and then brought to approximately the optimum moisture content (as determined

by ASTM D-1557). The optimum moisture content equates to approximately 80 to

85% of saturation. After setting for 6 to 30 hours, the moisture-conditioned soil is

compacted into a 4-in diameter mold. The moisture content is then adjusted, if

necessary, to bring the sample to 50% saturation. A 144 psf surcharge is applied and

the sample is wetted and monitored for 24 hours, measuring the volumetric swell. The

Expansion Index is calculated as follows:

EI = 100 x Δh x F (2.7)

Where; Δh = percent swell and F = fraction passing No. 4 sieve


23

Section 1803.2 of the 1994 Uniform Building Code directs expansive soil tendency be

graded by this method. The UBC mandates that “special [foundation] design

consideration” be employed if the Expansion Index is 20, or greater (UBC Table 18-1-

B). UBC Table 18-1-C may be applied to gain a “weighted index”, allowing for a

lessening of expansion with increasing depth (confinement).

Table (2.14): Expansion potential from Expansion Index after Uniform Building Code

(1968). Expansion Potential Expansion Index, EI

Very low 0-20

Low 21-50

Medium 51-90 High 91-130

Very high >130

According to ASTM, “The expansion index has been determined to have a greater range and better sensitivity of expansion potential than other indices” (such as Atterberg limits).

2.3.1.6 Cation Exchange Capacity:

The cation exchange capacity (CEC) has been used to estimate shrink-swell potential,

in addition to approximating Atterberg limits and other engineering properties of a

soil. CEC is related to amount and type of clay present in a soil. As clay content and

swelling clays increase in a soil, CEC should elevate and be reflected in an increase in

shrink-swell potential. An Alabama study conducted by Gill and Reaves, 1957 a high

correlation between CEC and plasticity indices of clayey Ultisols, Alfisols, and

Vertisols was observed.


24

Cation exchange capacity (CEC) is the quantity of exchangeable cations needed to

balance the negative charge on the surface of clay particles and is usually expressed in

milli-equivalents per 100 grams (meq/100g) of dry clay. Figure (2.7) is a relationship

developed by Holt , 1969 in which a normalized cation exchange capacity, CEC , and

the activity are used to indicate the presence of expansive clay minerals without the

need to perform the identification procedures mentioned above. Normalized cation

exchange capacity, CEC, is the conventional cation exchange capacity in milli-

equivalents per 100 grams divided by the colloids content in percent.

High CEC values indicate a high surface activity. In general, swell potential increases

as the CEC increases and the total CEC value is the summation of the individual CEC

values of each of the present clay minerals, as a fraction of the total clay content.

2.3.1.7 Cation Exchange Activity:

Pearring , 1963 used cation exchange capacity, CEC, and plasticity as two parameters

to classify soils as to a predominant mineral type. Pearring normalized these two

parameters based on the percent fine clay content.

This normalization yielded two new parameters, the activity ratio (Ac) and the cation

exchange activity (CEAc) as follows, Figure (2.8) illustrates the classification

developed by Pearring (1963).

c%

% % .

(2.5)

.

% % .

(2.6)


25

McKeen, 1981 used a mineralogical classification similar to that of Pearring, 1963

defining regions charted against Ac and CEAc axes including predicted COLE values.

Hamberg, 1985 updated the classification chart and included an adjustment for clay

percentage.

Then the approach continued to be refined in Nelson and Miller , 1992 that produced

a more simple general classification scheme using CEAc and Ac axes as shown in

Figure (2.9).

As can be surmised from the discussion above, several physical, chemical, and

mineralogical soil properties influence shrink-swell behavior, with no one property

accurately predicting shrink swell potential for all soil types. Often, most expansive

soils are clayey with high cation exchange capacities, high specific surface areas, and

high liquid limits. Smectite typically comprises a significant portion of the soil clay

fraction.

Figure (2.7): Clay mineralogy as a function of Activity and Cation Exchange Capacity

(Holt, 1969)


26

Figure (2.8): Mineralogical classification from Pearring,1963.

Figure (2.9): Expansion potential as a function of CEAc and Ac from Nelson and

Miller, 1992.


27

2.3.1.8 Coefficient of Linear Extensibility (COLE)

COLE is a kind of reverse swelling and is determined in a test involving finding the

dry unit weight of the soil for the two specified conditions (Nelson and Miller, 1992).

The coefficient of linear extensibility (COLE) is used routinely by the National Soil

Survey Laboratory to characterize shrink-swell potential of soils (Soil Survey Staff,

1996).

The COLE test determines the linear strain of an undisturbed unconfined sample on

drying from 33 kPa suction to oven dry suction. The procedure involves coating

undisturbed soil samples (clods), with a flexible plastic resin. The resin is

impermeable to liquid water, but permeable to water vapor. Natural clods of soil are

brought to a soil suction of 33 kPa in a pressure vessel. They are weighed in air and

water to determine weight and volume using Archimedes principle. The samples are

then oven dried and another volume measurement is performed in the same manner.

COLE is a measure of the change in sample dimension from the moist to dry state and

is estimated from the bulk densities of the clod at a suction of 5 psi and oven dry

moisture conditions. The value of COLE is given by:

COLE = ΔL /ΔLD

= (γdB

/γdM

)0.33

-1 (2.8)

where ;

ΔL /ΔLD

= linear strain relative to dry dimensions

γdB

= bulk density of the oven dry sample

γdM

= DBM is bulk density of the sample at field capacity

The National Soil Survey uses COLE as an estimator of clay mineralogy. The ratio of

COLE to clay content is related to mineralogy as shown in Table (2.15).


28

Table (2.15): Estimation of clay mineralogy using cole:

COLE /Percent Clay Mineralogy

>0.15 Smectites

0.05-0.15 Illites

<0.05 Kaolinites

The COLE index was found to be closely related to a number of soil variables (Parker

et al., 1982). McKeen and Hamberg, 1981 extended the CEAc-Activity relationships

of Figure (2.9) to develop both an approximate method for estimating directly the

qualitative swell potential of a soil based on the COLE and a method for estimating

the COLE value itself.

Once a COLE has been determined, the qualitative swell potential of the soil can be

estimated from COLE and the colloids content using Figure (2.10). If the COLE

values are not available, but activity and cation exchange capacity data are, Figure

(2.11) can be used to identify the appropriate region of Figure (2.10) and thus provide

estimates of both soil expansiveness and COLE. According to the calculated COLE, a

range of soil swell-shrink potential can be distinguished based on data in Table (2.16).

The quantitative swell potential can also be estimated by correlating the colloids

content and the COLE using Figure (2.10).

Table (2.16): Ranges of COLE to determine soil swell-shrink potential (Thomas et al.,

2000).

Soil swell-shrink potential COLE

Low <0.03

Moderate 0.03-0.06

High 0.06-0.09

Very high >0.09


29

2.3.1.9 Soil water characteristic curve

McKeen, 1992 incorporates the dimensionless slope of the soil suction-water content

relationship of a soil, Δh/Δw, and the suction compression index, Ch. An example of

the suction-water content relationship for a soil (also called a soil water characteristic

curve or SWCC) is shown in Figure (2.12).

The dimensionless slope of the SWCC is obtained when suction is expressed in units

of pF, and water content is expressed in percent. The suction unit pF is the logarithm

to the base 10 of the suction head, h, expressed in cm of water.

The dimensionless suction compression index, Ch, is the ratio of the change in

dimensionless vertical strain, εvolume, produced by drying or wetting a soil divided by

the change in the suction, in units of pF, accompanying the change in water content.

Figure (2.13) illustrates the definition of suction compression index for one

dimensional volume changes. Classification of soil expansiveness using these criteria

is given in Table (2.17) and (2.18).

Table (2.17): Soil expansiveness classification based on soil properties related to soil

suction (McKeen, 1992)

Category Δh/ Δw (pF/%)

Ch εvol. / pF

Expansiveness Classification

I > -6 -0.227 Very High

II -6 to -10 -0.227 to -0.120 High

III -10 to -13 -0.120 to -0.040 Moderate

IV -13 to -20 -0.040 to NE1 Low

V < -20 1Non-Expansive 1Non-Expansive


30

Table (2.18): Expansive Soil Classification based on Atterberg Limits and in situ

suction after Snethen, 1984.

WLL (%) PI (%) Natural Soil Suction Potential Swell % Potential Swell Classification

<30 <25 <1.5 <0.5 Low

30-60 25-35 1.5-4 0.5-1.5 Marginal

>60 >35 >4 >1.5 High

The testing effort required to evaluate the expansiveness of a soil using these criteria

goes well beyond measuring soil index properties, but as was the case for the approach

described in the previous paragraphs, it is possible to use the criteria even if only

index properties are available.

For example, empirical relationships to estimate the soil water characteristic curves

from plasticity and grain size data have been published in both the soil science and

geotechnical literature.

Figure (2.14), from Zapata et al., 2000 gives soil water characteristic curves of fine-

grained soils as functions of plasticity index and percent fines. The curves in the figure

could be used to estimate the slope of the SWCC, and Figure (2.15) could be used to

estimate the suction compression index, Ch, as a function of the slope of the SWCC .


31

Figure (2.10): Swell potential as a function of colloids content and COLE (McKeen

and Hamberg, 1981)

Figure (2.11): Soil expansiveness and COLE regions as a function of Activity and

Cation Exchange Capacity (McKeen and Hamberg, 1981)


32

Figure (2.12): Example of the relationship between soil suction and water content

(McKeen, 1992)

Figure (2.13): Example of the relationship between volume strain and soil suction

(McKeen, 1992)


33

Figure (2.14) Relation of soil water characteristic curves, soil plasticity and percent

fines (Zapata, et al., 2000)

Figure (2.15): Relation of suction compression index, Ch, to the slope of the soil water

characteristic curve (McKeen, 1992).


34

2.3.2 Direct techniques

Direct measurements are considered the most satisfactory and convenient method for

determining the swelling characteristics of expansive soils. Direct measurements of

expansive soils can be achieved by the use of the conventional one-dimensional

consolidometer. The direct measurements were developed to determine the swelling

pressure developing in expansive soil when it flooded both from the bottom and from

the top. Researchers used different methods and equipment to investigate the swelling

behavior, there are "four" main methods for measuring the swelling pressure.

2.3.2.1 Constant volume test

In this method, the soil sample involves inundating, the sample in the oedometer while

preventing the sample from swelling. The swell pressure is reported, as the maximum

applied stress required maintaining constant volume. Once the swelling pressure stops

increasing after soaking, the sample may be rebounded by complete load removal or

incremental load removal. Idealized plots of constant volume tests data are shown in

Figure (2.16) (Porter and Nelson, 1980). The analysis of oedometer tests must take

into account the loading and wetting sequence, surcharge pressure, sample disturbance

and apparatus compressibility.

2.3.2.2 Double oedometer test

The double oedometer procedure involves testing two undisturbed samples. One

sample is consolidated at its natural moisture content, the other sample is undated

while subjected to a small initial load and then consolidated under saturated

conditions. Typical results for initially moisture and initially dry sample pairs are

shown in Figure (2.17a). The curve for the sample tested at natural moisture content is

used to obtain the in situ void ratio (e0), corresponding to the total in situ stress, σ0.

The final void ratio (ef), is found from the saturated compression curve after

calculating the final effective stress.


35

The change in void ratio during heave is:

∆e = ef - e0 (2.9)

where;

e0 = initial void ratio corresponding to the initial total stress(σ0) on the natural

moisture consolidation curve, and

ef = final void ratio corresponding to the final effective stress(σf) on the

saturated consolidation curve.

For initially dry sample pairs, the natural moisture content curve is sometimes

displaced above the saturated curve at high loads as shown in Figure (2.17b) [Burland,

1962]. Burland noted that the amount of displacement between the straight-line

portions of these curves was largely dependent on the initial moisture content. He also

observed that; the rebound portion of the natural moisture content curve was flat when

unloading took place quickly. Burland, 1962 suggested a revised method analysis.

This procedure is shown in Figure (2.17b).

2.3.2.3 Simplified oedometer test

The simplified oedometer test is a modified consolidation swell test. The simplified

procedure was devised as an alternative to the double oedometer test method, initially

proposed by Jennings and Knight, 1957. The simplified procedure involved as a result

of observations made during testing of the natural moisture content samples (Jennings

et al., 1973). In the original tests, the consolidation of a specimen at natural moisture

content was performed solely for the purpose of obtaining the initial condition (e0, σ0)

Figure (2.17a).

Fredlund, 1983 named this method as free swell in oedometer and showed that the

swell pressure of soil is determined through the one-dimensional restrained swell test

by utilizing the oedometer apparatus. The undisturbed soil specimen is cut at its in-situ

moisture content, put in an odometer, saturated and brought to equilibrium under a


36

surcharge of about 1 kPa. The load on the specimen is increased periodically until the

height of the specimen returns to origin. For each increment of load, the specimen is

allowed to consolidate fully before the application of the next load. The amount of

swell is recorded with the dial gauge and the maximum vertical stress necessary to

attain original height of sample is the swelling pressure. A graph can be obtained of

height or void ratio against stress (Figure 2.18). This test has the advantage that only

one sample is required and apart from free swell, the consolidation characteristics can

be determined.

That value provided an estimate of the initial in situ void ratio of the saturated

specimen for the prediction of heave. It was recognized that (e0, σ0) could be obtained

by loading a single specimen to (σ0) at its natural moisture content, then unloading to a

light seating load of 0.01 ton/ft2 (0.14 psi or 1.0 kpa) and performing the saturated

swell -consolidation test as usual. The results are illustrated in Figure (2.19a).Other

investigators observed that the slopes of the natural moisture content compression and

rebounded curves were very flat up to the pressure (σ0) [Burland, 1962, Ralph and

Nagar, 1972]. Thus, little error was introduced by assuming that the in situ void ratio

(e0) corresponding exactly with the initial sample void ratio (e0 sample) was as shown

in Figure (2.19b). This simplified oedometer procedure; Jennings et al., 1973 analyzed

previous results obtained from double oedometer tests [Jennings and Kerrich, 1962].

These researchers found that the heave values predicted by the simplified procedure

were close to that predicted by double oedometer analysis. Therefore the simplified

test eliminated the uncertainties associated with effects of very dry soils and

differences in initial void ratios of sample pairs also it eliminated the need to carry

tests to very high loads to locate the virgin compression lines. For large initial loading

values and /or where the curve for the sample at natural water content has significant

slope the simplified procedure can under predict heave ,the double oedometer test

would be preferred in such cases.


37

Figure (2.16): Typical constant volume swell test results (After Porter and Nelson,

1980).

Figure (2.17): Double oedometer test results, Initially moist sample pair (Jennings

and Kerrich, 1962)


38

Figure (2.18): Free swell under load in the oedometer (after Fredlund, 1983)

Figure (2.19): Simplified oedometer test analysis (After Jennings et al., 1973)


39

2.3.2.4 Pre-swell sample method

In the pre-swell method, the specimen is allowed to swell under a token pressure by

submerging the specimen in distilled water. The soil specimen is then loaded and

unloaded following the conventional oedometer test procedure.

The swelling pressure is usually defined as the pressure required recompressing the

fully swollen sample back to the initial volume. An idealized plot of consolidation

swell test data is shown in Figure (2.20), where (σo) represents the stress at which the

sample is wetted and (σs) represents the swelling pressure according to the above

definition.

Figure (2.20): Typical plot of consolidation - swell test results (After Jennings et al.,

1973)

Table (2.19) defines and describes some of the various published procedures in which

the swell and swelling pressure of both undisturbed and remolded soils have been

measured.


40

Table (2.19): Direct Techniques for Quantitatively Measuring Volume Change of

Expansive Soils (after Snethen et al., 1975).

Method Description Navy method Odometer test on remolded undisturbed samples in deformations under various surcharges are

measured to develop a surcharge versus percent curve. The surcharge versus percent swell curve is related to the depth of clay percent swell curve from which the magnitude of volume change is calculated as the area under the curve

Potential vertical rise method

The correlation of measured volumetric swell of a specimen around pressure of 1 psi) with classification test data PI, and percent soil to determine the Family Number(predetermined correlations) for the soil. The vertical pressures at the midpoints of strata are calculated and used in with Curves to obtain percent volumetric swell under actual loading conditions in each strata. The linear swell is take" as one-third of the volumetric swell which is cumulatively summed to calculate the potential vertical rise

Noble method Odometer on statically compacted samples (total four, two initial moisture contents under two surcharge pressures) measuring deformation. Previously correlated data are consulted to determine the magnitude of volume change with changing loading and initial moisture conditions

Double odometer method Odometer test in which two adjacent undisturbed are subjected to differing conditions. One sample is inundated and allowed to swell to equilibrium, then consolidation-tested using routine procedures. The second sample is consolidated-tested using routine procedures at its natural moisture content The virgin portion of the NMC curve is to coincide with the swell- consolidation curve, and relationships from consolidation theory are used to estimate volume change

Simple odometer method Odometer test using one undisturbed sample which is loaded to its in situ overburden pressure the" unloaded to seating load, inundated, and allowed to swell to equilibrium, the" consolidation- tested using routine procedures. Analytical procedures double odometer method

Sampson, Schuster, and Budge method

Odometer test in which two undisturbed or remolded samples are subjected to different loading conditions. One sample is loaded to the testing machine capacity (32 tsf reported) and consolidated to equilibrium, inundated, unloaded to 0.1 tsf, and allowed to swell to equilibrium. The second sample is loaded to its in situ overburden pressure, inundated, unloaded to the planned structure load, and allowed to swell to-equilibrium. The swelling index and changes in void ratio and consolidation theory are used to determine amount of volume change

Lambe and Whitman method

Odometer test in which undisturbed or remolded samples are consolidation-tested using routine procedures including rebound. Effective stresses are calculated before and after testing, and the associated void ratio changes are determined. From this ∆e/l + eo or ∆H/H* versus depth curves are plotted. Magnitude of volume change is equal to area under the curve

Sullivan and McClelland (constant volume swell) method

Odometer test in which a" undisturbed sample is loaded to its in situ overburden pressure, inundated, and swell pressure by maintaining constant volume, then unloaded to light seating load and the swell measured. Changes in void ratio are taken from the curve corresponding to the initial and final effective stress conditions of the in situ soil. Consolidation theory is used to estimate volume change

Komornik, Wisema", and Ben-Yaacob method

Odometer test on undisturbed samples in which swell is measured under corresponding overburden pressures to develop depth versus percent curve. Magnitude of volume change is equal to area under curve

wong and Yang method Same as previous procedure except that a" additional surcharge equal to the pore water suction at hydrostatic conditions is added. Same analytical procedures

Expansion Index (Orange County) method

Odometer test on compacted samples measuring volume change under l-psi surcharge

Third cycle expansion pressure test method

Used in conjunction with standard R-value test. Swelling pressure is measured at the end of the third cycle of volume development (i.e., swell pressure is developed and relieved twice, then measured after developing the third time)

* ∆e = change in void ratio ; eo = initial void ratio ; ∆H = change in height ; A = height.


41

2.3.3 Combination techniques

Combination techniques involve the correlation of indirect and direct techniques to

provide better classification groups with regard to severity of volume change and

develop quantitative estimation techniques for ultimate volume change. Commonly

used correlation parameters include Atterberg limits (liquid limit, plastic index,

shrinkage limit), colloidal content, activity, and swell or swelling pressures from

odometer test under various loading conditions.

Generally the techniques result in a categorization of the relative severity of volume

change; however, in some cases prediction equations are obtained from Statistical

comparison of measured properties (Snethen et al., 1975). The following paragraphs

present some of the more widely published techniques with brief descriptions of their

application for heave prediction.

2.4 Heave prediction

The total heave is the maximum potential magnitude of heaving of structure. The

relation between swelling potential in the laboratory and the total heave depends on

some variables such as climate, soil profile, g round water and drainage. Prediction

methods of the amount of total heave can be separated into two broad categories.

These are described as semi-empirical methods and empirical methods (Nelson and

Miller 1992).

2.4.1 Semi-Empirical Methods

They are all based on the oedometer test, which similarly attempts to model field

behavior.

2.4.1.1 Heave prediction using (Texas Method No. 1)

In this method, the oedometer is used directly to model the field conditions. An

undistributed sample from depth "z" is placed in the oedometer and it is loaded with


42

the total overburden pressure (σv = σz). Care is taken to preserve the water content.

The additional pressure due to the structure (σv) is then added and the soil is flooded

with water. At this stage the specimen is left until all swelling movements cease. From

then, the test proceeds as a normal consolidation test, the resulting curve is as shown

in Figure (2.19a). The total heave is calculated as consolidation settlement as-

following;

∆H =H×∆e

1+eo (2.10)

Where; H refers to any particular stratum or layer thickness appropriate to the test

which yields e and eo as in Figure (2.19a). This was one of the first methods tried in

South Africa about (1950) on a site where the heave of a structure was already known,

it yielded a result which was about one half of observed heave. This was attributed to

the fact that when the specimen was loaded with the pressure (σv), the fissure in the

soil closed making it very difficult to secure water entry.

2.4.1.2 Heave prediction using double oedometer test

This was devised in the attempt to overcome the difficulties described in the above

method. It was based on the effective stress change theory proposed by Jennings and

Kerrich, 1962. Two adjacent samples are tested, one at the natural moisture content

and the other in a soaked or loaded condition. Flooding is performed at a low vertical

pressure (normally 10 kpa), since at this low pressure the fissure will not be closed and

the water easily enters the soil structure.

A most important observation is that swelling must be observed to take place. To take

account of the commonly observed differences in initial void ratio between adjacent

specimens, the two curves are adjusted to make their virgin lines coincide as shown on

Figure (2.19a) where (eo) is the initial void ratio of the sample which will be soaked


43

under the low pressure (1.0 kpa), (es)s is the void ratio after soaking under this low

pressure, and (es) is the void ratio of the soaked specimen under the starting pressure

(10.0 kpa). For the sample at natural moisture content (eo)n is the void ratio under the

saturation pressure (10.0 kpa).

It will be noted that in the natural moisture content test, (σo) is in terms of total stress

while for the soaked test, (σo) is in terms of effective stress.

The first mentioned test is used to obtain the point (eo, σo), the second for the point (ef ,

σf ). Being in terms of effective stress, the final stress (Pf) must therefore include

overburden pressure (σz), pressure increment (∆σz) and the pore water which here is

the negative capillary pressure. The heave is again given by equation (2.11) as

follows;

∆ H = H × ∆e 1+eo

(2.11)

The double oedometer test has worked very well on many sites in south Africa. In

certain circumstances it has been suspected that the predicted heave may be too large

and this observation applies particularly to very dry sand clays as pointed out by

Burland, 1962. This makes the overproduction more important in the upper portions of

expansive clay stratum.

2.4.1.3 Heave prediction using simplified oedometer test

Firth, 1971 drew attention to the fact that in the double oedometer test the natural

moisture content curve is used only for obtaining the point (eo , σo). He suggested that

this could be found by loading the natural moisture content sample (σo) then

unloading to the small pressure (1.0 kpa) and then flooding the sample and proceeding

in the ordinary way employed for the soaked test. All the other features of the double

oedometer test are retained.


44

The principles of Firth's modification are shown in Figure (2.19a). Ralph and Nagar ,

1972 used the method proposed by Firth to calculate the heave for a profile in

Vereeniging Transvaal an area which is notorious for the heaving subsoil conditions.

These workers made a further observation. The slope of the natural moisture content

curves was so flat up to the loading(σo) that little error was introduced by accepting

that (eo) at (σo) was identical to the (eo) at the (10.0 kpa) load on the specimen.

The simplified test procedure is illustrated in Figure (2.19b). These workers show that

where the soils are drier, the unit heave ∆e1+eo

calculated from simple test (after re-

analyzed by Jennings and Kerrich, 1962 using the principles of the simple oedometer

test) is generally lower than that found from the double oedometer test. The simple

oedometer test appears to be most suitable and convenient method for the prediction of

total heave.

2.4.1.4 Heave prediction using constant volume test

Richard et al., 1969 presented a method for predicting the heave of light structures

from constant volume swell test. The sample in this method is trimmed into the

consolidometer and it is loaded incrementally to a vertical pressure equivalent to total

overburden pressure in the field as shown by dash curve in Figure (2.21).

The load increments are usually held for time intervals ranging between half an hour

and hour. The specimen is then submerged in water, the sufficient load is applied in

small increments to prevent swelling unit the swelling pressure is fully developed. The

measured swelling pressure is an effective stress, because the soil suction has been

nullified and the pore water pressure is zero. The submerged sample is then unloaded

in increments from the swelling pressure to a pressure of 1 kg/cm2. Initial soil suction

equal to the swelling pressure determined from a laboratory test minus the total

overburden pressure, as shown by the horizontal dash line in Figure (2.21).


45

The amount of heave occurring in an expansive stratum of thickness will be;

∆ ∆

∆ ∆

(2.12)

Where;

∆f is the unit swell of the soil due to stress decrease, and fo , is zero or negative value

indicating slight compression of the sample under overburden pressure before

submergence. The predicted heave from this method over - estimates the observed

heave at the center of one story office building in Texas by about 10 -20 % and this

method is unsuitable for estimating time rate of heave.

Correction factor for oedometer test data using constant volume test

The constant volume oedometer test was recommended by Fredlund, 1983 as the best

testing method for predicting expansive soil movement using the concepts of state of

stress presented above Figure (2.22). The effect of the matric suction stress state

variable is assumed to be transferred upon soaking to the saturated effective stress

plane along the idealized stress path ø-A-2 shown in Figure (2.23) (Fredlund, 1983).

The magnitude of the transferred stress or "matric suction equivalent" will generally

be less than the in site matric suction [Yoshida et al., 1983]. The maximum swell

pressure (ps) as shown in Figure (2.21) theoretically represents the initial stress state,

in terms of an equivalent saturated effective stress. The corrected swell pressure (ps')

may be determined from constant volume test data using a graphical procedure similar

to Casagrande's construction.


46

Figure (2.21): Laboratory relationship between void ratio and effective pressure(After

Richard et al., 1969).

Figure (2.22): Idealized three dimensional loading surface for unsaturated soils in

terms of void ratio versus indepented stress (After Fredlund, 1983).


47

Figure (2.23): Correction of constant volume swell test data for sample disturbance

(After Fredlund, 1983).

Figure (2.24): Idealized and actual versus analysis stress path for prediction based on

constant volume (After Fredlund, 1983).


48

Fredlund, 1983 suggested that a correction should first be applied for the

compressibility of the consolidation apparatus itself. The graphical correction for

sampling disturbance is shown in Figure (2.24). The corrected swelling pressure is

designated as the intersection of the bisector of the formed by these lines and a line

tangent to the curve, which is parallel to the slope of the rebound curve. The corrected

swelling pressure may be as much as two to three times the magnitude of the

uncorrected swelling pressure. The equation for the rebound portion of the oedometer

test can be written as follows;

(2.13)

Where ;

ef = final void ratio.

eo = initial void ratio.

cs = swelling index.

'δf = final effective stress (δo + ∆δ - Uwf)

'δs = corrected swelling pressure.

Total heave is the sum of the displacement in each soil layer and is written in terms of

change in void ratio;

∆ ∆ ∆ (2.14)

Where ;

∆H = total heave,

∆zi = heave of layer (i),

Zi = thickness of layer(i),

∆ei = change in void ratio of layer (i) = (ef - eo) = [ Cs log δf / δs ]i , and

n = number of layers


49

2.4.1.5 Heave prediction using stress change method "Closed Form Heave

Equation"

The closed form heave solution was carried out by Rama et al., 1988 for evaluating

the total heave in swelling soils, where the heave prediction requires a knowledge of

some important variables first; the initial in situ state of stress (σi), the second is the

swelling index (Cs), and the third is the finial state of stress (σf).

The initial stress state and the swelling index are commonly obtained from one

dimensional oedometer tests, the final stress state may be strongly influenced by local

experience [Fredlund, 1983]. The stress state variable change between the initial and

final conditions together with the swelling index is used to predict the amount of

heave.

A general equation for the prediction of heave was published by several researches

Rama et al., 1988 stated that the heave stress path follows the rebound curve (i.e., Cs)

from the initial stress state to the final stress state, the equation of rebound portion of

the oedometer test can be written as follows ;

∆ (2.15)

Where;

∆e =change in void ratio (i.e; ef - eo) corresponding to the corrected swelling pressure

(σs ),

eo = initial void ratio,

ef = final void ratio,

Cs = swelling index,

σf = final stress state, and


50

σi = initial stress state = corrected swelling pressure σS`

The initial stress state (σ0) can be formulated as the sum of the overburden pressure

and the matric suction equivalent as follows;

σo = ( σv - ua ) + ( ua - uw ) (2.16)

Where;

σv = total overburden pressure,

ua = pore-air pressure, uw= pore-water pressure,

σv - ua = net overburden pressure, and

ua- uw = matric suction equivalent.

Equation (2.20) can be simplified as follows;

σo = σv - uw (2.17)

The final stress state (σf) must account for total stress changes and the final pore water

pressure conditions

σf = σv ± ∆σv - uwf (2.18)

Where :

∆ σv = change in total stress, and

uwf = estimated final pore water pressure.

Rama stated that one of the three possibilities provide the most logical estimation of

the final pore water pressure. First, it can be assumed that the water table will rise to

ground surface creating a hydraulic condition and produces the greatest heave

prediction.


51

The second, it can be assumed that the pore water pressure approaches zero

throughout its depth, where this assumption may be a realistic assumption, however, it

should be noted that it is not an equilibrium condition. A good agreement between

heave analyses and field data is when zero final pore water pressure was assumed. The

third, it can be assumed that under long term equilibrium conditions the pore water

pressure will remain slightly negative, this assumption produces the smallest

prediction of heave. This is due to the fact that most of the heave occurs in the

uppermost soil layer where the matric suction change is largest. The heave of an

individual soil layer can be written in terms of change in void ratio as follows;

∆

∆ (2.19)

Where;

∆hi = heave of an individual layer,

hi = thickness of the layer under consideration, and

∆e = change in void ratio of the layer.

The change in void ratio in equation (2.15) can be substituted by equation (2.19) to

give the following form

∆ (2.20)

Where;

σf = final stress state in the soil layer, and

σo = intial stress state in the soil layer.


52

The total heave from several layers (∆H) is equal to the sum of the heave for each

layer;

∆ ∆ (2.21)

2.4.1.6 Heave prediction using suction change method

Hamberg, 1985 showed that the relationship between void ratio and water content, to

be linear over a range of water contents greater than the shrinkage limit, and this

relationship is shown in Figure (2.25). The slope of the curve is designated Cold index

(Cw) and is equivalent to the compressibility factor (α).

The parameter (Cw) is an index of volumetric compressibility with respect to water

content and is defined as the ratio between the suction index with respect to void ratio

and the suction index with respect to water content. The parameter (Cw) is defined as

the modulus ratio and is given by;

∆∆ 2.22

The heave (∆Zi) for a uniform layer of thickness (zi) can be determined from this equation;

∆∆e1 eo

∆w

1 eo 2.23


53

The total heave (∆H) is the sum of all increments of heave for each layer and equal

∆ ∆∆w

1 eo (2.24)

Hamberg, 1985 concluded that the equation is similar in concept to that of Fredlund,

1983 but it considers heave only due to suction changes and not changes in effective

stress. Initial water content profiles can be measured during the preliminary site

investigation. The final water content profile after construction must be predicted on

the basis of soil and ground water conditions and environmental factors. In areas of

shallow ground water, it may be reasonable to assume that full saturation with zero

suction could develop to the near surface. Experiments conducted at Colorado State

University have provided data on seasonal variations of water content profiles under

floor slabs (Hamberg, 1985). On the basis of those experiments and measurements of

suction on the pierre shale to depths greater than 25 ft. (8m) using the filter paper

method, an idealized initial and finial water content profile was developed by

Hamberg, 1985 as shown in Figure (2.26). Although the active zone was deeper than 6

ft (1.8m) moisture variation below 6 ft was small at the time of heave prediction. It

was observed that generally the initial water content near the surface did not fall much

below the shrinkage limit. Also the maximum water content under the simulated floor

slab did not significantly exceed the plastic limit. Consequently, as shown in Figure

(2.26), these values represented the minimum and maximum water contents at the

surface for initial and finial water contents profiles respectively.

An extensive laboratory and field study was conducted by the U.S. Army corps of

Engineers Waterways Experiment Station (WES), to evaluate soil suction and

mechanical prediction models for foundation design (Johnson, 1977). Comparison of

laboratory procedures between suction test methods and the oedometer test method

showed that, suction test methods were simpler, more economical and more expedient


54

(Johnson, 1977). The suction index was not measured directly, but was calculated as

follows;

Cw = α Gs /100 B (2.25)

Where; α =compressibility factor (slope of specific volume versus water content

relationship),

B= slope of suction versus water content relationship, and

Gs= specific gravity of solids.

The compressibility factor (α) may be estimated from the following empirical

relationships (Croney et al., 1958)

α = 0 P.I<5

α = 0.0275 PI-0.125 5<P.I < 40

α =1.0 P.I > 40

The suction index (Cw) appears to be related to the swell index Cs, when α is less than

unity and to the compression index Cc when α is equal to one.

The following equation was found to be adequately represent in the suction - water

content relationships for numerous clay soil with suction ranging from 15 psi to 750 psi

(10.0 to 5000kpa):

log (Ua – Uw) =A + B WC (2.26)

Where;

Ua - Uw = soil suction, and

Wc = gravimetric water content.


55

Figure (2.25): Void ratio versus water content (After Hamberg, 1985),

Predicted soil suction (Kpa)

Figure (2.26): Idealized moisture boundary profile for the Pierre Shale, fort collins

(Hamberg, 1985).


56

A modified test procedure was developed at the new Mexico Engineering Research

Institute to predict heave beneath air field pavements [Mckeen and Hamberg, 1981].

The CSU-COLE test was developed to provide a simpler and more informative test

for- routine engineering application.

The basic difference between the CSU-COLE test and the COLE test procedures is

that samples are brought to a variety of initial moisture conditions before the first

Saran coating in the CSU-COLE test. The initial soil suction of each sample is then

measured independently, using thermocouple psychrometer or filter paper denser.

After the initial suction measurement, the soil samples are coated with resin according

to the COLE procedure specifications, and the initial sample volumes are measured.

The samples are allowed to dry slowly in the air, with periodic volume and weight

measurements taken until each sample reaches a content weight under laboratory

humidity conditions. The samples are then oven dried for 48 hours and a final volume

and weight measurement is taken.

This data provides the void ratio and water content. Although void ratio is related to

soil suction, it can also be related to water content because the latter is also directly

related to suction.

From the heave prediction using [CSU -COLE] can be concluded that;

a) Heave prediction in this method only due to suction changes and not due to

changes in effective stress.

b) Final moisture content did not exceed the plastic limit near the water.

c) Moisture cont did not change after the active depth approximately reached zero.


57

Hamberg, 1985 summarized the definitions of Volume Change Indices by different

methods with Respect to Suction Changes as shown in Table (2.20).

Table (2.20): Definitions of Volume Change Indices with Respect to Suction Changes

(after Hamberg, 1985) 2.4.1

Reference Symbol Definition Typical Values

Formation or Soil Type

Location

Fredlund, 1979 Cm Slope of void ratio versus log matric suction: Cm = ∆e/∆ log (ua - uw)

0.1-0.2 Regina clay Canada

Johnson, 1977 CT Slope of void ratio versus log matric suction, approximated by CT = αG,/100 B where α= compressibility factor (0 < α < 1), B = slope of the suction versus water content relationship

0.07 0.15-0.21 0.09-0.23 0.07-0.15 0.13-0.29

Loess Yazoo clay

Upper Midway Pierre Shale Marine clay

Mississippi Mississippi

Texas Colorado

Sicily

Lytton. 1977 McKeen, 1981

γh Slope of the volumetric strain versus the log of total suction: ∆ /

∆ where

∆e/(1 + eo) = volumetric strain, h = total suction

0.02-0.18 0.02-0.20 0.05-0.22

Engle ford Yazoo clay

Mancos

Texas Mississippi

New Mexico

Aitchison and Martin (1973)

I''pt Slope of vertical strain versus the log of total suction: I''pt = εv / ∆ log h (ua - uw)

0-0.08

Fargher et al. (1979)

I''pm

Slope of vertical strain versus the log of matric suction: I"pm = εv /∆ log (ua - uw)

0-0.11 Red-Brown Clay

Adelaide, S. Australia

I"ps Slope of vertical strain versus the log of solute (osmotic) suction: I"ps = εv /∆ log л

0-0.20

Grossman et al., 1968 U.S.D.A. Soil Conservation Service, 1971

COLE Value of linear strain corresponding to a suction change from 33 kPa to oven dry: COLE = ∆L/∆LD = ( γd / γw ) 0.33 –1 where ∆L/LD = linear strain relative to dry dimensions, γD = bulk density of oven dry sample, γw = bulk density of sample at 33 kPa suction

0-0.17 Western and Midwestern U.S. soils


58

Abdel-Moaty, 1999 compared between observed and predicted swelling percentages

of soil specimens beneath isolated footing model rested directly on expensive soil. The

prediction was done using stress change method (Rama et al., 1988), suction changes

method (Hamberg, 1985), Brackely et al., 1983 and Seed et al., 1962 empirical

equations, the results of comparison were summarized in Table (2.21).

Table (2.21): Comparison between measured and predicted footing heave percentage

(after Abdel-Moaty, 1999)

Sample No.

Footing Stress

σ (kg/cm2)

Final Measured Footing Heave

Percentage (∆h/H)fm %

Final Predicted Footing Heave Percentage (∆h/H)fp %

Rama et al., 1988

Hamberg, 1985

Brackely et al., 1983

Seed et al., 1962

C-1 0.5 22.38 36.2 57.4 21.25 47.1

0.0 23.38 55.5 57.4 21.82 47.1

C-2 0.5 21.50 48.6 78.6 29.1 89.0

0.0 25.00 98.2 78.6 29.9 89.0

C-3 0.5 21.30 26.5 64.9 15.3 52.5

0.0 21.40 54.9 64.9 15.8 52.5

The Comparison between measured and predicted heave of footing model test show

that:

a) The predicted footing heave using stress change method ranging from one to three

times the measured footing heave. For the soil have small clay content and small

swelling pressure values the stress changes method give the good agreement results.

b) The predicted footing heave using suction changes method ranging from two to

three times the measured footing heave.

c) The predicted footing heave using empirical heave equation suggest by Brackley et

al., 1983 ranging from 0.75 to 1.0 the measured footing heave


59

d) The predicted footing heave using empirical heave equation suggest by Seed et al.,

1962 ranging from two to four times the measured footing heave.

The comparison shows that the value of predicted heave of footing using equation

suggested by Brackley et al., 1983 is a good agreement with that measured from

laboratory footing model test .This means that the predicted footing heave is mainly

affected by the geotechnical of expansive clay deposits within the active depth.

2.4.2 Empirical Methods

These empirical expressions relate the swelling parameters to the geo technical

parameters that are determined by means of identification tests (Yilmaz, 2006).

• Seed, Woodward and Lundgren, 1962 established the following simplified

relationship;

S = 60 K (P.I) 2.44 (2.27)

Where;

S = swell potential,

K = 3.6 X 10-5 and is constant, and

P.I= plasticity index.

This equation is applied only to soils with clay contents between (8% -65%) and the

computed value is probably accurate to within about 33% of the laboratory determined

swell potential.

• Van der Merwe, 1964 developed a simple formula using potential expansiveness

and a factor to account for decreasing heave with depth. The potential expansiveness


60

of the soil is determined from Figure (2.27) based on the plasticity index and clay

content.

The expansive soil layer is divided into (n) layers. The total heave (p) is estimated by

the following equation;

∆ (2.28)

Where ,

D =Depth of soil layer in increments of 1 foot

Fi = reduction factor for layer (i) where ( F= 10 –D/20 ) , and

PEi = Potential expansiveness in inch / foot of depth.

The potential expansiveness is assumed as follows:

Very high PE=1 inch per. foot depth

High PE=l/2 inch per. foot depth

Medium PE=l/4 inch per. foot depth

Low PE=0 inch per. foot depth

The reduction factor curve was developed for an area in South Africa using double

oedometer test results. This method is used often in South Africa. However, it does not

consider initial soil conditions such as water content, suction or density. It should be

used only as an indicator of heave and not for quantitative predictions.


61

(a)

(b)

Figure (2.27): Relationships for determining (a) Plasticity index (P.I) and (b)

Reduction factor (P) for Van Der Merwe’s empirical heave prediction methods (After

Van Der Merwe, 1964).


62

• Schneider and Poor, 1974 developed statistical relationships for Texas clays

between measured swell for various surcharges, and the plasticity index and water

content. They presented the equations shown in Table (2.22) for prediction of

percent swell (Sp).

Log Sp = F ( P.I / Wc ) - C (2.29)

Where;

Sp = percent swell,

P.I = plasticity index,

Wc= initial water content, and

F , C = factors depend on the surcharge.

Table (2.22): Prediction of swell percent (After Schneider and Poor, 1974). Surcharge

(Kg/cm2) Log Sp

0.0 0.90(PI/W)-1.19

0.3 0.65 (PI/W)-0.93

0.5 0.51 (PI/W)-0.76

1.0 0.41 (PI/W)-0.69

2.0 0.33 (PI/W)-0.62

• Nayak and Christensen method, 1974 The method involves the development of

two statistical relationships, one for swell and the other for swelling pressure, in terms

of plasticity index, percent clay content, and initial moisture content. The developed

relationships are:

. . . (2.30)


63

where

SP= predicted swell percentage

PI = plasticity index, percent

C = clay content, percent

Wi = initial moisture content

and

. . . (2.31)

where

Pp = predicted swelling pressure, psi

The correlation with measured odometer data was very good. Here again, experience

with the method outside the area of its development is somewhat limited.

• Weston, 1980 presented a method of calculating swell based on the liquid limit

which can be determined more accurately than plastic index. This method is an

improvement on Van der Merwe’s method to take into account the moisture content.

Swell (%) = 0.00041(WLW)4.17(P)-0.386 (w i)-2.33 (2.32)

Where;

% 0.425 (2.33)

P = vertical pressure in kN/m2 (kPa), under which swell takes place

wi = initial moisture content (%)


64

• Pidgeon, 1987 produced a more friendly-user empirical relation for the

determination of free swell was proposed by Pidgeon (1987). The swelling potential is

calculated as follows:

%

(2.34)

where

FS = free swell (%)

P = pressure induced by the foundation and the overburden (kPa)

Ps = swelling pressure of the soil (kPa)

• Vijayvergiya and Ghazzaly, 1973 proposed a model gives the following

relationships:

log S = 1/12 ( 0.4LL − Wn +5.5) (2.35)

log S =1/19.5(6.242γd +0.65 LL−130.5) (2.36)

where

s = swell (%)

γd = the dry weight density in kN/m3

LL and Wn are liquid limit and the natural water content.

• Komornik and David method, 1969 suggested another statistical comparison of

measured data which provides a relationship for predicting swelling pressure using

liquid limit (LL), natural dry density (γd), and natural moisture content (Wi).


65

The relationship for predicted swelling pressure is:

. . . . (2.37)

where

The dry density in Kg/cm3 and swelling pressure are in kg/cm2.

• Bekkouche et al, 2001 presented a model which estimates the swelling magnitude

as a function of the load present in the soil. This model is written as follows:

. . . . . (2.38)

Thus, when the magnitude of swelling is nil (εs= 0), the free swelling pressure will be

given by

..

(2.39)

In the two relationships above, the plasticity index Ip and the natural water content wn

are expressed in percentages while the swelling pressure ps is expressed in kPa.

• Brackley et al., 1983 presented an empirical equation for the prediction of the

swell of expansive clay. It is derived from the results of laboratory swell tests upon

compacted samples, the equation takes account of the soil composition, initial

moisture content, density and the magnitude of the external pressure applied to the

soil.


66

The relative importance of these different factors is difficult to gauge for this the

equation is introduced to model their inter-relationship to give an understanding of

swell behaviour and provide an indication of the effect of changes in the variables; the

equation is presented as follows;

SP=(5.3-(147e/PI)-log10P)(0.525PI+4.1-0.85Wc) (2.40)

Where;

Sp = potential percentage swell,

e = original void ratio,

P = external pressure in kilopascals,

Wc = original moisture content in percent by mass, and

P.I = the plasticity index.

Brackley et al. (1983) developed the relationship between swell percent and original

moisture content for four sites; he found that the swollen moisture content (Ws)

could be expressed as a function of the plasticity index.

Ws = 0.62PI + 4.S (2.41)

And the swelling pressure could also be defined in terms of plasticity index,

log10 Ps = 5.3- (147e/PI) (2.42)

Predictions from equation (2.40) have been compared with data of Yevnining

and Zaslavsky, 1970 where the prediction values of swell were much lower than that

actually measured, but the relationships found by Yevnining and Zaslavsky, 1970

were qualitatively similar. The discrepancy is attributed to differences in test


67

methods. Thus the equation (2.40) will give an indication of the effects of changes in

the variables that cause swell, and may be used for swell prediction. However, it is

sensitive to small changes in the variables, and more tests results are required to

produce a more accurate form of this equation.

Many other empirical procedures exist which have not been discussed here, but they

could be summarized as shown in Table (2.23) [El-Sohby et al, 1995], which

classified the empirical methods to three categories; The first includes those methods

which correlate the swell percent and swelling pressure to some of the physical

properties of soil such as liquid limit, plasticity index, clay content, type of clay

minerals represented by clay activity. The second includes those methods which take

into consideration the above factors as well as the initial dry density as a descriptive

factor for the soil fabric (voids and solids) and soil structure. The third includes those

methods which take into account all the factors considered in the first and second

category in addition to the pressure applied during the soil deformation.

The empirical methods have a major disadvantage that they are based on a limit

amount of data and actually apply only in the region for which they were developed.

Caution should be exercised in their use. Their primary value is as an indicator of

expansion potential.

The shown empirical methods (models) are very useful to determine the swelling

properties of soils but cannot be precisely used for all types of clay. There is a need to

collect enough data in each area in order to develop specific models for specific type

of clay.

As with the direct and indirect techniques, no universally applicable technique has

been described for an accurate assessment of the potential volume change. However,

experiences within localized areas have indicated reasonably good results using many

of the techniques previously described.


68

Table (2.23): Correlations for swelling soils. (After El-Sohby et al, 1995).

Category Reference Correlation Factors

(1) Seed et al. (1962) Sp= 2.16* 10 ^-3(PI)2.44

= 4.13 * l0-4(Sl)2.67

= 3.6* 10-5(A)2.44*(C)3.44 Sp: Swell percent under 7kpa

PI = WL - Wp SI = WL - WS A = PI/C C: % <2u

Van Der Merewe (1964) Sp= 100/n ΣFHPE

PE to be determined from PI-C chart FH = 10H/20 and H: layer thickness (foot)

PI and C

Zacharias & Ranganathan (1972)

Sp = -225 + 290 (WL - W100) / SI + 1.2 SI/Sr

SI = WL - WS

W100 at Sr =100 % WL

Vija & Ghazzaly(1973) Log Sp = 1/12 (0.40 WL - Wo +5.5) Sp: swell percent under load=10 Kpa

WL , W0

Schneider & Poor (1974) Log Sp = 0.9 (PI/W) - 1.19 for no surcharge PI , W

Nayak & Chritensen (1974) Sp = 0.0229 (PI) (1.45 C/Wo) - 6.38 for load 7kpa log Ps = 0.0358 (PI)0.5 (C/W)2 + 3.7912 Ps : swelling pressure in t/m2

Pl , C , Wo

Dedier el al. (1973) Log Ps = 0.0294 C- 1.923 Ps in kg/cm2 C %

Vija & Ghazzaly(1973) Log Ps = 1/12 (0.40 WL - Wo - 0.4) Ps in kg/cm2 WL , W0

Zacharais & Ranganathan (1972)

Ps=(-225/6.4)+(229/6.4)(WL-W100)/SI+(1.2/6.4)(SI/Sr) WL , Wl00 , Sr

Popescu (1983) Ps = 0.5735 PI - 10.9196 PI

(II) Komorinik & David (1969) Log Ps = 2.132+0.0208WL-6.65* 104γd -0.0269Wo

Ps in kg/cm2, γd in kg/cm3 γd , WL ,W0

Dedier et al. (1973) Log Ps = 2.55 (γd/γw) - 1.705 γd

Vija&Ghazzaly (1973) LogSp= 1/19.5 (γd + 0.65 WL- 130.5) LogPs= 1/19.5 (γd + 0.65 WL- 139.5) Ps in t/ft2, γd = lb/ft3

γd , WL

El Sohby & Rabbaa ( 1981) Log Ps = 2.17 (γd + 8.4*10-3C - 1.8) for sand clay Log Ps = 2.5(γd + 6*10-3 C-1.6) for silt clay Ps in kg/cm2, γd in t/m3

γd , C

El Sohby & Mazen (1987) Log Ps = 2.17 (γd + 0.1 WL - 2.0) for sand-clay Log Ps = 2.5 (γd + 7* l0-3 WL - 1.83) for silt-clay

γd , WL

El Khoraibi et al. (1991) logP,=0.2258Ac[L(As-2)/Bs]1/2 + 3.1623L(γd)l/2 -0.034 L(Wc/Bs) + 2 where Ps in KN/m2, γd in KN/m3

Ac=PI/Cn% Gs , γd ,Wc% α=Wc%/γd L=αGs , As,Bs suction parameters

(III) Brackley et al. (1975) Sp = (5.3 - 2.77e-log10P)(32.4 - 0.85W) Log Ps = 5.3 - 2.77e where P and Ps in kpa

P , e , W e=(Gs.γw/γd)-1

Weston (1980) Sp = 4.11 *10-4 (WL) 4.17 (P) -0.386 (W)-2.33 P , e , W

Brackley et al. (1983) Sp = [5.3-(147 e/PI)- logP) (0.525PI - 0.85 W + 4.1) Log Ps = 5.3 - 147 (e/PI) where P and Ps in kpa

P , PI , W e=(Gs. γw/γd) -1


69

2.5 Treatment of expansive soils

Many treatment procedures are available for expansive soils (Chen, 1988; Nelson and

Miller, 1992). Removal of expansive soil and replacement with a non-expansive

material is a common method of reducing shrink-swell risk. If the expansive soil or

stratum is thin, then the entire layer can be removed. However, frequently the soil or

stratum extends to a depth too great to remove economically.

Pre-wetting a site to increase the moisture content can eliminate an expansive soil

problem if the high moisture content can be maintained. Soils with low hydraulic

conductivity may take years to pre-wet and conversely soils with high hydraulic

conductivity may never sufficiently wet. Most pre-wetted clay sub-grades are injected

with water to a depth that will nearly eliminate sub-grade swell potential. After this

process, they are either covered with a polyethylene sheet retain moisture, kept wet

by sprinkling, or built on fairly quickly. Vertical or horizontal moisture-loss

barriers have been installed in only a few cases in the past, but their use is becoming

more frequent. The combined effect of a moistened sub-grade and moisture barriers,

when applied correctly, can be very successful. The application of moisture treatment

coupled with proper degrees of compaction has been utilized extensively (Petry and

Little, 2002).

Stabilization of clay sub-grades is a popular alternative for geotechnical engineers

considering the economics of construction with expansive clay soils. They must

choose among traditional stabilizers and new, nontraditional stabilizers that are not

backed by as much field experience or independent research results. Two construction

factors have strongly affected the quality of chemical stabilization: moisture and

degree of pulverization. First, sub-grades are rarely pre-wetted but are brought to the

required water content during construction; often they are worked dry of optimum,

resulting in incomplete chemical reactions due to lack of water. Second, the degree of

pulverization is often substandard. Lime must react intimately with particles to enact

the pozzolanic process; the larger the soil particles, the longer that process will take.

A summary of widely used treatment techniques is shown in the following flow chart.


70


71

2.5.1 Miscellaneous Treatments

Methods categorized as miscellaneous in this report include replacement fill,

remolding and compaction, pre-wetting, and sub-drainage.

2.5.1.1 Replacement Fill

Removal of natural expansive sub-grade material and replacement with a

nonexpansive material is a most obvious method of eliminating swell problems. In

some cases this approach may be economical if the expansive stratum is thin and

replacement materials are available. Unfortunately, this is generally not the case, and

the excavation and replacement solution is extended only to a depth which will reduce

swelling to a tolerable minimum. Hence the required depth of excavation depends

upon the expansiveness of the soil and the anticipated weight of backfill and structure

which counteract the uplift forces of the swelling soil. The selection of the particular

non expansive backfill material is critical. Replacement materials have included non-

expansive borrow as well as expansive soils treated with lime or other swell-

prevention chemicals.

Abdullah, 2002 found that the direct method (Texas Highway Department Method

TEX-124-E) and Double oedometer Method overestimate the footing heave. The

reason for such over prediction is that the swell condition under a finite loaded area is

not one-dimensional (manifested in the vertical direction) as assumed by these

methods. Soil expansion under the influence of finite loaded areas is multidimensional

where vertical and lateral swells take place simultaneously. The moisture content and

the amount of applied footing pressure are two important factors, which influence the

value of the induced heave. Awad, 2005 and Abdel-Hady, 2007 in arecent study

explained that the swelling potential of untreated expansive soil is not affected by the

size of tested soil sample

Oedometer results:

Numerous works have been done to study the swelling properties of treated expansive

soil in the oedometer using sand replacement (e.g. Satyanaryana, 1969; Moussa et al.,


72

1985). Marie et al. (2000) studied the improvement of expansive soil using three types

of cushions: sand+5% lime cushion, sand cushion, and clayey silt cushion, for clay

content of 20%. They concluded that: i) sand+5% lime cushion proved to be the best

cushion to achieve a higher reduction in the swelling potential and swelling pressure,

ii) sand cushion or clayey silt cushion causes almost the same reduction in swelling

potential and swelling pressure.

Laboratory modeling results:

The behavior of footing resting on treated expansive soil using sand replacement has

been investigated (Abouleid and Reyad, 1985, Awad 2005, Awad and Abdel-Hady,

2005, Abdel-Hady, 2007).

Abouleid and Reyad, 1985 carried out a series of tests using laboratory model for a

shallow footing. The heave of footing for different parameters was measured as a

function of time due to wetting. According to this study, the following was concluded:

i. Providing sand cushions beneath isolated footing is considered avery effective

arrangement to minimize the foundation displacement tp a reasonable and

practical value that the structure could colerate.

ii. Increasing the thickness of sand cushions beneath footing reduces the maximum

vertical rise of foundation to acertain small value beyond which any increase in

the cushion height results in avery slight and rather negligible reduction in

foundation movement.

iii. The effective height of sand depends of footing breadth and the thickness of the

clay layer under the foundation.

iv. The effective value of cushion height was found to be 1/3 of the foundation

breadth (B).

v. Providing sand cushion with breadth equal to double the footing width reduces

the expected maximum vertical rise of the foundation by 95% of its value in

case of no cushion used and by 77% of its value in case of both the cushion and

footing have of equal breadths.


73

vi. Applying surcharge pressure to the clay surface around the footing area

resulting ahigher up raising pressures on the footing and prevents further

settlement.

Although the utilization of sand replacement below shallow isolated footings is

considered, as an adequate technique for reducing both swelling and swelling pressure

of expansive soils, but Ahmed, 2000 investigated that its capability tends to be less for

raft foundations than for isolated footings, after discussing the field monitoring results

of this building of isolated reinforced concrete footings resting on plain concrete raft

and cushioned with sand. He indicated that the utilized foundation arrangement leads

to negligible reduction in the swelling pressure in the range of 5.78 to 6.67 %, while

the reduction in swelling percent is ranging between 10.4 to 20%, these results

attributed why the building does not reach the stability condition as expected with the

assignment of sand cushion. Accordingly Ahmed, 2000 confirmed that the validity of

sand cushions in damping swelling depends on the presence of unconfined free zones

between footings permitting the upward movement of soil outside the boundaries of

shear zones. Therefore, the use of cushioned raft foundations without some foundation

details or special technique to allow the upward movement of soil will be insufficient.

Awad and Abdel-Hady , 2005 measured the final heave of laboratory model footings

models resting on untreated and treated expansive soils using sand-lime cushion (soil

clay content of 35%) and concluded that: i) the dry density of medium sand-lime

cushion was maximum at lime content 10%, ii) the optimum water content increased

with the increase of lime ratio, iii) the measured final heave of the footings and ground

surface decreased as lime ratio of the cushion increased to a certain value (lime ratio

of 10%) then, it increased as lime ratio increased, iv) the relative movement between

the footings was approximately equal to zero (lime ratio ≤ 10%), then it increased as

lime ratio of the cushion increased, v) the measured final footing heave (sand-30 %

lime) was approximately 1.5 times that of the untreated expansive soil.


74

Comparison between oedometer and laboratory modeling results:

From the abovementioned mentioned literature review, it is clear that the

investigations in the Egyptian practice are limited to one type of Egyptian expansive

clay whose clay content is less than or equal to 50%, using oedometer tests.

Awad, 2005 in a present work carried out more investigations using soil specimens

with different clay contents treated with different lime contents. He used both

oedometer and one dimension laboratory model tests (Figure 28) to study the effect of

sample size on the swelling properties of the tested soils and to investigate the effect

of sand-lime cushion on suppressing the heave of expansive soil. He also determined

the swelling potential and swelling pressure also using direct measurement tests for

treated and untreated expansive soils. Table (2.24) shows the soil properties used in

test. Some of the results are shown in Figures (2.29, 2.30 and 2.31)

Table (2.24): Properties of untreated expansive soils (after Awad, 2005)

Property Soil A Soil B Soil C Classification

Clay Content (Cn) % 100 75 50 Very highly expansive (Holtz, 1959)

Liquid Limit (wL.L) %

136.3 98.8 91.8 Montmorillonite mineral (mg)

(Chen, 1988)

Plastic Limit (wP.L) % 33.9 26.4 19.7 Critical expansive soil

(Altmeyer, 1955) Shrinkage Limit

(wS.L) % 9.8 9.6 9.4

Critical expansive soil (Altmeyer, 1955)

Plasticity Index (P.I) %

102.4 72.4 72.1 Very high expansive (Holtz and

Gibbs,1956)

Activity 1.024 0.96 1.44 -

Swelling Pressure (kg/cm2) *

8 7 1.9 -

Swelling potential (%)*

23 19 7 -

* Jennings et al., 1973


75

Figure (2.28) Laboratory model test set up (after Awad, 2005).

Figure (2.29): Swelling- log stress of expansive soil model and oedometer ( clay

content 100%) (after Awad, 2005).

Loading plate

Dial gauge

Screw pin

Mould

Soil specimen

Mould base

Porous plate

Mould seat

Weights

Water tank

Main lever

Sand cushion


76

Figure (2.30): Swelling potential of treated soil-lime content in oedometer and model

(after Awad, 2005).

Figure (2.31): Swelling potential of treated soil-lime content in oedometer and model

(after Awad, 2005).

Based on the experimental study, it was found that:

(a) The swelling properties of treated soil are the same for sand cushion and sand

mixed with 5 % lime, while the sand+5% lime replacement improved the compaction

degree of sand replacement layer.


77

(b) For clay contents of 50 and 75%, the swelling potential of treated expansive soil is

approximately constant with lime content. While in case of clay content 100%, the

swelling potential is approximately constant up to 5% lime content, then it increases

as lime content increases to a certain value (lime content=10%), then it becomes

approximately constant (lime content > 10%).

(c) The swelling pressure of treated soil is approximately constant at lime content ≤

5%, then it increases as lime content increases up to a lime content of 10% in

oedometer tests, and 20 % for laboratory model tests, then it is approximately constant

with lime content.

Awad and Abdel-Hady, 2005 investigated the final heave of the laboratory footings

models resting on untreated and treated expansive soil using sand-lime cushion. The

moisture content distribution under the footing and ground surface due to leakage of

the water source were also determined. The parametric study of this research took into

consideration the effect of water source distance from the footings and lime content of

the cushion on the heave-time and moisture content profile relationships. The lime

content of the cushion was ranging from 0 to 30%. Two identical square footing (FA

and FB) were investigated for untreated and treated soils.

The footing width (B) was 4cm and the footing stress was 23kN/m2.The water feeding

source distance from the footings was varying from 0.5 to 10.5 footing width (2 to 42

cm) and was entering at the bottom of the specimen. The thickness of sand cushion

was kept constant at half footing width. A very high swelling potential expansive soil

was used to evaluate the effectiveness of sand-lime cushions.

Two analytical heave equations were proposed by Jenning et al., 1973 and Hamberg,

1985 for quantitative assessment of the effect of sand-lime replacement on damping

the heave of expansive soil are applied to the case study. Finally, the predicted heave

results using these equations were compared with that measured in laboratory tests.


78

Figure (2.32): Laboratory model test set up (after Awad and Abdel-Hady, 2005).

SB= 32 and 42 cm

SA = 2 and 5 cm

50 cm

25 cm FAFB

2cm

2.5 B

(G. S.)A (G. S.)B

B = 4.0cm

LoadLoad

Water Inlet

Dial gauge

Sand cushion

Vertical sand filter

Square footing

BxB Vent FAFB


79

Based on their experimental and prediction results the following conclusions were

obtained:

1- The measured final heave of the footings and ground surface decreases with the

increasing of water source distance in case of untreated soil.

2- For untreated soil, the final moisture content values increases as the water source

distance from the footing decreases. The final moisture content at water source inlet is

less than the plastic limit and it is greater than the shrinkage limit at the bottom of

specimen beside the water source. No wetting gradients observed at water source

distance equal to eight times footing width.

3- The measured final heave of the footings and ground surface decrease as lime

content of the cushion increases up to lime content =10% then, it increases.

4- For treated soil, the relative movement between the footings is a nearly equal to

zero for lime content ≤ 10% then; it increases as lime content of the cushion increases.

5- Treatment of swelling soils using sand-lime cushion should be limited to lime

content up to 10%.

6- The distribution of moisture content through of the treated expansive soil is not

affected by lime content of the cushion and footing stress for lime content ≤ 10%.

7- The predicted footing heave according to Hamberg, 1985 increases as the water

source distance from the footing and the footing stress decrease. While, the predicted

footing heave according to Jenning et al., 1973 is not affected by the water source

distance and it increases as the footing stress decreases for the untreated and treated

expansive soil.


80

8- A comparison between the predicted and the measured results revealed that the two

analytical equations proposed by Jenning et al., 1973 and Hamberg, 1985 are the

suitable for the heave prediction of the treated soil. However, the effect of distance

between the footing and water source can be taken into consideration in the first

method.

Abdel-Hady, 2007 in a recent study measured the properties of treated expansive soil

using sand-lime cushion and clay-lime cushion. The parametric study of the research

took into consideration the effect of lime content of the sand cushion, size of sand

cushion and the lime content which mixed with expansive soils.

An experimental testing program has been performed using oedometer and large-scale

one dimension laboratory model tests. Three types of expansive soil, medium and fine

sand were used. The hydrate lime content of sand cushion is ranging from 0 to 100%.

The lime content which mixed with expansive soil is ranging from 5 to 20 %. The

results are shown in Table (2.25).

From the obtained results, the following conclusions can be summarized:

i. The effect of increasing lime content on maximum dry density of sand-lime

cushion is more considerable for medium sand than fine sand.

ii. The maximum dry density of compacted sand-lime cushion increases with

increasing lime content and the optimum moisture content decreases with it.

iii. The improvement of the fine sand compaction is approximately constant up to

lime content ≤ 15 %.

iv. The swelling potential is not affected by the size of sand cushion at lime

content = 5%.

v. Increasing of lime content on sand-lime cushion decreases the swelling pressure

and it’s a pronounced effect at 5 and 10 % lime content for oedometer tests.


81

Table (2.25): Properties of untreated and treated expansive soil using sand-lime

cushion (after Abdel-Hady, 2007).

Property of soil

Type of test Type of soil

Sand cushion

type

Lime content of sand cushion (%)

- 0 5 10 15 20 25 100

Swelling potential (Sp) %

Oedometer

A

Untreated 23.86

Medium sand - 15.86 12.86 12.86 13.73

15.35 14.02

Fine sand - 12.45 12.67 17.52 16.97

16.5

B Untreated 23.24

Medium sand 14.67 14.5 15.62 13.42

15.6 12.2

C Untreated 19.13

Medium sand 14.52 12.83 14.94 12.33

11.24 14.6

Laboratory model

A Untreated 24

Medium sand 13.55 13.22 14.4

12 12.42

Swelling pressure σS

(kg/cm2)

Oedometer

A

Untreated 13.5

Medium sand - 10.9 9.7 7.65 9

7 9

Fine sand - 7.3 4.2 12 10

11.5

B Untreated 20

Medium sand 17 17 15 15.8

22.9 15

C Untreated 15

Medium sand 12.6 8 15.8 8.4

8 9

Laboratory model

A Untreated 63.82

Medium sand 49 44 54.8

53 30


82

Replacement thickness prediction:

According to (Chen, 1988), the Federal Highway Administration has recommended

that the required thickness of the, replacement material can be estimated from the

expansive soil’s plasticity using the relationships of Table (2.26).

Table (2.26): Relation of required surcharge fill thickness to soil plasticity (Chen,

1988)

Plasticity Index, percent

Thickness of Undercut and Replacement Fill, feet

Interstate Highways Secondary and State Highways

10-20 2 2 20-30 3 2

30-40 4 3 40-50 5 3 >50 6 4

Ahmed, 1985 concluded that the providing isolated footing with sand cushion as a

good practice in order to minimize the swelling pressure of under lying expansive soil

by more than 50 %. Presence of sand layer under the expansive soil deposit results a

reduction in the swelling pressure by 35% or more. The optimum height of sand

cushion is 1/5 of the footing breadth.

Abdel-Moaty, 1999 determined the swelling deposits at three new communities

around the greater Cairo; El-Obour, El-Mokattam and El-Qattamiya, as representative

examples for the geological formation and geotechnical properties of this region. The

three new cities were considered the northeast extension of the new communities

around the greater Cairo. The expansive soils at the selected sites in the three cities

were classified as high to very high swelling clays. The ground surface heave was

predicted analytically using different approaches; semi-empirical methods as

suggested by two different procedures (Stress Changes and Suction Changes) as well

as the empirical methods employing heave equations.


83

Effects of the magnitude of active depths and the thickness of non expansive soil

overlying the expansive clay layer on the expected heave values were investigated in a

parametric study for the three new cities. The methods were also used to evaluate the

contribution of soil replacement to damping of heave movements.

The predicted heave values using these methods were compared with that measured in

large-scale laboratory tests. The comparison revealed that the empirical heave method,

suggested by Brackely et al., 1983 was the best method for heave prediction. The

results also showed that optimum treatment of expansive deposits using soil

replacement with adequately compacted sand can be achieved when the thickness of

compacted sand was about 40% of the active depth.

El-Nahhas et al., 1998 studied the effects of the magnitude of active depth and the

thickness of soil replacement on the expected heave using stress change method

(Rama et al., 1988). Figure (2.33) shows the predicted ground surface heave versus

thickness of replacement using Stress Change method. Figure (2.34) shows the

reduction of heave versus thickness of replacement using Stress Change Method; the

results are summarized in Table (2.27).

These results show if 40% of maximum active thickness of expansive layer at the sites

B and C is removed and replaced with compacted non-expansive soil, the maximum

movement would be only 54% of that anticipated with no soil replacement.

Therefore, analytical results confirm quantitatively that the replacement of the top

portion of expansive clay layer can significally reduce the potential heave that could

occur during lifetime of the structure. The reduction of ground surface heave increases

as the thickness of replacement increases up to the maximum value at the active depth

or full replacement of the expansive clay layer which ever smaller.


84

Table(2.27): Predicted ground surface heave for treated and untreated expansive soil

using soil replacement.

Thickness of replacement (meters)

Site A (H=1.80m) Site B (H=6.00m) Site C (H=6.00m)

0 93 299.5 299.5

0.6 72.7 263.9 263.9

1.2 34.8 229.7 229.7

1.8 0 195 195

2.4 0 160.3 160.3

3.6 0 106.9 106.9

4.8 0 51.8 51.8

6 0 0 0

6.5 0 0 0

Figure (2.33): Predicted ground surface heave versus thickness of replacement using

Stress Change method

0

50

100

150

200

250

300

350

0 1 2 3 4 5 6 7

Groun

d Surface he

ave (cm)

Thickness of replacement (m)

Site A (H=1.80m(

Site B (H=6.00m(

Site C (H=6.00m(


85

Figure (2.34) :Reduction of heave versus thickness of replacement using Stress

Change Method

Awad, 1999 summarized the reduction of final footing heave in case of treated

expansive soil using soil replacement in Table (2.28 )

Table (2.28 ): reduction of final footing heave in case of treated expansive soil using

soil replacement:

Refrences Soil Replacement thickness Footing heave reduction =

∆ . ∆∆ .

Abouleid and Reyad, 1985 0.38 B 31 %

El-Sabaie, 1992 0.50 B 12.5 %

Awad, 1993 0.50 B 30 - 24 %

Awad, 1993 1.0 B 52 - 40 %

Awad, 1993 0.50 Ha 66 %

El-Kadi and Awad,1998 0.35 Ha 73 %

El-Kadi and Awad,1998 0.50 Ha 92 %

El-Nahhas et al., 1998 0.40 Ha 54 %

0

20

40

60

80

100

120

0 1 2 3 4 5 6 7

Redu

ction of heave %

Thickness of replacement (m)

Site A (H=1.80m)

Site B (H=6.00m)

Site C (H=6.00m)


86

2.5.1.2 Remolding and compaction

Compacting sub-grade soils to low dry unit weights on the wet side of optimum water

content using kneading methods has been found to be an effective means to reduce

swelling for moderately plastic soils (Seed, Mitchell and Chan, 1960). Holtz, 1959

shows that compaction at low densities, and at water contents above the optimum

water content, as determined by a Standard Proctor test produces less expansion

potential than compaction at high densities and low water contents.

Holtz and Gibbs, 1956 illustrate that an increase in molding water content for a given

density decreased the swell and swell pressure. However, an increase in density at any

given water content may increase or decrease the swell, depending on the range of

densities involved (generally, an increase in density causes an increase in swell).

Hence low densities and high water contents are conducive to smaller expansion.

It was observed that soils compacted dry of optimum exhibit higher swelling

characteristics and swell to higher water contents than do samples at the same density

compacted wet of optimum. The method of compaction also influences swelling

characteristics of compacted swelling soils. An expansive soil with a dispersed,

(deflocculated) structure swells less than one with a flocculated structure for the same

water content and density.

Compaction of clays under conditions that produce high shear strains and locally

oriented clay particle arrangements has been demonstrated to result in less water

deficiency and therefore, reduced swelling. The limitation on this approach is that the

lowered dry unit weights and higher water contents are also accompanied by lower

sub-grade strength and stiffness and an increased tendency for soil shrinkage to occur

if the soils are subjected to drying (Snethen et al., 1975).


87

2.5.1.3 Pre-Wetting

Pre-wetting is based on the theory that increasing the moisture content in the

expansive foundation soils will cause heave to occur prior to construction and thereby

eliminate problems afterward. It is assumed that if the high moisture content is

maintained, there will be no appreciable increase in soil volume to damage the

structure. This procedure may have serious drawbacks that limit its application

(Nelson and Miller, 1992).

Experience with the techniques of pre-wetting, sprinkling or injecting large amounts

of water into natural sub-grade soils prior to pavement construction for the purpose of

reducing post-construction swelling has been reviewed by Chen, 1988. Pre-wetting

programs are designed to provide water to satisfy the clay’s water deficiency.

However, the reductions in soil swelling are usually accompanied by weakening,

softening and decreases in the workability of the soil. Depending on the permeability

of the treated soil, the pre-wetting process may require months or even years to be

completed.

2.5.1.4 Sub-Drainage

The use of sub-surface drain tile or perforated pipe wrapped in fabrics or graded

granular filters below the edges of pavements (Figure 2.35) has also been reviewed by

Chen, 1988 and Nelson and Miller, 1992. Properly designed and constructed side

drains parallel to the pavement alignment can be effective in intercepting and rapidly

removing subsurface water flowing at atmospheric pressures through materials of low

suction potential. Sub-drains will not have any effect on water moving into expansive

soils in the vapor phase or water moving in unsaturated soil in response to gradients in

soil suction.


88

In fact, subsurface drains may enhance the opportunity for atmospheric water to enter

the sub-grade by exposing more surface area of expansive high suction sub-grade soils

to water at atmospheric pressure. Good surface drainage and the prevention of pre-

wetting is more likely to be effective than sub-surface drains placed in highly

impermeable, highly plastic soils (Hardcastle, 2003).

(a)

(b)

Figure (2.35): embedded drains around structures


89

2.5.2 Hydraulic Barriers

Since the change in moisture content is the main factor influencing the volume change

of swelling soils, it is obvious that if the soil could be isolated from any moisture

changes, volume change could be reduced or minimized.

The basic principle on which hydraulic barriers act is to move edge effects away from

the foundation or pavement and minimize seasonal fluctuations of water content

directly below the structure. Also, Barriers may be designed to completely encapsulate

the expansive soil or they may be intended to act as cutoffs for either horizontal or

vertical flows so that the time during which moisture changes occur is longer. This

allows for water content to be more uniformly distributed due to capillary action in the

subsoil. Thus the heave will occur more slowly and in a more uniform fashion (Nelson

and Miller, 1992). Moisture may still increase beneath or within areas surrounded by

the moisture barriers leading to steady but uniform heave of the foundation or slab-on-

grid.

2.5.2.1 Horizontal moisture barriers

This treatment provides for the complete isolation of expansive soils from all water

sources. Guidelines for successful applications of horizontal barriers have been

presented by Woodward-Clyde-Sherard, 1968 and Hammitt and Ahlvin, 1973.

Horizontal barriers have been constructed using both membranes and rigid barriers.

a) Horizontal Membranes

Single waterproof layers of asphalts, Polyethylene ranging from 4to20 mil in

thickness, polyvinyl chloride (PVC), polypropylene, high-density polypropylene, geo-

membranes, and other types of nonwoven fabrics have been used with varying degrees

of success. They have been installed both above and below compacted and natural

sub-grades for the purpose of intercepting liquid water and preventing its entry into


90

expansive soils. Horizontal barriers installed above the expansive soil layer are

designed to prevent water present in base courses from being drawn into the expansive

soil, which has a much higher affinity for water (suction) than typical aggregate base

materials.

Membranes placed below the potentially expansive sub-grade during new construction

or rehabilitation are designed to prevent water from being transported by capillary rise

into the sub-grade. The horizontal barrier approach is obviously not available for

remediation of existing expansive soil sub-grades. Furthermore, when membranes are

placed on top of compacted or natural sub-grades in semiarid climates, it’s virtually

impossible to prevent an increase in the water content of the potentially expansive sub-

grades. This situation is due to the fact that water in its vapor phase will continue to

move into the potentially expansive soil as a result of changes in temperature and

evapo-transpiration regime brought about by the membrane and the overlying

pavement.

Research at Fort Collins, Colorado, has shown that while horizontal membranes can

be effective in reducing the entry of surface water, their major benefits in arid and

semiarid climates are to increase the time required for water content increases and

swelling to occur. They also tend to make the water content increase and thus the

swelling more uniform when it does occur (Nelson and Miller, 1992).

Other applications include the use of horizontal moisture barriers around the perimeter

of structures to reduce lateral variations in moisture changes and differential heave in

the foundation soil. Plastic or other thin membranes around the perimeter should be

protected from the environment by a 6- to 12-inchthick layer of earth (Tm–Army,

1983).

A disadvantage of these barriers is that they are not necessarily reliable and may be

detrimental in some cases. For example, most fabrics and plastic membranes tend to

deteriorate with time. Undetected (and hence unrepaired) punctures that allow water to


91

get in, but not to get out, commonly occur in handling on placement. Punctures may

also occur during planting of vegetation. If the barrier is a concrete slab, the concrete

may act as a wick and pull water out of the soil, Figure (2.36) illustrates a useful

application of horizontal membranes (Tm–Army, 1983).

Figure (2.36): Application of a horizontal membrane (Tm–Army, 1983).

b) Rigid Barriers

Concrete aprons or sidewalks are commonly used as horizontal moisture barriers

around building foundations. Concrete, coupled with flexible membrane or asphalt,

has been used successfully as a remedial measure to establish constant uniform

moisture contents within foundation soils (Mohan and Rao, 1965; Lee and

Kocherhans, 1973; Najder and Werno, 1973).The design and function of joints and

seals are important when using a rigid horizontal barrier. Horizontal barriers will be

more efficient if surface drainage is provided so that bonding is prevented.

2.5.2.2 Deep Vertical Moisture Barriers (DVMB)

Deep moisture barriers have been used in airports, highways, railroads, pipelines,

canals, and sidewalks and at the perimeter of buildings. The deep barriers solution can

be used in deep highly expansive soils when seasonal moisture variations and

differential movements are exclusively caused by rainfall; a typical DVMB is shown

in Figure 2.37 (Snethen, 1979).


92

Installing the hydraulic barriers in a vertical orientation at the margins of the pavement

permits their use both in new construction and as a remedial technique for existing

pavements. DVMB’s have been constructed mainly with Portland cement concrete and

geo-membranes. More recently, geo-composites and highly plastic clays have been

used. For the reasons mentioned above for horizontal membranes, DVMB’s have been

shown to be effective in reducing but not entirely eliminating soil expansion beneath

pavements. DVMB’s also delay the swelling and tend to make it more uniform

(Nelson and Miller, 1992; Steinberg, 2000).

The vertical barrier (Fig. 2.38) should extend to the depth of the active zone and

should be placed a minimum of 3 feet from the foundation to simplify construction

and to avoid disturbance of the foundation soil. The barrier may not be practical in

prevent ing migration of moisture beneath the bottom edge for active zones deeper

than 8 to 10 feet. The granular barrier may also help reduce moisture changes during

droughts by providing a reservoir of moisture. The placement of a filter fabric around

the trench to keep fine particles from entering the perforated pipe will permit use of an

open coarse aggregate instead of a graded granular filter. In some cases, the perforated

pipe could be eliminated from the drain trench.

In another experience with a barrier placed in the perimeter of foundation slabs in a

residential area in Dallas, the heave inside the slab ranged from 5 to 8 cm and

stabilized in 1 to 2 years time (O’Neill and Ghazzaly, 1977). Movements of similar

unprotected foundations were somewhat higher but far more variable. Finally, in

another experience in the Valley View Airport in Alberta, Canada, an impervious

membrane was placed at both sides of the runway, 3 m deep, to prevent lateral

migration of soil moisture. The pavement showed improved performance in the short

term (3 years); however, after 5 years operation, the pavement distress was quite

severe showing bumps varying from 11 to 29 mm with cracks ranging from slight to

moderate (Diyaljee and Wiens, 1995).


93

Approximately more than half of the economic losses due to expansive soils come

from deterioration of transportation facilities: Much of this damage may be attributed

to the broad use of empirical methods incapable of correctly defining total and

differential movements, expansive pressures on soil-structure interfaces and the rate of

movement of the soil under different wetting-drying conditions.

Figure (2.37): Deep vertical moisture barrier, DVMB (Snethen, 1979)

Figure (2.38): Vertical and horizontal moisture barriers (Tm–Army, 1983).


94

Rojas et al., 2006 attempts to contribute to the development of more reliable analytical

methods to design civil infrastructure on expansive soils. Accordingly, a constitutive

model for expansive soils based on laboratory experimental observations was

proposed. The resulting numerical tool had the capacity of computing soil suction and

volumetric strain changes of expansive soils under a defined wetting–drying regime.

To verify the capabilities of this computer code, a laboratory barrier model was built.

The model was instrumented to measure soil suction changes and the corresponding

surface displacements. The experimental and theoretical results were compared.

Finally, the numerical model was applied to a design example of a deep moisture

barrier.

Figure (2.39): Lateral view of laboratory barrier model (after Rojas et al., 2006).

The intent of this model was to study the effectiveness of vertical barriers to prevent

soil volume change on the dry side of the barrier upon wetting. The model consists of

a rigid rectangular acrylic box with a slot in its center where a vertical barrier may be

inserted to different depths. The soil was placed by compaction into the box, at either

side of the vertical barrier Figure (2.39). A small flow rate was applied on the upper

left side of the barrier and vertical displacements were measured on the upper right

side. Thus, the left side represents the wetting source and the right side is the soil to be


95

protected by the moisture barrier. Soil suctions were measured at six ports located at

both sides of the moisture barrier, as shown in Figure (2.39), and then a constitutive

model for expansive soils has been coupled with the flow equation in a deforming

media resulting in a finite-element code able to simulate the flow of water under

different boundary conditions, Figure (2.40).

Figure (2.40): Finite-element grid for vertical moisture barrier model (after Rojas et

al., 2006).

Figures (2.41 a–d and f) show the isovalue–suction curves _in meters of water (for an

8 cm) depth barrier for Days 2, 5, 10, 12, and 15, respectively. The assembly of curves

observed in these figures represents the wetting front. It can be seen that this front

advances practically uniformly on the left side.

However, when it reaches the tip of the barrier, the wetting front gets distorted and its

pattern is modified significantly as it advances upward at the other side of the barrier.

This inclined wetting front generates distortions to the finite element grid on this side,

owing to the differential expansion of the nodes.

Swelling p

Figure (

Figure (2

at the cen

surface e

displacem

small met

erformance

(2.41): Iso

.42) show

nter of the

expansion

ment transd

tal plate w

e of some e

ovalue–suc

ws the theo

right side

on the

ducer (LV

was set dow

expansive s

ction curv

oretical an

e of the m

laborator

VDT) plac

wn.

soil treatme

96

ves in mete

nd experim

odel for th

ry model

ced at the

ent techniq

ers of wate

mental res

he three b

l was re

e center o

ques

er (after R

sults of the

arrier dep

ecorded w

f the mod

Rojas et al

e surface

pths consid

with the

del surfac

l., 2006).

expansion

dered. The

aid of a

e where a

n

e

a

a


97

Figure (2. 42): Theoretical and experimental result comparisons for surface heave

and different barrier depths

Accordingly he concluded that the finite-element code developed was a useful tool for

the design of deep moisture barriers. It is also helpful for the analysis of long-term

behavior of foundations or pavements built over expansive soils. Furthermore, it can

be expanded to account properly for the soil–structure interaction phenomenon.

Finally moisture barriers have been used both as a preconstruction technique and as a

remedial measure. Methods of construction are important in the success of moisture

barriers. Care should be taken to seal joints, scams, rips, or holes in the barrier. Deep-

rooted plants should be planted well beyond the perimeter of the moisture barrier.


98

2.5.3 Chemical Soil Treatments

Portland cement, lime, asphalt, calcium chloride, sodium chloride, and paper mill

wastes are common chemical stabilization agents. The effectiveness of these additives

depends on the soil conditions, stabilizer properties, and type of construction (i.e.,

houses, roads, etc.). The selection of a particular additive depends on costs, benefits,

availability, and practicality of its application.

Chemical stabilization includes the mixing or injecting of chemical substances into the

soil, these chemicals effectively reduce the affinity of clay minerals for water or

change the clay minerals into non-expanding lattice materials. The latter effect is

achieved by either destroying the clay minerals through what has been termed

artificial weathering or by reactions between additive and clay that produce

cementations materials at the expense of the clay. Changes in the chemical

environment brought about by chemical additives also may make the clay require less

water to satisfy its charge deficiency through the phenomenon of ion crowding, that is,

by replacing higher volume hydrated monovalent cations with divalent cations and

increasing the concentration of cations in the soil water.

A variety of stabilizers are available. These may be divided into three groups

according to their use: traditional stabilizers (hydrated lime, portland cement, and fly

ash); byproduct stabilizers (cement kiln dust, lime kiln dust, and other forms of

byproduct lime); and nontraditional stabilizers (sulfonated oils, potassium compounds,

ammonium chloride, enzymes, polymers, and so on). The traditional stabilizers

generally rely on calcium exchange and pozzolanic reactions to effect stabilization.

This is also the primary mechanism for stabilization with many of the byproducts. The

nontraditional stabilizers are identified as such because they rely on a different

stabilization mechanism. For example, sulfonated oils typically rely on hydrogen ion

penetration into the clay lattice. Hydrogen alters the clay structure primarily by

reducing its water-holding capability (Petry and Armstrong, 2001).


99

Stabilizers also can be divided according to the presence of calcium ions in their

chemical composition in to calcium based and non calcium based stabilizers.

2.5.3.1 Calcium-Based Stabilizers

Portland cement and the various forms of lime are examples of the calcium-based

materials used to modify and stabilize clayey soils. Soil modification with lime refers

to the immediate reduction in plasticity and water content that results from the

flocculation and agglomeration of the clay particles produced by the depression of the

adsorbed water films of the clay. Adsorbed water film thicknesses are reduced both as

a result of Ca+2 cations replacing monovalent cations (Na+ and K+) and the increase in

the total electrolyte concentration of the soil water. Both these phenomena result in a

decrease in the soil suction or water deficiency and the swelling potential. Soil

stabilization with lime is due to pozzolanic reactions in which new crystalline

cementitious materials are produced as the added lime raises the soil pH and thus the

solubility of silica and alumina (Transportation Research Board, 1987).

The principal materials used for the cementations stabilization and modification of

highway pavement materials are lime, fly ash, and Portland cement. Whereas lime

and Portland cement are manufactured products, fly ash is a by-product of the burning

of coal at electric power generating stations. As a consequence, fly ash generally

exhibits greater variability than is seen in the other products. By-products such as kiln

dust and fluidized bed ash from various manufacturing and energy generating

processes are used to a lesser extent.

a) Lime stabilization

It is an age-old practice to use lime in one form or the other to improve the

engineering behavior of clayey soils. Because of the proven success of lime

stabilization in the field of highways and air-field pavements, this technique is now

being extended for deep in-situ treatment of clayey soils to improve their strength and


100

reduce compressibility. The improvements in the properties of soil are attributed to the

soil-lime reactions (Yesilbas, 2004). Lime stabilization is covered extensively in the

literature (Rogers and Glendinning, 2000).

Lime or calcium hydroxide is the most widely used chemical stabilizer for clay soil

sub-grades. Lime stabilization is particularly important in road construction for

modifying sub-grade soils, sub-base materials, and base materials. The improved

engineering characteristics of lime-treated materials provide important benefits to both

portland cement concrete (rigid) and asphalt (flexible) pavements. Lime will primarily

react with medium, moderately fine, and fine-grained soils to create a number of

important engineering properties in soils, including improved strength; improved

resistance to fracture, fatigue, and permanent deformation (decrease elasticity),

increase workability; improved resilient properties; reduced swelling; and resistance to

the damaging effects of moisture. The most substantial improvements in these

properties are seen in moderately to highly plastic soils, such as heavy clays (Little et

al., 2000). Such improved soil properties are the result of three basic chemical

reactions (Fang, 1991): Cation exchange and flocculation-agglomeration; Cementation

(pozzolanic reaction); and Carbonation.

The cation exchange process involves an agglomeration of the fine clay particles into

coarse particles. The cementation process develops from the reaction between calcium

present in lime and silica and alumina in the soil, forming calcium-silicate and

calcium-alumina or calcium-alumina-silicates. The cementations compounds produced

are characterized by their high strength and low-volume change. Previous researchers

reported that small lime additions (from 2–8%) significantly decrease the liquid limit,

plasticity index, maximum dry density, and swell, and increase plastic limit, the

optimum water content, and strength of expansive soils (Basma et al., 1998). It was

reported by Sivapullaiah et al., 1997 that lime added in excess of the amount required

for cations exchange could only produce cementations compounds, which bind the

flocculated particles and develop extra strength (Al-Rawas et al., 2002).


101

The most commonly used products are hydrated high calcium lime Ca (OH) 2, MgO,

calcitic quick lime CaO, and dolomitic quick lime CaO.MgO. Quick lime is used

widely for soil stabilization (TRC180, 1982). Hydrated lime is a fine powder, whereas

quicklime is a more granular substance. Quick lime is more caustic than hydrated

lime, so additional safety procedures are required with this material. The type of the

lime used as a stabilizing agent varies from country to country. Although using quick

lime is more popular in Europe, hydrated lime is used mainly for stabilization but

proportion of quick lime that is used increased to about 25% in 1987 from about 15%

in 1976 (Rollings and Rolling, 1996). According to McCallister and Petry, 1988 both

calcium hydroxide [Ca (OH2] and quick lime (CaO) are common and effective for the

physicochemical treatment of expansive clays.

Although lime is generally used to transform fine-grained soils permanently, it may be

used for shorter-term soil modification for example, to provide a working platform at a

construction site.

• Reaction Mechanisms

Lime soil reactions are complex; however, understanding of the chemistry involved

and the results of field experience are sufficient to provide design guidelines for

successful lime treatment of a range of soils. The sustained and relatively slow

pozzolanic reaction between lime and soil silica and soil alumina (released in the high-

pH environment) is key to effective and durable stabilization in lime-soil mixtures.

Mixture design procedures that secure this reaction must be adopted (Petry, and Little,

2002). In addition to stabilizing materials, lime plays an increasing role in the

reclamation of road bases. Lime has been used effectively to upgrade or reclaim not

only clay soils, but also clay-contaminated aggregate bases and even calcareous bases

that have little or no appreciable clay. Work in the United States, South Africa, and

France has established the benefits of lime stabilization of calcareous bases. The

process results in significant improvements in strength, moisture resistance, and


102

resilient modulus without transforming the calcareous bases into rigid systems that

could be susceptible to cracking and shrinkage, results of field experience are

sufficient to provide design guidelines for successful lime

When lime is added to the soil, hydration of the lime causes an immediate drying of

the soil. Anhydraous quicklime will have a more pronounced drying effect than

hydrated lime. Consequently, lime can prove to be an effective construction expedient

for drying out wet sites.

If lime is added to a plastic soil, plasticity drops, and texture changes. The chemical

changes occurring in the soil are usually explained with the help of some established

mechanisms suggesting cation exchange, flocculation, and aggregation. The first two

reactions are known to occur immediately after lime is either added or allowed to

diffuse into the soil whereas the third reaction is time bound and temperature

dependent and can be considered as a long term reaction. Cation exchange is an

important reaction and is believed to be mainly responsible for the changes occurring

in the plasticity characteristics of the soil.

Depending on the availability of various types of cations in the pore fluid, cation

replacement can take place. In general, the cations are arranged in the order of their

replacing power according to the Iyotropic series, Li+< Na+< H+< K+< NH4+<

Mg2+< Ca2+< Al3+, i.e., any cation will tend to replace the left of it and monovalent

cations are generally replaced by multivalent cations. The replacement of sodium or

potassium ions with calcium will significantly reduce the plasticity index of a clay

mineral.

The addition of lime increases the soil pH, which also increases the cations exchange

capacity. Consequently, even calcium-rich soils may respond to lime treatment with a

reduction in the soil’s plasticity. A reduction in plasticity is usually accompanied by

reduced potential for shrinking or swelling.


103

Due to the addition of lime to the soil the texture of the soil is also changed. As a

result of particle agglomeration clayey soils become more silty and sandy in behavior.

The amount of clay-sized particles (2μm) decreases as the amount of lime in the soil

lime mixtures increases.

Stabilization occurs when the proper amount of lime is added to reactive soil. Ingles

and Metcalf, 1972 recommended the criteria of lime mixture as shown in Table (2.29).

Table (2.29): Suggested Lime Contents (Ingles and Metcalf, 1972)

Soil Type Content for Modification Content for Stabilization

Fine crushed rock 2 – 4 percent Not recommended

Well graded clay gravels 1 – 3 percent ~3 percent

Sands Not recommended Not recommended

Sandy clay Not recommended ~5 percent

Silty clay 1 – 3 percent 2 – 4 percent

Heavy clay 1 – 3 percent 3 – 8 percent

Very heavy clay 1 – 3 percent 3 – 8 percent

Organic soils Not recommended Not recommended

Stabilization differs from modification in that significant level of long-term strength

gain is developed through a long-term pozzolonic reaction. This pozzolonic reaction is

the formation of calcium silicate hydrates and calcium aluminate hydrates as the

calcium from the lime reacts with the aluminates and silicates solubilized from the

clay mineral surface. This reaction can begin quickly and is responsible for some of

the effects of modification. However, research has shown that the full term pozzolonic

reaction can continue for a very long period of time- even many years- as long as

enough lime is present and the pH remains high (Above about 10). As a result of this

long-term pozzolonic reaction, some soils can produce very high strength gains when

lime treated. The key to pozzolonic reactivity and stabilization is a reactive soil and a

good mix design protocol. The results of stabilization can be very substantial increase

in resilient modulus values (by a factor of 10 or more in many cases), very substantial


104

improvements in shear strength (by a factor of 20 or more in some cases), continued

strength gain with time even after periods of environmental or load damage

(autogenously healing) and long-term durability over decades of service even under

severe environmental conditions. (Wibawa, 2003) The change after adding lime to the

soil is shown in Figure (2.43).

Figure (2.43)The Visual Effect of Lime Addition (Wibawa, 2003).

Hisham and Safwat , 1998 found that the swelling pressure of the treated expansive

soil with lime is equal to 3.5 times of that of the untreated expansive soil. Yesilbas,

2004 introduced lime as an admixture up to a maximum of 9 percent. He found that

addition of 3% lime to the samples reduces the swell percentage significantly but after

3 % lime addition there is no significant change in swell percentages.

El-Hoseiny et al., 2005 found that the best amount of mixing lime to be use in

stabilization of the investigated expansive soil is about 6 %.

Abdel-Hady, 2007measured the properties of treated expansive soil using sand-lime

cushion and clay-lime cushion. The lime content which mixed with expansive soil was

ranging from 5 to 20 %. The results showed that the swelling potential expansive soil-


105

lime mix decreased with increasing lime content for laboratory model test, and the free

swell of expansive soil-lime mix decreases with increasing the lime content.

The effect of lime content on the swelling pressure of expansive-lime mix was more

significant for lime content ≤ 5 % (oedometer and laboratory model) as ahown in

Table (2.30).

Table (2.30) Properties of untreated and treated expansive soil-lime mix

Property of soil Type of test Type of soil

Lime content of sand cushion (%)

0 5 10 15 20

Swelling potential (Sp) %

Oedometer

A 23.86 20.21 23.77 20.43 16.8

B 23.24 23.26 22.22 19.44 18.98

C 19.13 18.4 27.7 18.4 26

Laboratory model A 24 19.91 16.24 14.8 13.93

Swelling pressure σS

(kg/cm2)

Oedometer

A 13.5 5.8 15 21 23.3

B 20 18.4 27.7 18.4 26

C 15 10 10 10.8 20

Laboratory model A 63.82 40 47 54 56

• Mixture Design, Pavement Design, and Performance Considerations

Design of lime-stabilized mixtures is usually based on laboratory analysis of desired

engineering properties. Several approaches to mix design currently exist. In addition to

engineering design criteria, users must consider whether the laboratory procedures

adequately simulate field conditions and long-term performance. In terms of roadway

sub-grades, aspects of these procedures are likely to be superseded as AASHTO shifts

to a mechanistic-empirical approach. Laboratory testing procedures include

determining optimum lime requirements and moisture content, preparing samples, and

curing the samples under simulated field conditions.


106

Curing is important for chemically stabilized soils and aggregates particularly lime-

stabilized soils because lime-soil reactions are time and temperature dependent and

continue for long periods of time (even years). Pozzolanic reactions are slower than

cement hydration reactions and can result in construction and performance benefits,

such as extended mixing times in heavy clays (more intimate mixing) and

autogenously healing of moderately damaged layers, even after years of service. On

the other hand, longer reactions may mean that traffic delays are associated with using

the pavement. In addition, protocols for lime-soil mixture design must address the

impact of moisture on performance.

Males states in Little et al., 2000 that lime-stabilization construction is relatively

straightforward. In-place mixing (to the appropriate depth) is usually employed to add

the proper amount of lime to a soil, mixed to an appropriate depth. Pulverization and

mixing are used to combine the lime and soil thoroughly. For heavy clays, preliminary

mixing may be followed by 24 to 48 h (or more) of moist curing prior to final mixing.

This ability to mellow the soil for extended periods and then remix it is unique to lime.

During this process, a more intimate mixing of the lime and the heavy clay occurs,

resulting in more complete stabilization. For maximum development of strength and

durability, proper compaction is necessary; proper curing is also important.

The performance of lime-stabilized sub-bases and bases has been somewhat difficult

to assess in the current AASHTO design protocol because the measure of structural

contribution the structural layer coefficient cannot be ascertained directly.

Indirectly determined coefficients for lime-stabilized systems, however, have been

found to be structurally significant. As AASHTO shifts to a mechanistic-empirical

approach, measurable properties, such as resilient module, were used to assess stress

and strain distributions in pavement systems, including stabilized bases and sub-bases.

These properties were coupled with shear-strength properties in assessing resistance to

accumulated deformation.


107

The lime industry (Little, 2000) has submitted a three to four-step design and testing

protocol to be considered for inclusion in the AASHTO design protocol:

• Step 1: Determine optimum lime content using the Eades and Grim pH test (ASTM

D-6274);

• Step 2: Simulate field conditions. Use AASHTO T-180 compaction and 7-day

curing at 40°C to represent good-quality construction techniques. After curing,

subject samples to 24 to 48 h of moisture conditioning via capillary rise (soak);

• Step 3: Verify compressive strength, stiffness, and moisture sensitivity, and

measure unconfined compressive strengths using ASTM D-5102 methods.

For most applications, the above three steps are sufficient because design

parameters such as flexural strength, deformation potential, and stiffness (resilient

modulus) can be approximated from unconfined compressive strength. For more

detailed (Level 2) designs, a direct measure of resilient modulus may be required;

and

• Step 4: Perform resilient modulus testing using AASHTO T-294-94 or expedited

(and validated) alternatives, such as the rapid triaxial test. This protocol and its

mechanistic-empirical basis provide a sound foundation for future lime-

stabilization applications.

In addition to conventional mix in-place or batch mixing, other methods for

incorporating lime include electrical, drill-hole, pressure, and deep-plow whom are

basically used to increase the rate of lime migration through soil.

b) Coal fly ash stabilization

Large quantities of coal are being burnt in thermal power stations to meet the ever

increasing demand for thermal power. Combustion of coal results in a residue

consisting of inorganic mineral constituents and organic matter which is not fully


108

burned. The inorganic mineral constituents from ash: About 80% of this ash is fly ash.

Environmentally safe disposal of large quantities of ash is not only tedious but also

expensive. To reduce the problems of disposal, great efforts are being made to utilize

fly ash. The use of fly ash as a soil-stabilizing agent is beneficial for improving the

engineering properties of the soil, while at the same time it provides an opportunity for

the utilization of an industrial waste that will otherwise require costly disposal

(Çetiner, 2004).

Fly ash produced from the burning of pulverized coal in a coal-fired boiler is a fine-

grained, powdery particulate material that is carried off in the flue gas and collected

from the flue gas by means of electrostatic precipitators (Vassilev et al., 2003).

Fly ash is useful in many construction applications because it is a pozzolan, meaning it

is a siliceous or alumina-siliceous material which in itself possess little or no

cementations value but will, in finely divided form and in the presence of moisture,

chemically react with calcium hydroxide at ordinary temperatures to form compounds

possessing cementations properties (ASTM, 1993). A microscopic view of fly ash

reveals mainly glassy spheres with some crystalline and carbonaceous matter. The

principal chemical constituents are silica (SiO2), alumina (Al2O3), ferric oxide (Fe2O3),

and calcium oxide (CaO). Other components are magnesium oxide (MgO), sulfur

trioxide (SO3), titanium oxide (TiO2), alkalies (Na2O and K2O), phosphorous oxide

(P2O5), and carbon (related to loss on ignition).

Water added to fly ash usually creates an alkaline solution, with pH in the range 6 to

11. Because of the variations in coals from different sources, as well as the differences

in the design of coal-fired boilers, not all the fly ash is the same (Çetiner, 2004).


109

Factors affecting the physical, chemical, and engineering properties of fly ash include

(TFHRC, 2003):

• Coal type and purity

• Degree of pulverization

• Boiler type and operation

• Collection and stockpiling methods

Two classes of fly ash are defined in ASTM C 618: Class F fly ash, and Class C fly

ash. Class F fly ash is normally produced from burning anthracite or bituminous coal.

This class fly ash has pozzolanic properties. Class C fly ash is normally produced from

burning lignite or sub-bituminous coal. This class of fly ash, in addition to having

pozzolanic properties, also has some self-cementing properties, meaning that it has

ability to harden and gain strength in the presence of water alone. Typical chemical

compositions of Class F and Class C fly ashes are given in Table (2.31).

Table (2.31): Typical Chemical Compositions of Class F and Class C Fly Ashes

(expressed as percent by weight) (TFHRC, 2003).

Component Class F Fly Ash Class C Fly Ash

SiO2 20 – 60 40 – 60 Al2O3 5 – 35 10 – 30 Fe2O3 10 – 40 4 – 15 CaO 1 – 12 5 – 30 MgO 0 – 5 1 – 6 SO3 0 – 4 0 – 4

Na2O 0 – 4 0 – 6 K2O 0 – 3 0 – 4

Loss on Ignition 0 – 15 0 – 3


110

i) Stabilization with Non-Self-Cementing Coal Fly Ash

Little et al. 2000 summarized that the use of class F fly (non-self-cementing coal ash)

is useless as a stabilizer without the addition of lime as a source of calcium. This fly

ash is produced from the combustion of bituminous, anthracite, and some lignite coals

is pozzolanic but not self-cementing. To produce cementations products, an activator

such as Portland cement or lime must be added. Non-self-cementing fly ash can be

used to produce a lime/fly ash/aggregate (pozzolanic-stabilized mixture [PSM]) road

base. The mixture developed must possess adequate strength and durability for its

designated use, be easily placed and compacted, and be economical. Quality mixtures

have been produced with lime content ranging from 2 to 8 percent (by weight). Fly ash

content may range from 8 to 15 percent (by weight). Typical proportions range from 3

to 4 percent lime and 10 to 15 percent fly ash. When needed, 0.5 to 1.5 percent

portland cement can be used to accelerate the initial strength gain. The resulting

material is similar to cement-stabilized aggregate base in its production, placement,

and even appearance. PSM bases can be placed with conventional equipment and used

with recycled base materials. The cost of PSM bases varies significantly from area to

area, but is often lower than that of alternative base materials. PSM bases are not

heated and require less energy to place than asphaltic bases. The strength development

of a PSM is highly dependent on curing time and temperature. In many cases,

minimum curing times are specified, along with an allowable curing temperature

range that will produce the required strengths. Target strengths for mix design

development should allow for the fact that PSM bases will continue to gain significant

strength after the curing period has ended. PSM bases also have the inherent ability to

heal or re-cement across cracks if moisture is present, and if unreacted lime and fly

ash are available. This phenomenon is known as autogenous healing. The deleterious

effects of shrinkage cracks may be reduced in PSMs designed to produce slower rates

of pozzolanic reaction over longer periods of time.


111

ii) Stabilization with Self-Cementing Fly Ash

Stewart (Little et al., 2000) explains that lime-Class C fly ash combinations have been

successfully added to clay soils for stabilization. Successful application is normally

achieved by adding lime first to modify soil properties and then adding fly ash to react

with the lime pozzolanically to achieve a specified strength gain. This approach is

attractive when a specified strength level cannot be achieved with fly ash alone.

Although a two-stabilizer, two step process may be economically disadvantageous

under certain circumstances, it offers certain advantages, including the ability and time

to modify the soil with lime before adding the fly ash for stabilization. Lime allows an

extended mellowing period, which is necessary with some expansive clays.

Stewart (Little et al., 2000) further explains that with the passage of the Clean Air Act

in the 1970s, many utilities began burning low-sulfur sub-bituminous coals. An

unexpected benefit of burning this lower-sulfur coal was the production of a new type

of fly ash, designated by ASTM as Class C coal fly ash. This material is self-

cementing because of the presence of calcium oxide (CaO) in concentrations typically

ranging from 20 to 30%.

Most of the CaO in Class C fly ash, however, is complexly combined with pozzolans,

and only a small percentage is ‘‘free’’ lime. This characteristic may impact the

suitability of the material for stabilization of plastic clay soils. Sub-bituminous coals

are now shipped by rail to power plants throughout the United States, although the

largest concentrations of sub-bituminous coal combustion are west of Ohio. Class C

fly ashes are shipped by truck and rail into many construction markets throughout the

United States.

According to ASTM D 5239, ‘‘Standard Practice for Characterizing Fly Ash for Use

in Soil Stabilization,’’ the use of self cementing fly ash can result in improved soil

properties, including increased stiffness, strength, and freeze-thaw durability; reduced

permeability, plasticity, and swelling; and increased control of soil compressibility and


112

moisture. Although these ashes have properties similar to those of Portland cement,

they also have unique characteristics that must be addressed by both the mix design

and construction procedures.

The primary design consideration is the rate at which the fly ash hydrates upon

exposure to water. Recognizing and properly addressing the hydration characteristics

of the ash can result in a significant enhancement of the potential benefits of its use.

Even self-cementing ashes can be enhanced with activators such as Portland cement or

hydrated lime. This is particularly true if the self-cementing ash does not have enough

free lime to fully develop the pozzolanic reaction potential.

In clay soil stabilization the activator (lime or cement) may play a dual role:

modifying the clay and activating the fly ash. A key role for fly ash in lime-clay

stabilization is to provide the required pozzolanic-based strength gain for ‘‘non-(lime)

reactive’’ clays.

Another classification by ASTM D 5239 classifies fly ashes into three categories

according to their soil stabilization performances (Çetiner, 2004).

a) Non Self-Cementing (Class F) Fly Ash Stabilization

Non self-cementing fly ash, by itself, has little effect on soil stabilization. It is a poor

source of calcium and magnesium ions. The particle size of fly ash may exceed that of

the voids in fine-grained soils, precluding its use as filler. However, this fly ash in

poorly graded sandy soils may be suitable filler and, as such, may aid in compaction,

may increase density, and may decrease permeability.

b) Non Self-Cementing (Class F) Fly Ash Mixed With Cement or Lime

The advantage of adding fly ash to fine-grained soils, along with cement or lime, is for

its pozzolanic properties and improved soil texture. Some clay is pozzolanic in nature

and only requires lime to initiate the pozzolanic reaction. The use of this fly ash is


113

suitable with clays requiring lime modification, provided lime is added to promote the

pozzolanic reaction.

c) Self-Cementing (Class C) Fly Ash Stabilization

This fly ash is a better source of calcium and magnesium ions although not as good as

lime or Portland cement. Self-cementing fly ash contains varying amounts of free

(uncombined) lime (0 to 7% CaO by weight) that can provide cations exchange and

ion crowding to fine-grained soils when used in significant amounts. It has been used

successfully to control swell potential of expansive soils. It has also been used to

stabilize coarse-grained soils.

• Mix Design and Construction Considerations

Stewart (Little et al., 2000) explains that to achieve optimum results, a thorough

understanding of the influence of the compaction delay time and moisture content of

the stabilized materials is essential. Ash hydration begins immediately upon exposure

to water. Strict control of the time between incorporation of the fly ash and final

compaction of the stabilized section is required. A maximum delay time of 2 h can be

employed if contractors are not experienced in ash stabilization or if achieving

maximum potential strength is not a primary consideration for the application. This

limits the clay mellowing period; thus a two-step process is often required in lime-fly

ash clay stabilization. The first step is lime modification followed by fly ash addition.

A maximum compaction delay of 1 h is commonly specified for stabilization of

pavement base or sub-base sections when maximum potential strength is required.

Achieving final compaction within the prescribed time frame generally requires

working in small, discrete areas, an approach that differs from the methods used for

lime stabilization. Stewart (Little et al., 2000) explains that the second major design

consideration is that there is an optimum moisture content at which maximum strength

will be achieved. This optimum moisture content is typically below that for maximum

density often by as much as 7 to 8%. The strength of the stabilized material can be


114

reduced by 50% or more if the moisture content exceeds the optimum for maximum

strength by 4 to 6%. An understanding of both the influence of compaction delay and

moisture control of the stabilized material is essential to achieving the optimum

benefit from stabilization applications that use self-cementing fly ashes.

c) Portland cement stabilization

Little et al. 2000 states that Since 1915, more than 100,000 miles of equivalent 7.5 m

(24 ft) wide pavement bases has been constructed from cement-stabilized soils.

Cement has been found to be effective in stabilizing a wide variety of soils, including

granular materials, silts, and clays. Cement stabilization develops from the

cementations links between the calcium silicate and aluminate hydration products and

the soil particles. Cement addition to clay soils reduce the liquid limit, plasticity index,

and swelling potential, and increase the shrinkage limit and shear strength (Nelson and

Miller, 1992). Previous work carried on by Basma et al., 1998 showed that the

addition of cement in small percentages (3–9% of the dry weight of the soil) resulted

in a decrease in the swelling characteristics.

i) Definitions and Applications

Cement-stabilized materials generally fall into two classes soil-cement and cement

modified soil. Soil-cement is a mixture of pulverized soil material and/or aggregates,

measured amounts of Portland cement, and water that is compacted to a high density.

Enough cement is added to produce a hardened material with the strength and

durability necessary to serve as the primary structural base layer in a flexible

pavement or as a sub-base for rigid pavements. Cement-treated aggregate base and

recycled flexible pavements are considered soil-cement products. Cement-modified

soil is a soil or aggregate material that has been treated with a relatively small

proportion of Portland cement (less cement than is required to produce hardened soil-

cement) with the objective of altering undesirable properties of soils or other materials

so they are suitable for use in construction. Cement-modified soil is typically used to


115

improve sub-grade soils or to amend local aggregates for use as base in lieu of more

costly transported aggregates. Alternative terms include cement-treated or cement-

stabilized soil or sub-grade (Little et al., 2000).

ii) Stabilization Mechanisms

Portland cement is composed of calcium-silicates and calcium-aluminates that, when

combined with water, hydrate to form the cementing compounds of calcium silicate

hydrate and calcium-Aluminate-hydrate, as well as excess calcium hydroxide. Because

of the cementations material, as well as the calcium hydroxide (lime) formed, Portland

cement may be successful in stabilizing both granular and fine-grained soils, as well as

aggregates and miscellaneous materials.

A pozzolanic reaction between the calcium hydroxide released during hydration and

soil alumina and soil silica occurs in fine-grained clay soils and is an important aspect

of the stabilization of these soils. The permeability of cement stabilized material is

greatly reduced. The result is a moisture-resistant material that is highly durable and

resistant to leaching over the long term.

iii) Mix Design Considerations

Mix design requirements vary depending on the objective. Soil-cement bases generally

have more stringent requirements than cement-modified soil sub-grades. For soil-

cement bases, two types of testing have typically been used—durability tests and

strength tests.

The Portland Cement Association has developed requirements for AASHTO soils A-1

to A-7 that make it possible to determine the durability of cement on the basis of

maximum weight losses under wet-dry (ASTM D559) and freeze-thaw (ASTM D560)

tests. Many state departments of transportation (DOTs) currently require minimum


116

unconfined compressive strength testing (ASTM D1633) in lieu of these durability

tests (Little et al., 2000).

This requirement is often based on many years of experience with soil-cement. The

advantage of using these strength tests is that they can be conducted more rapidly than

the durability tests (7 days vs. 1 month) and require less laboratory equipment and

technician training. However, achievement of a specified strength does not always

ensure durability. Typical minimum strength varies from 200 to 750 pounds per square

inch. For cement-modified soils, the engineer selects an objective and defines the

cement requirements accordingly.

Objectives may include one or more of the following: reducing the plasticity index

(Atterberg limits, ASTM D4318); increasing the shrinkage limit; reducing the volume

change of the soil (AASHTO T116); reducing clay/silt-sized particles (hydrometer

analysis); meeting strength values/indexes such as the California Bearing Ratio

(ASTM D1883) or triaxial test (ASTM D2850); and improving resilient modulus

(ASTM D2434). Cement has been incorporated successfully into soils in the field with

plasticity indexes ranging as high as 50 (Little et al. 2000).

iv) Construction Considerations

Construction of soil-cement and cement-modified soil is normally a fast,

straightforward process. Cement can be incorporated into soil/aggregate in a number

of ways. The most common method is to spread dry cement in measured amounts on a

prepared soil/aggregate and blend it in with a transverse single-shaft mixer to a

specified depth. Cement slurries in which water and cement are combined in a 50/50

blend with a slurry-jet mixer or in a water truck with a recirculation pump have been

used successfully to reduce dusting and improve mixing with heavy clays (Little et al.,

2000). Sometimes, central mixing plants are employed. Twin shaft continuous-flow

pug mills are most common, although rotary-drum mixers have been used as well.


117

Although construction procedures are similar for soil-cement and cement-modified

soil, pulverization requirements need to be adjusted accordingly.

The recommended pulverization for both granular and fine-grained soil (for soil

material exclusive of gravel or stone) is as follows:

Sieve Size Soil-Cement Cement-Modified Soil

45 mm (1 3/4 in.) - 100

25 mm (1 in.) 100 - 4.75 mm (#4) 80 60

Compaction is normally a minimum of 95 percent of either standard or modified

proctor density (ASTM D588 or ASTM D1557, respectively), with moisture content

+2 percent of optimum. Prusinski (Little et al., 2000) explains that Soil-cement shrinks

as a result of hydration and moisture loss. Shrinkage cracks develop in the base, and

can reflect through thin bituminous surfaces as thin (< 3 mm [1/8 in.]) cracks at a

spacing of 2 m (6 ft) to 12 m (40 ft). If proper construction procedures are followed,

shrinkage cracks may not reflect through, and if they do, they generally pose no

performance problem. However, cracks can compromise performance if they become

wide and admit significant moisture. A number of techniques have been used to

minimize this problem, including:

• compaction at a moisture content slightly drier than optimum;

• precracking through inducement of weakened planes or early load applications;

• delayed placement of surface hot mix; reduced cement content; and use of

interlayers to absorb crack energy and prevent further propagation.

Prusinski further explains (Little et al., 2000) that though portland cement no longer

widely used, probably because of higher first costs, combination treatment with lime

and portland cement of plastic clay soils offers an attractive alternative. The initial


118

lime treatment offers the ability to reduce plasticity and improve mixing efficiency,

and since the lime allows a mellowing period of up to three or four days, time is not a

constraint.

The addition of a relatively low amount of portland cement (approximately 2 to 3%)

upon final mixing provides the required strength even in clays that have moderate or

low reactivity with lime.

d) Cement By-Pass Dust (CBPD) Stabilization

Cement by-pass dust (CBPD) is a fine powder material produced as a by-product of

the manufacturing of Portland cement (Al-Rawas et al., 2002), which is also referred

to in the literature as “cement kiln dust.” It is generated during the calcining process in

the kiln. The composition of CBPD is quite variable from source to source, due to raw

materials and process variations. It is primarily made up of variable amounts of fine

calcined and uncalcined feed materials, fine cement clinker, fuel combustion by-

products, and condensed alkali compounds. Morgan and Halff, 1984 reported that

cement kiln dust is a cost effective and efficient solidifying agent compared to lime,

cement, fly ash, and sulphur. It was reported that cement kiln dust results in significant

reduction in the plasticity index and swell percent, and an increase in the compressive

strength of soil (Zaman et al., 1992). Zaman et al., 1992 showed that the addition of

25% cement kiln dust of the dry weight of a highly expansive soil resulted in reducing

the swell percent from 9.1% to 0%, and increasing the unconfined compressive

strength from 103 kPa to 290 kPa in a 56-day curing period.

e) Slag Stabilization

Slag materials are produced as by-products, and they can be grouped, based on their

composition, into two categories: iron and steel slags, and nonferrous slags (Al-

Rawas, et al., 2002). Blast furnace slag is derived from producing iron in a blast

furnace. The slag consists mainly of silicates and alumino-silicates of lime. Blast


119

furnace slag can be produced in three forms: air-cooled, granulated, and expanded

forms. Air-cooled slag is commonly used in concrete, asphalt, and road bases, and as

fill material. Granulated slag is used as slag-cement. Expanded slag is used as an

aggregate. Steel slags consist mainly of calcium, iron, unslaked lime, and magnesium.

Steel slag usually contains sufficient amounts (on the order of 30–50%) of lime, which

can be mixed with fly ash to provide lime for pozzolanic reactions (Chu and Kao,

1993). A useful review of the composition, types, and uses of various types of slags

was presented by Collins and Ciesielski, 1994.

Calcium-based stabilizers have been around for many years and have a well-

documented record of improving workability and decreasing expansiveness in soils

with expanding lattice clay minerals. Lime treatment offers the advantage of fairly

simple and reliable test methods to determine the required treatment level achieve the

required amount of lime modification and swelling reduction (Eades and Grim, 1966).

The major disadvantage of all calcium-based stabilizers that increase the pH of the

treated soil is the possible reaction of lime with soluble sulfates. This so-called

“sulfate reaction” reduces the amount of lime available to participate in the stabilizing

pozzolanic reactions, and even worse, the lime can react with the sulfates to produce

the highly expansive hydrated minerals ettringite and thaumasite (Mitchell, 1986) In

order for the sulfate reaction to occur in soils, at least ten percent clay and a pH greater

than 10.5 are required (Hunter, 1988). The minimum amount of soluble sulfate

required has been estimated as ranging from 2,000 ppm (Perrin, 1992) to as low as 700

ppm (Hunter, 1988). Puppala, et al., 2000 detected ettringite in soils having as little as

320 ppm soluble sulfates. Unlike clay swelling which can take place as the colloids

condense water from its vapor in the soil atmosphere, the sulfate reaction requires that

there be sufficient liquid water in the soil to maintain the sulfates in solution. The

reaction is believed to occur in as little as a “few months “ after lime treatment

(Hunter, 1988) and can continue to occur for years as long as the soil pH remains

above 10.4 and sulfates remain in solution (Puppala, et al., 2000).


120

The use of lime to stabilize expansive clay soils is considered to be most effective

when dry CaO or Ca(OH)2 or slurried Ca(OH)2 can be mixed directly with the soil

along with an appropriate amount of water, allowed to cure, followed by mixing with

additional water and possibly lime for compaction, and finally, compacted to dry unit

weights close to the T-99 maximum. There are several variations on this approach

(Transportation Research Board, 1987). The technique of injecting lime slurries under

pressure directly into soils beneath the surface has had mixed success. Problems are

usually attributed to incomplete penetration of homogeneous, intact clays (Pengelly, et

al., 1997).

2.5.3.2 Non-Calcium Based Chemical Stabilizers

A variety of chemical additives which have demonstrated ability to improve the

engineering properties and workability of soils and soil-aggregate mixtures have been

developed. More recently, partly in response to the increasing recognition of the

sulfate problem described above, some of the old and some new chemical admixtures

have been put forth as being effective alternatives for the calcium-based treatments of

expansive clay soils. Many of the non-calcium based stabilizers are proprietary, and

the mechanisms of their stabilizing action are not clearly understood. Some are

believed to act in like calcium-based stabilizers in that they break down the clay

minerals (weathering) thereby reducing the amount of clay available to expand. Other

suggested mechanisms include replacement of adsorbed monovalent cations (ion

exchange), which lowers the water deficiency (suction) in the soil in a manner similar

to the role of the Ca+2 in lime treatment. Table (2.28) is a partial list of non-calcium

based chemical stabilizers purported to be effective in reducing soil swelling.

The review of the available literature for non-calcium based stabilizers revealed that

many of the non-calcium based chemicals seem to be most effective in treating soils,

aggregates and soil-aggregate mixtures that contain only small amounts of clay of low

plasticity.


121

The applications have been mainly to surface courses and bases for low volume roads.

Beneficial effects (stabilization) in these applications refer to increases in strength and

cohesiveness, improvements in workability, and decreases in compressibility of the

treated materials. With the exception of EcSS-3000, EMC-SQUARED, HIExC, and

Roadbond EN-1, there is no experience, nor claim for the use of most of the

proprietary chemicals specifically to reduce swelling of highly plastic clay soils. As

indicated in Table (2.32), the advantages claimed for the first three products relates to

their ability to stabilize expansive clay soils containing high levels of sulfates.

Some additional perceived limitations of the use of the non-calcium based stabilizers

include the as yet relatively small number of well-documented field trials and

demonstrations and difficulty in determining appropriate treatment levels. Currently

there are no simple, rapid, widely recognized test methods to determine appropriate

treatment levels. The materials may not be available everywhere, and costs may be

high at the treatment levels required for very highly plastic, expansive clays

(Hardcastle, 2003).

Other chemical compounds, such as sodium silicate, calcium hydroxide, sodium

chloride, calcium chloride, and potassium nitrate, have been utilized in stabilizing

soils. However, there is no supporting evidence that any of them has economically

worthwhile benefits (Gromko 1974).

Furthermore, Nelson and Miller, 1992 reported that there is insufficient evidence that

salts other than sodium chloride and calcium chloride have adequate soil stabilization

capabilities to be economically justifiable.


122

Table (2.32): Non-calcium based chemical soil stabilizers (after Hardcastle, 2003).

Stabilizing Chemical or Trade Name Manufacturer or Supplier Remarks

Ammonium and Potassium Lignosulfates Hayward-Baker Fort Worth, TX Injectable

Barium Chloride Barium Hydroxide Various Expensive, forhigh

SO4 clays

BIO-CAT (bioenzyme) Soil Stabilization Products, Inc. Merced, CA

CBR-PLUS CBR-PLUS North America, Inc. Vancouver, B.C.

Condor SS (sulfonatednaphthalene)

Earth Science Products Corp.Wilsonville, OR

Not effective with Smectite

Consolid-444 (ammonium chloride) American Consolid, Inc. Davenport, IA No contact

EcSS 3000 (sulfonated oil) Environmental Soil Stabilization, L.L.C. Arlington, TX

Injectable, highSO4 clays

EMC Squared(bioenzyme) Soil Stabilization Products, Inc. Merced, CA

Proprietary, for high SO4 clays

HIExC Environmental Soil Stabilization, L.L.C. Arlington, TX

Injection in highSO4 clays

Perma-Zyme(enzyme) International Enzymes @perma-zyme.com

PSCS-320 (bioenzyme) Alpha Omega Enterprises

Road Bond EN1 Sulfonated D-Limonene

C.S.S. Technology, Inc. Weatherford, TX 76086


123

2.5.4 Mechanical Treatments

The three types of treatments described as mechanical are surcharge stresses, fiber

reinforcement and Geogrids.

2.5.4.1 Surcharge Stress

Swell can be prevented if expansive clays can be loaded with a surcharge large enough

to counteract the expected swelling pressures. This is generally applicable only for

soils with Slow to moderate swelling pressures (Nelson and Miller, 1992). If the

surcharge material has a dry unit weight of 125 lbs/ft3, each foot of thickness provides

a surcharge stress of 125 lbs/ft2 or 0.063 tons/ft2. It’s clear that unless high fills are

required for highway grade considerations, the surcharge method is likely to be cost-

effective only for expansive soils with low swelling pressures. The surcharge material

itself can be a stabilized or encapsulated expansive soil or more commonly, a

nonexpansive borrow material. In some applications part of the expansive soil is

removed prior to placement of the surcharge material. The successful application of

this approach requires that laboratory swell tests be performed to determine the swell

pressure as the soil as goes from its initial moisture condition to the final stable

moisture condition. This treatment method can only be used in conjunction with new

construction of the pavement section (Hardcastle, 2003).

2.5.4.2 Fiber Reinforcement

Hardcastle, 2003 stated that the use of randomly oriented polypropylene fibers 25 to

50 mm in length for stabilizing expansive soils has been investigated by Puppala and

Musenda, 2000. In tests involving 0.3, 0.6 and 0.9 percent fibers by dry weight of

compacted soil, the inclusion of the fibers increased the unconfined compressive

strength over that of the untreated high plasticity CH soils. The strains required to

mobilize the increased strengths were larger than for the untreated soils. While the

fibers slightly reduced the shrinkage of soil pastes, the conventional swelling


124

measured in oedometer tests increased when the soils were treated with fibers. The

increased free swell was attributed to a more uniform distribution of moisture in the

compacted samples caused by the moisture paths created by the fibers. The use of 0.9

percent fiber reduced the swell pressures in the soil with a liquid limit of 82 percent

from 0.26 tons/ft2 to 0.22 tons/ft2. In the slightly less plastic soil (liquid limit of 73

percent), the swelling pressure was reduced from 0.39 tons/ft2 to 0.22 tons/ft2. The

investigators concluded that “the mechanisms causing the swelling pressure reductions

still need to be evaluated”. The use of fibers for reducing swelling doesn’t address any

of the three required conditions for swelling. Fibers don’t change the swelling clay

minerals in any way; fibers don’t reduce the water deficiency condition of the soil

from what it is in its compacted condition and fibers don’t make water unavailable to

the potentially expansive soil.

2.5.4.3 Geogrids

Geogrids can provide mechanical reinforcement of soft, compressible clay sub-grades.

Geogrids increase the strength and bearing capacity of the soil and can therefore be

considered to be a soil stabilizer. Geogrids are promoted by their manufacturer as

being an alternative to lime stabilization for heavy clays containing sulfates (Tensar

Earth Technologies, Inc., 2001), but it is difficult to see how they can prevent swelling

of soils containing highly expansive clay minerals. Like fibers, geogrids don’t alter the

structure of expanding lattice clay minerals in soil; geogrids can’t reduce the water

deficiency if the soil’s mineraology, condition and environment create a water

deficiency; and geogrids cannot prevent the entry if water becomes available to a soil

possessing the first two requirements. It does seem likely that like fibers, geogrids may

reduce swell pressures by providing some restraint in the form of tensile

reinforcement. By providing paths for the entry of moisture, geogrids may also

promote more uniform soil swelling when used with uniform soils.


125

Chapter 3

BEHAVIOR OF SUPPER-STRUCTURE RESTING ON TREATED EXPANSIVE SOIL

“CASE STUDY”

3.1 General

This chapter presents a case study of a severely cracked reinforced concrete building

constructed on the middle Mokattam plateau. The stratigraphy of the middle

Mokattam plateau was presented and the geotechnical properties of the soil at the site

area were investigated using direct and indirect measurement tests. The building was

founded on a shallow foundation resting on treated expansive clay using medium sand

replacement. The field investigation revealed that the damage was caused by

differential heaving as a result of water leakage from a nearby drainage utility. The

movement of the building was simulated depending on six survey readings during a

time period reached about four years nearly after the building inclination. Also the

horizontal displacement for the building was predicted using two semi empirical

equations. The predicted results compared reasonably well with the actual

observations in the field.

3.2 Description of building and cracks

The case study described herein deals with expansive soil damage to a multi-story

reinforced concrete consists of six floors residential building located in El-Mokattam

medium plateau. The building was constructed in 1989 in El-Mokattam medium

plateau as a part of a large housing project, Figure (3.1) shows a satellite photo for this

project. The foundation system consists of isolated reinforced concrete footings

connected with tie beams (25x60 cm) resting on plain concrete raft (50 cm thick),

Figure (3.2).


126

The preliminary geotechnical investigation data provided some clues for the presence

of expansive soils in the construction area. Accordingly some recommendations were

taken into consideration before construction and can be summarized in the following

notes:

1. The site should be excavated in dimensions exceeds the plain concrete dimensions

by 1.00m, and to level (-3.00 m) from the boreholes zero level.

2. Soil replacement with clean course sand as a method of treatment with thickness

1.50m, and dimensions exceeds the plain concrete dimensions by 1.00m shall be

applied.

3. The replacement soil should be good compacted in layers 25 cm each using roller

vibrators, such that the minimum required density using the Modified Proctor test

method shall be at least 95% of the maximum dry density with optimum moisture

content ±2%.

4. Bearing capacity at the replacement soil top equal 1.2 kg/cm2.

5. Plain concrete raft with thickness 50cm to be laid over the replaced soil.

6. The suggested foundation system is isolated reinforced concrete footings connected

with rigid ground beams of 30cm x 80cm in dimension.

7. All ground beams should be laid over the isolated footings and have a continuous

RFT 6 Ø 16 along the top and bottom sides tied by 6 Ø 10/m shear stirrups.

8. Sidewalks with 1.50m wide should be paved around buildings with taking all the

precautions to prevent any leakage from any sources of water to the soil.

9. Any backfilling shall be with compacted clean sand as the excavated soil cannot be

used in backfilling.


127

A lawsuit was filed in 1998 by the existing homeowners (plaintiffs), claiming that the

soils and foundations did not adequately support the structure, causing instability and a

dangerous condition that required extensive repairs. The plaintiffs also claimed that

they were unaware of the defective and dangerous soil conditions, or the extent of the

repairs necessary to correct them.

With time, the damage progressed, and in October 1998, a consulting engineer carried

out an investigation of the building damages and stated the deviations that occurred in

the following:

• Damage to the building superstructure consisted of noticeable floors sloping

especially in the upper floors.

• Cracks in the longitudinal beams adjacent to the neighbor building specially at

the third, fourth and fifth floors.

• Significant separation between the upper floor slabs and the connecting

columns.

• Also it was observed that the greatest amount of distortion occurs at the side of

building adjacent to the neighbor one as the vertical articulation joint between

them had opened up significantly at the top.

• Some of the exterior columns exhibited horizontal and vertical displacements.

Figures (3.3a & 3.3b) show some of the damages to the building superstructure,

according to this investigation notes it was recommended the following:

1. Periodic monitoring of the cracks probagation.

2. Periodic monitoring of the building movement.


128

Figure (3.1): A satellite photo for the housing project in the Mokattam middle plateau.

Figure (3.2): The utilized foundation system.

Expansive Soil Layer = 2.00 m

Porous Calcareous Sand Stone Layer

Sand Replacement Layer = 1.20 m

Fill

P.C thickness = 0.50 m

Tie Beams= (30 x 60 cm)Isolated footings

(0.00)

(‐2.00)

(‐2.50)

(‐3.70)

(‐5.70)


129

Figure (3.3a): Photo for defects and cracks on front 1

Figure (3.3b): Photo for defects and cracks on front 2

Differential movement between buildings

Differential movement between buildings


130

Accordingly a comprehensive surveying field-monitoring program has been designed

and executed. This program starts from 15 December 1998 to 9 July 2002 .The main

aim of the program was to detect the spatial mode of movement propagation, in the

horizontal and vertical direction, referring to a fixed measuring reference bench marks.

Therefore, twelve vertical lines of inclination-marks (signed as A to H) are used to

monitor the building inclinations in X & Y directions. Another twelve settlement-

marks (signed as 1 to 12) have been fixed on 8 columns located on the severe three

sides of the building to calibrate the developed vertical displacement. The locations of

vertical lines of inclination-marks and the settlement-marks are shown in Figure (3.4).

The building inclinations in both X and Y directions, as well as in the vertical

displacements have been monitored four times in 15/12/1998, 31/12/1998, 22/2/1999,

and 30/6/1999, and these results were presented by (Ahmed, 2000), and monitoring

was continued on 22/2/2000, 5/7/2001 and finally finished on 9/7/2002.

The vertical displacements for the taken measurements are summarized in Table (3.1).

While the differential horizontal movements or inclinations in the (X-Y plane) with

respect to the fixed vertical planes of inclination are outlined in Table (3.2).The

measurement accuracy where about 2mm, and some points where missed with time as

shown in Table (3.2).

Although some repair works were carried out at that time, the remedial measures

showed that these works failed to stop the movements, and since December 1998 the

condition of the building has started to deteriorate.


131

Table (3.1): Vertical displacement in centimeters for the specified points.

Observation date

Location 31/12/1998 22/2/1999 30/6/1999 22/2/2000 5/7/2001 9/7/2002

A (1) -0.4 0.0 0.0 0.00 0.00 0.00

B (2) -0.35 2.8 3.5 5.30 5.45 5.45

C (3) 0.05 5.50 7.4 13.17 15.25 15.95

D (4,5) -0.2 5.50 9.90 19.40 23.35 25.15

E (6,7) 0.18 5.60 9.80 20.7 24.20 26.85

F (8,9) 0.20 5.10 7.30 15.4 18.20 19.83

G (10,11) -0.18 3.60 4.00 8.60 9.18 9.97

H (12) -0.1 3.00 2.40 3.18 3.20 3.20

Average vl. displacement -0.1 3.88 5.54 10.72 12.35 15.20

Figure (3.4): Locations of vertical lines of inclination-marks and settlement-marks for

the defected building.


132

Table (3.2): The detected structure horizontal displacement in centimeters: Observation

date 15/12/1998

* 31/12/1998 22/2/1999 30/6/1999 5/7/2001 9/7/2002

Point

No.

Floor

No. X-dir. Y-dir. X-dir. Y-dir. X-dir. Y-dir. X-dir. Y-dir. X-dir. Y-dir. X-dir. Y-dir.

A

1 0.00 0.00 0.00 0.00 0.00 0.00

3 -2.50 -2.80 -2.80 -2.80 -3.60 -3.90

4 -0.40 -0.80 -0.90 -0.90 -1.90 -2.30

6 -2.30 -2.70 -2.90 -2.90 -4.35 -5.0

roof 0.60 0.00 -0.10 -0.10 missed missed

B

1 0.00 0.00 0.00 0.00 0.00 0.00

2 2.90 2.90 2.90 2.90 2.60 2.60

3 -10.0 -10.0 -10.0 -10.0 -10.7 -10.8

4 -5.90 -6.10 -6.20 -6.20 -7.0 -7.1

5 -3.20 -3.50 -3.60 -3.60 -4.9 -4.90

6 -0.70 -0.90 -1.10 -1.2 -2.60 -2.90

roof 12.45 11.9 11.95 12.05 10.35 missed

C

1 0.00 0.00 0.00 0.00 0.00 0.00

2 5 4.9 4.9 4.9 4.9 4.9

3 1.90 1.9 1.80 1.80 1.60 1.60

4 -0.20 -0.30 -0.40 -0.60 -1.00 -1.20

roof -20.5 -20.7 -20.8 -21.0 missed missed

D

1 0.00 0.00 0.00 0.00 0.00 0.00

2 -2.40 -2.70 -2.50 -2.80 -2.40 -2.40

3 -2.60 -2.80 -2.70 -3.10 -3.10 -3.10

4 -4.70 -5.10 -5.10 -5.40 -5.45 -5.80

6 -4.20 -4.70 -4.70 -5.30 -5.80 -5.90

roof -7.90 0.60 -8.40 0.40 -8.40 0.10 -9.2 0.80 -10.1 missed -9.90 missed

E 1 0.00 0.00 0.00 0.00 0.00 0.00

roof 7.80 13.90 7.70 14.30 7.70 14.20 7.0 13.95 6.10 missed 6.10 missed

F 1 0.00 0.00 0.00 0.00 0.00 0.00

roof 28.0 28.30 28.1 28.60 27.7 28.60 27.35 28.35 missed missed missed missed

G

1 0.00 0.00 0.00 0.00 0.00 0.00

3 -1.10 -0.90 -0.90 -1.45 -2.1 -2.50

roof -0.60 0.40 -0.50 0.70 -0.40 0.60 -1.05 0.20 -3.55 0.10 -4.2 0.10

H 1 0.00 0.00 0.00 0.00 0.00 0.00

roof -1.40 -1.50 -1.70 -2.25 -3.60 -3.80 * The horizontal displacement readings in 15/12/1998 means either construction defects or horizontal

movements, therefore it was taken as a reference for the other following readings.


133

3.3 Geological formation

The building lies at middle plateau of Gebel Mokattam Figure (3.1). Gebel Mokattam

is located east of Cairo and is bounded on the north by Gebel Ahmer, on the east by a

gently sloping desert area leading to the New Cairo city, on the south by the Greater

Cairo ring road, and on the west by the autostrade and Salah Salem road. Gebel

Mokattam is a topographically high area that has a relief equal to about 160 meters. It

includes two main flat-topped areas locally referred to as the middle and upper

plateaus. Mokattam city was built on to upper plateau whereas the middle plateau is

the site of a large housing project that started in the 1980's and has not been completed

so far (Yousif, 2000). The exposed rocks of Gebel Mokattam have a total thickness of

166 meters Figure (3.5). These rocks are Middle and Late Eocene in age.

The stratigraphy of the Mokattam rocks was studied by several workers. The exposed

Middle Eocene rocks of Gebel Mokattam are 77 meters thick and are composed

mainly of white, hard, micritic, fossiliferous limestone with some marly limestone

beds. These Middle Eocene rocks are divided into three formations which are, from

base to top, the Gebel Hof Formation, the Observatory Formation, and the lower part

of the El Qurn Formation. The overlying Upper Eocene rocks have brown to yellowish

brown color and are 89 meters thick. They include the following formations, from

base to top, the Upper part of El Qurn Formation, the Wadi Garawi Formation, the

Wadi Hof Formation, and the Anqabia Formation (Yousif, 2000).

The Middle Eocene rocks of Gebel Mokattam form the middle plateau and its

bounding scarps. The latter are characterized by their overall white to grayish white

color and their resistance to erosion. On the other hand, the Upper Eocene rocks form

the slope and upper surface of the upper plateau and are generally less resistant to

erosion and form recessive slopes. The exposed rocks of G. Mokattam have a gentle

northeastward dip equal to about 3° (Yousif, 2000).


134

From a geotechnical point of view, the Middle Eocene rocks of the middle plateau of

G. Mokattam have a low-medium compressive strength equal to 40 MPa and a

Young's modulus equal to 5600 MPa (Moustafa et al., 1991).

Figure (3.5): Composite stratigraphic section of G. Mokattam (after El-Nahhas et al.,

1990).


135

3.4 Geotechnical Properties

A preliminary study was conducted in 1989 to present the stratigraphy and properties

of the soil at the site area. Twenty six boreholes each of depth 10 m were drilled at the

site before construction to evaluate the soil reactivity. Results of the study showed that

the soils at site are quietly complex and vary from place to place at the local scale.

A geotechnical group was retained in November 1999 as an expert for the owner

company to investigate the causes of cracking and to propose and implement

appropriate remedial repair work. At the time of the investigation, the house was

approximately 10 years old. The depth of foundation was about 2.50m below ground

surface, then a replacement soil layer (sand cushion), 1.20m consists of graded sand,

were placed and compacted on four layers. The relative compaction results of

replacement soil are shown in Table (3.3)

Table (3.3): Relative compaction of replacement soil for building (102)

Building Position Relative compaction ( % ) for layer

1 2 3 4

102 1 96.8 96.9 96.3 97.4

2 96.8 97.9 97.9 95.8

As shown in the table that the replacement soil is densely compacted, and this means

that there are no problems in the compaction process.

The investigation soil program includes drilling three boreholes of depth 10, 10 and 6

m respectively at the site to evaluate the soil properties, the boring locations are

illustrated in Figure (3.6). The geological soil profile at the site of borings is presented

in Table (3.4). It must be mentioned that borehole No. 1 & 2 have been drilled outside

the foundation area, while borehole No. 3 have been executed within the foundation

area (sand cushion).


136

The soil stratification at the site starts with a surface fill layer extended to a depth of

about 2.20m from the ground surface, as shown in Table (3.4). Then it is followed by

an expansive clay layer intercalated with some graded sand, tubes of gypsum crystals

and iron oxides. The expansive soil layer extended to a depth of about 5.70m below

ground surface. The expansive soil layer rests on a bed of dark yellowish calcareous

sandstone. Figure (3.7) shows photos for an open cut near the site and samples for that

expansive soil.

Table (3.4): The soil profile for the boreholes around the defected building:

Soil Profile

Description

Bore hole No.

1 2 3

FILL:

calcareous clayey silty graded sand,

some gravel & lime stone fragments

(0.00m) to (-1.00m) ---------- (0.00m) to (-1.00m)

SILT-CLAY:

calcareous, some graded loose sand

(yellow)

(-1.00m) to (-2.20m) (0.00m) to (-1.00m) ----------

SANDSTONE:

calcareous porous (yellow) ---------- (-1.00m) to (-2.20m) (-1.00m) to (-2.20m)

CLAY:

contains some silt, iron oxides &

gypsum (light brown to grey)

(-2.20m) to (-5.70m) (-2.20m) to (-5.70m) (-2.20m) to (-5.70m)

SANDSTONE:

calcareous porous (yellow) (-5.70m) to (-8.00m) (-5.70m) to (-8.00m)

CLAY:

contains some silt, iron oxides &

gypsum (light brown to grey)

(-8.00m) to (-9.00m) (-8.00m) to (-9.00m)

SANDSTONE: calcareous porous

(yellow) (-9.00m) to (-10.00m) (-9.00m) to (-10.00m)


137

The site investigation revealed that the building distortion could be attributed to soil

heave due to the inadequate provision of site drainage. There was evidence of water

leaking near the building and also the courtyard was not paved which make led it to be

affected by any surface water. Also it was found a number of trees planted close to the

buildings, their roots can reach under the buildings, sucking moisture and causing

excessive drying and shrinkage to soil.

The geotechnical investigation properties of the three expansive soil samples are

summarized in Table (3.5). The results of water content after water leakage were

approximately equal to plastic limit for the three expansive soil samples. This means

that the expansive soil in the site becomes saturated.

Table (3.5): Geotechnical properties of the existing clay:

Boring Number 1 2 3

Depth (m) 4 3 4

Soil Classification Clay Clay Clay

Natural unit weight (gm/cm3) 2.013 1.702 1.827

Clay Content (%) 84.13 85.36 88.32

Moisture Content after water leakage (%) 29.3 33.4 33.6

Liquid Limit (%) 95.6 84.8 92.90

Plastic Limit (%) 31.5 24.9 31.5

Shrinkage Limit (%) 14.9 12.5 15.2

Plasticity Index (%) 64.1 59.9 61.4

Activity= %PI/(CC-5) (according to Carter and Bentley,1991) 0.81 0.75 0.74


138

Figure (3.6): Schematic layout for the described building and boreholes, the defected

building is the shaded one.


139

`

Figure (3.7a): Photo for the expansive soil layer bearing on calcareous sandstone

layer

Figure (3.7b): Photo showing the shrink – swell effect for expansive soil layer.

Expansive soil layer

Shrink –Swell effect

Calcareous Sand Stone layer


140

Figure (3.7e): Undisturbed expansive soil sample before preparation

Figure (3.7f): Samples from the expansive soil layer showing the apparent

composition of this layer

Iron oxide

Gypsum crystals

Sand traces


141

3.4.1 Indirect measurements

Indirect estimation of swelling soil activity and swelling potential was performed

using charts and empirical formulas. Van der Merwe classification chart (Figure 3.8)

and the proposed chart according to Carter and Bentley, 1991(Figure 3.9) were used to

define the category of fine grained materials.


1964.


modified by Carter and Bentley, 1991.

Bore(2) depth(3m)

Bore(3) depth(4m)

Bore(1) depth(4m)

Bore(2) depth(3m)Bore(3) depth(4m)

Bore(1) depth(4m)


142

Figure (3.9) shows that the swelling potential equal to 25 % and is classified very high

swelling soil according to Van der Merwe, (1964) and Seed et al. (1960).

The results indicate that the soil is associated with clay, which could swell

considerably when wetted. The soils proved to have the ability to absorb and retain a

great deal of water and undergo significant volumetric changes with moisture

fluctuations (i.e. clay having high to very high swelling potential). The swelling

potential of expansive soil is discussed using empirical equations as shown in Table

(3.6).

Table (3.6): Classification of swelling potential of expansive soil samples as assessed

using empirical equations

Boring

type

Sample

depth (m)

Altmeyer

(1955)

Ranganatham and

Satyanarayana (1965)

Seed et al.

(1962)

Snethen

(1980)

Holtz et

al. (1973)

BRE

(1980)

Ghen

(1988)

Carter and

Bentley, (1991)

1 4 Not critical Very high High Very highModerate to

high

High to

Very high

High to

Very highVery high

2 3 Not critical Very high High Very high High High to

Very high

High to

Very highVery high

3 4 Not critical Very high High Very highModerate to

high

High to

Very high

High to

Very highVery high

The results illustrate the following comments on the existing clays:

i) All the classifications do not give the same assessment of the swelling or

shrinkage potential of the soils.

ii) The swelling of the examined soils is clearly apparent from all the

classifications and it's extremely high to very high swelling, except Altmeyer’s

classification which appears to underestimate the swelling potential.


143

3.4.2 Direct Measurements

The results of simple modified oedometer test according to Jenning et al., 1973

yielded a swelling pressure ranging from 3.5 to 6.5 kg/cm2 and swelling potential

varies from 7.5% to 9.8% as shown in Table(3.7) and Figure (3.10).

Table (3.7): Swelling & swelling pressure of expansive soil samples*:

Boring Number 2 3

Depth (m) 3 4

Swelling (%) at ( 1.03 Kg/cm2 ) = ∆e1+eo

7.5 9.8

Swelling pressure ( Kg/cm2 ) 3.5 6.5

* Stress at center line of expansive soil layer = Δσ + σo= 1.03 kg/cm2


according to Jenning et al., 1973.

eo=0.47σf

0.35

0.40

0.45

0.50

0.55

0.60

0.65

0.1 1 10 100

eo

Stress (kg/cm2)

Bore (2)

Bore (3)

`

σs3

`

σs2


144

The difference in swelling properties of expansive soil between the two samples can

be attributed to the difference in clay content, atterberg limits and the presence of sand

traces, gypsum and iron oxide with different ratios in the two samples.

3.5 Heave prediction

3.5.1 Heave prediction using semi-empirical heave equations

The final heave of expansive clay layer is predicted using two methods for the

expansive soil samples. The first method used is the Stress changes method

(According to Rama et al., 1988). In this method the pore water pressure was taken

equal to zero (according to Fredlund and Rahardjo, 1993) as the water was flowing

downward. The second method is the Suction changes method (According to

Hamberg, 1985). The predicted heave values are shown in Table (3.8) and Figure

(3.11).

Table (3.8): Predicted heave according to Hamberg, 1985 and Rama et al., 1988.

Bore

Bore (2) depth= 3.00m Bore (3) depth= 4.00m

eo ef WS.L WP.L eo ef WS.L WP.L

0.47 0.59 12.5 24.9 0.47 0.61 15.2 31.5

Cw1 0.01 0.01

∆w 28.40 28.60

Sp according to Rama et al., 1988 0.075 0.095

Sp according to Hamberg, 1985 0.287 0.287

H (m) thickness of layer 2.00 2.00

∆H (cm) Hamberg,1985 57.4 57.8

∆H (cm) Rama et al.,1988 15 19

Figure (3

The resul

utility and

The heav

through e

(2) and 19

the two bo

The heav

through e

for the tw

two boreh

can be at

leakage.

Pred

icted Heave (cm)

3.11): Com

lts of heav

d at the da

e predicti

xpansive

9 cm at b

oreholes l

ve predicti

expansive

wo boreho

holes locat

ttributed

0.00

10.00

20.00

30.00

40.00

50.00

60.00

70.00

Bo

S

mparison b

ve were p

ate of bore

on results

soil. The h

ore (3). A

ocations i

ion results

soil. The

les. Accor

tions is ap

to the di

ore (2)

Swelling per

between th

and Ra

predicted

eholes drill

s accordin

heave pre

Accordingl

s equal to

s accordin

total heav

rdingly th

pproximate

fference b

rformance

145

he predict

ama et al.

after leak

ling, this m

ng Rama e

diction fo

ly the rela

4 cm.

ng Hambe

ve predict

he relative

ely 0.4 cm

between w

B

of some exp

ted heave a

, 1988.

kage of w

means tha

et al., 198

or expansiv

ative heav

erg, 1985

tion for ex

e heave fo

m. The rea

water con

Bore (3)

Ha

Ra

pansive soil

according

water from

at the soil i

88 depend

ve soil is e

ve for exp

depends

xpansive

or expansiv

son of Ha

ntent befo

amberg,

ama et al

l treatment

g to Hamb

m a nearby

is under sw

ds on stres

equal 15 c

ansive soi

on suctio

soil is abo

ve soil be

amberg, 19

ore and a

1985

l, 1988

t techniques

berg, 1985

y drainage

welling.

ss changes

cm at bore

il between

n changes

out 58 cm

etween the

985 results

fter water

Average

s

5

e

s

e

n

s

m

e

s

r


146

The results of predicted heave according to Hamberg, 1985 is about 3.5 times the

value obtained using Rama et al., 1988 and this result agree with Abdel-Moatty, 1999.

3.5.2 Heave prediction using empirical heave equations

The swelling potential is predicted using empirical heave equations and is summarized

in Table (3.9) also the swelling pressure is predicted and summarized in Table (3.10).

Table (3.9): Comparison between the swelling potential according to Hamberg, 1985

(Spo) and the predicted swelling potential (Sp) for other different equations: (H=

2.00m thick)

Bore (2) (3)

Depth (m) 3.00 4.00

Swelling potential

according to Hamberg (1985) Spo (%) 29 29

Vijayvergiya and Ghazzaly (1973) Predicted swelling (%) 3.04 3.06

Sp / Spo 0.41 0.31

Weston (1980) Predicted swelling (%) 99.58 167.92

Sp / Spo 13.28 17.13

Schneider and Poor (1974) Predicted swelling (%) 4.E+09 7.E+09

Sp / Spo 5.33E+08 7.14E+08

Brackley et al. (1975) Predicted swelling (%) 89 89

Sp / Spo 11.87 9.08

Brackley et al. (1983) Predicted swelling (%) 67 69

Sp / Spo 8.93 7.04

Zacharais & Ranganathan (1972) Predicted swelling (%) 118.1 124.9

Sp / Spo 15.75 12.74

Nayak and Christensen (1974) Predicted swelling (%) 154 164.8

Sp / Spo 20.53 16.82


147


iii) All the empirical equations do not give the same assessment of the swelling

potential of the soils.

iv) All the empirical equations overestimate the swelling potential using oedometer

test, except Vijayvergiya and Ghazzaly (1973) equation which appears to

underestimate the swelling potential.

Table (3.10): Comparison between the swelling pressure using oedometer test and

predicted swelling pressure for different equations: Bore (2) (3)

Depth (m) 3.00 4.00

Swelling Pressure Psoed. (kg/cm2) 3.5 6.5

Popescu (1983) Predicted (Ps kg/cm2) 1.65 1.71

Pspred. / Psoed. 0.47 0.26

El Sohby & Mazen

(1987)

Predicted (Ps kg/cm2) 1.55 1.55


Dedier el al. (1973)

eqn(2)



El Sohby & Rabbaa

(1981)



Brackley et al. (1975) Predicted (Ps kg/cm2) 9.96 9.96


Brackley et al. (1983) Predicted (Ps kg/cm2) 14.01 14.95


Dedier el al. (1973)

eqn(1)



Zacharais &

Ranganathan (1972)



Nayak and Christensen

(1974)




148


v) All the classifications do not give the same assessment of the swelling pressure

of the soils.

vi) The predicted swelling pressure according to Popescu (1983), El Sohby &

Mazen (1987), Dedier el al. (1973) and Ranganathan (1972) underestimate that

using oedometer. While the swelling pressure results according to Brackley et

al. (1975), Brackley et al. (1983) and Nayak and Christensen (1974)

overestimate that using oedometer.

vii) The swelling pressure results according to El Sohby & Rabbaa (1981) and

Dedier el al. (1973) agree with that obtained from oedometer.

The results show that the above empirical equations (models) are someway useful to

predict the swelling properties of soils but cannot be precisely used for all types of

clay. There is a need to collect enough data in each area in order to develop specific

models for specific type of clay.

3.6 Measured building movements

Vertical displacement was plotted to determine the kind of movement and the state at

which the soil undergoes either heave or shrinkage with time. The positive

displacement values means settlement while the negative ones means heave. The

vertical displacement values are shown for different settlement marks distributed

along each front for the building as shown in Figures (3.12 to 3.15).

The vertical displacements for whole settlement points are shown in Figure (3.16), and

the relative displacement along each front is plotted in Figure (3.17). Horizontal

displacements for the building roof are presented in both directions (longitudinal and

transverse directions) as shown in Figure (3.18 and 3.19).

Figure (3

Figure (3

‐5

0

5

10

15

20

25

30

vertical displacem

ent (cm

)

2

3

vertical displacem

ent (cm

)

3.12): Ver

3.13): Ver

‐0.4

‐5

0

5

10

15

20

25

30

‐0

S

rtical displ

rtical displ

0

‐0.352

0.05‐0.

.13

‐0.180.2

Swelling per

lacement f

lacement f

0

.8 3.5

5.5.2

5.5

Measurin

2.4

3.6 4

5.10.185.6

Measurin

rformance

149

for the bui

for the bui

0

5.3

7.4

139.9

ng date

3.18

48.6

7.3

9.8

g date

of some exp

ilding fron

ilding fron

0

5.45

3.17 15.

19.4

3.2

9.18

15.418

20.7

pansive soil

nt (1) vers

nt (2) vers

0

5.45

.25 15.9

23.35 2

3.2

9.97

8.2 19.8

24.226

l treatment

sus measur

sus measur

A (1)

B (2)

C(3)D (

95

25.15

H (12 )

G (10,11

F (8,9E (6,

83

6.85

t techniques

ring date.

ring date.

(4,5)

1)

),7)

s

Swelling p

Figure (3

Figure (3

1

1

2

2

30

vertical displacem

ent (cm

)

‐

2

4

6

8

10

vertical displacem

ent (cm

)

erformance

3.14): Ver

3.15): Ver

‐5

0

5

10

15

20

5

0

0.2

‐2

0

2

4

6

8

‐0.4

e of some e

rtical displ

rtical displ

25.10.18‐0.2

4

0

‐0.1‐0.18

expansive s

lacement f

lacement f

7.35.6

5.5

Measurin

0

32

3.6

Measuri

soil treatme

150

for the bui

for the bui

15.49.8

2

9.9

ng date

0

2.4 3.

4

ng date

ent techniq

ilding fron

ilding fron

18.2

20.72419.4

0

18 3.

8.6

ques

nt (3) vers

nt (4) vers

19.8

4.2 26.823.35

0

.2 3

9.18

sus measur

sus measur

F (8,

E

83

8525.15

A (1)

H (

G

.2

9.97

ring date.

ring date.

,9)

(6,7)

D (4,5)

12 )

G (10,11)

Figure (

Figu

‐5

0

5

10

15

20

25

30Ve

rtical displacmen

t (cm

)

A

‐5

0

5

10

15

20

25

30

Rel

ativ

e di

spla

cem

ent (

mm

)

(3.16): Ver

ure (3.17):

(1) B (

S

rtical disp

: Relative

(2) C(3

Re

Swelling per

placement f

vertical d

me

3) D(4,

Measu

elative v

rformance

151

for the bu

date.

displaceme

asuring d

Measuring

,5) E (

uring date

vl. displ

of some exp

uilding ben

ent for the

date.

g date

6,7) F

lacemen

pansive soil

nch marks

e building f

(8,9)

nt

l treatment

s versus m

facades ve

G (10,11)

D‐A

H‐A

t techniques

measuring

ersus

A (1(

B (2(

C(3(

D(4,5(E (6,7(

F (8,9(

G (10,11(

H (12 (

H (12 )

E‐H

E‐D

s


152

Figure (3.18): Horizontal displacement at roof in the longitudinal direction relative to

the first reading versus measuring date.

Figure (3.19): Horizontal displacement at roof in the transverse direction relative to

the first reading versus measuring date.

24/07/1998

09/02/1999

28/08/1999

15/03/2000

01/10/2000

19/04/2001

05/11/2001

24/05/2002

10/12/2002

0 0.5 1 1.5 2 2.5 3 3.5 4

Measuring

date

Horizontal displacement (cm)

A B

C D

E F

G H

Boreholes drilling date

Average

Hz.

displacemen

t

Date at constant readings

FinalH

z. displacem

ent

01/11/1998

21/12/1998

09/02/1999

31/03/1999

20/05/1999

09/07/1999

28/08/1999

‐5 ‐4 ‐3 ‐2 ‐1 0 1 2 3 4 5

Measuring

date

Horizontal displacement (cm)

D

E

F

G


153

The observed results give the following comments:

• As shown in Figures (3.12 to 3.16) that all the readings were of positive values

which means that the building is undergoing settlement as mentioned before.

This also means that at the time of investigation, the building has completed its

heave, and all the deformation happening are due to water dissipation from soil.

• The large settlement values are of points E and D (front 3) and the smallest

values are for points A and H (front 4), this means that (front 3) is settling faster

than (front 4), and the decrease in moisture content below (front 3) is bigger

than (front 4) as shown in Figure (3.16).And this is compatible with the

presence of water leaking from a nearby drainage utility near (point A), as were

found in the investigation.

• According to Figure (3.17) the building inclination is towards the road, and the

main inclination is in the longitudinal direction as shown in the relative

displacement between points (D and A) on (front 1) and points (E and H) on

(front 2), while in the transverse direction the relative displacement between

points (H and A) on (front 4) and points (E and D) on (front 3) is small to make

problems to the building in this direction.

• All the vertical surveying lines in the longitudinal direction as in Figure (3.18)

were inclined in the same direction, unless some points were missed in the latest

readings but all the building tends to move away from the neighbor building,

and this is appeared in Figure (3.20). The last readings for vertical surveying

lines are approximately constant and tend to stop in some points.

• The vertical surveying lines in the transverse direction, Figure (3.19) have very

small displacements which vary from positive to negative values, and this can

be attributed to the allowable errors in surveying work.

• As shown in Figure (3.16) that the displacement of settlement points in the last

three readings tends to decrease and it stops in others, which means that the

building tends to return to its initial mode and began to gains its stability, and

this can be a sign in starting the repair works.

3

Swelling p

Figure (

3.7 Comp

comp

According

inclination

away from

drilling b

longitudin

predicted

measured

the longitu

A compar

displacem

direction

erformance

(3.20): Mo

parison b

onents:

g to the m

n of the b

m the neig

boreholes,

nal and tra

vertical

d tilting an

udinal and

rison was

ments, Tab

and comp

e of some e

ovement of

between t

measured r

building i

ghbor buil

, we can

ansverse d

displacem

ngle was ca

d transver

done betw

ble (3.11

ponents are

expansive s

of the appa

the measu

relative ve

s in the lo

lding. By

n get the

directions

ment using

alculated,

se directio

ween the m

). Schem

e shown in

soil treatme

154

arent defec

one.

ured and

ertical and

ongitudina

using Fig

e measur

at this tim

g (Hambe

then the h

ons were o

measured h

matic diag

n Figures

ent techniq

cted build

d predicte

d horizonta

al directio

gures (3.18

red horizo

me, accordi

erg, 1985

horizontal

obtained.

horizontal

grams for

(3.21 and

Significbetwee

ques

ding away f

ed horizo

al displace

on toward

8 and 3.19

ontal dis

ing to the

and Ram

l displacem

l displacem

r the bui

3.22)

cant Openn the two

from the a

ontal disp

ements, th

ds the adja

9) and at t

placemen

building h

ma et al, 1

ment comp

ments and

lding dis

ning o building

adjacent

placement

he induced

acent road

the date of

ts in the

height and

1988), the

ponents in

d predicted

placement

gs

t

d

d

f

e

d

e

n

d

t


155

Figure (3.21): Building rotation between bore (3) and bore (2).

Figure (3.22): Directions of horizontal displacement components.

G

E

F

AB

D


156

Table (3.11): Calculated horizontal displacement components using measured and

predicted values:

Data Measured Rama et al,1988 Hamberg,1985

Dist between D : A E : H Bore(3): Bore(2) Bore(3):Bore(2)

Distance (m) 19.64 19.64 11 11

Relative vertical disp. (Δvl) (cm) 2.40 2.30 4.08 (Δvlpred.) 0.4 (Δvlpred.)

Average horizontal disp. in Longitudinal direction at roof (cm)

Initial state date : Boring date 1.2* 1.2* ------------ 0.2 (ΔHzpred.)

Boring date : Final date 1.05* 1.05* 3.9 (ΔHzpred.) ------------

Average horizontal disp. in transverse direction at roof (cm)

0.0** 0.0** 5 (ΔHzpred.) 0.3 (ΔHzpred.)

* These values according to Figure (3.18)

** These values according to Figure (3.19)

In case of longitudinal direction, the building horizontal displacement direction

according to Rama et al, 1988 agree with the actual horizontal displacement direction

and the predicted value is about 4 times the measured value as shown in Table (3.11),

while by according to Hamberg, 1985 the predicted value is about 0.17 times the

measured value.

In case of transverse direction, the building horizontal displacement according to

Hamberg, 1985 is approximately equal to that measured value, Table (3.11).


157

Chapter 4

LABORATORY FOOTING MODELS RESTING ON UNTREATED

AND TREATED EXPANSIVE SOIL

4.1 General

Swelling behavior of shallow foundations rested on treated expansive soil is generally

affected by different factors. Some of these factors are related to the boundary

conditions controlling the site deposition and the mode of water migration, while the

others are related to the employed treatment technique for damping the heave potential

of the expansive soil. The main objective of this chapter is to investigate the

effectiveness of some widely used treatment techniques to eliminate or damp the

swelling behavior of expansive soil. The expansive soil used is classified using direct

and indirect tests. A laboratory testing program is designed to determine the swelling

behavior of shallow footings resting on treated expansive soil employing different

treatment techniques. The suggested experimental program consists of four groups,

Three cushion types are used in this chapter (medium sand cushion, medium sand

+5% hydrated lime cushion and expansive soil +5% hydrated lime cushion). The main

objective of this experimental program is to determine the effect of footing stress,

cushion type, polyethylene sheet location and water leakage space on the heave - time

relationship, moisture content profile and differential heave between footings. The

final heave of footings is predicted using two semi-empirical heave equations, Stress

Changes method (according to Rama et al., 1988) and Suction Changes method

(according to Hamberg, 1985). The measured and predicted footings heave are

compared and presented for different study factors.

4.2 Properties of expansive soil used

The geotechnical properties of the used expansive were determined using direct and

indirect methods and it was classified by different methods as shown in the following.


158

4.3 Indirect measurement tests results

Grain size characteristics and index properties of the expansive soil used (liquid limit,

plastic limit, shrinkage limit), clay content and free swell percent are summarized in

Table (4.1).

Table (4.1): Physical properties of the expansive soil used

Soil Property Value

Specific Gravity 2.73

Silt Size (%) 51.00

Clay Size (%) 49.00

Liquid Limit WLL 85.00

Plastic Limit WPL 34.32

Plasticity Index PI 50.68

Clay content% (<2μ) 49.00

Free swell (%) 102.5

The free swell test on the sample according to Holtz and Gibbs, 1956 indicated a value

about 103% (Table 4.1). This result indicates that the expansive soil used is associated

with clay, which could swell considerably when wetted.

The activity of the expansive soil is calculated using Skempton, 1953 and Cartel et al.,

1991 equations, the results are summarized in Table (4.2). The clay mineral of

expansive soil is identified according to the Casagrande A line chart, (Figures 4.1).

Swelling potential is obtained using Van de Merwe, 1964 chart (Figures 4.2), Seed, et

al., 1960 chart (Figure 4.3).

Table (4.2

Activ

Accordi

Valu

Classific

Figure

2): Classif

vity

ing to

ue

cation

e (4.1): Cl

S

fication of

Skem

lassificatio

Swelling per

f the expan

mpton (1953

1.03

Active

on of expa

rformance

159

nsive soil

3)

ansive soil

of some exp

used

l mineral u

pansive soil

Cartel et a

1.1

Very a

using Casa

l treatment

al., (1991)

15

active

agrande’s

t techniques

s chart.

s


160


1964.

Figure (4.3): Classification of expansive soil according to Seed et al., 1960 and

modified by Carter and Bentley, 1991.


161

The clay mineral of expansive soil is illite according to the Casagrande A line chart,

(Figures 4.1). The results indicate that the expansive soil used is high to very high

swelling according to Seed et al. (1960), Van der Merwe (1964) and Cartel et al.

(1991), swelling potential is equal to 25% according to Cartel et al. (1991). The

activity of expansive soil is ranging from active to very active according to Skempton,

(1954) and Cartel et al. (1991) equations.

The classification of swelling potential (Sp) is discussed by different classifications,

Table (4.3).

Table (4.3): Swelling potential of expansive soil used as assessed by different

empirical equations

Soil

property

Altmeyer

(1955)

Ranganatham and

Satyanarayana (1965)

Seed et al.

(1962)

Snethen

(1980)

Holtz et al.

(1973) BRE (1980)

Ghen

(1988)

Swelling

potential Low Very high High

Very

high Very high

Moderate

to

Very high

Very

high

The results of soil classification show that;

i) All the classifications do not give the same assessment of the swelling or

shrinkage potential of the soils.

ii) The swelling of the examined soils is clearly apparent from all the

classifications and it's extremely high to very high swelling, except Altmeyer’s

classification which appears to underestimate the swelling potential.

Accordingly the expansive soil used proved to have the ability to absorb and retain a

great deal of water and undergo significant volumetric changes with moisture

fluctuations (i.e. clay having high to very high swelling potential).

Swelling p 4.4 Dire

The swel

according

Figure (4

The resul

yielded a

swelling p

While in c

14 kg/cm2

Figure (4

pressure o

cementitio

with that

M. M., 19

eo

0.8

0.9

1

1.1

1.2

1.3

1.4

1.5

0.

Voi

d ra

tio (

e)

erformance

ect measu

lling prop

g to Jennin

(4.4): Resu

lts of sim

a swelling

potential a

case of ex2 and swel

4.4). Acco

of expansi

ous mater

proposed

979 and A

1

e of some e

rement te

perties ar

ng et al. (1

ults of exp

ac

mple mod

g pressure

at (0.5 Kg/

xpansive s

lling poten

rdingly u

ive soil; th

rials are p

by El-Ra

Abdel-Hady

expansive s

ests result

re determ

973), Figu

ansive soi

ccording to

ified oedo

e for untr

/cm2) of a

soil treated

ntial at (0.

using lime

his is due

produced

ayes, M. K

y, 2007.

1

soil treatme

162

ts

mined usi

ure (4.4).

il samples

o Jenning

ometer te

eated exp

approxima

d with 5%

.5 Kg/cm2

content

to pozzola

which ha

K., Hindi,

Stress

ent techniq

ing simpl

using sim

et al., 197

est accord

pansive so

tely 16.6 %

% lime, the2) of appro

in stabiliz

anic reacti

ave high v

A. O., R

s (kg/cm2)

E

E

σ

ques

le modifi

mple modif

73.

ding to Je

oil of abo

% as show

e swelling

oximately

zation inc

ions in wh

volume in

Radwan, A

10

Expansive

Expansive

σs

ied oedom

fied oedom

enning et

out 8.5 kg

wn in Figu

pressure i

16.5 % as

creases the

hich new c

ncrease, th

A. M. and

e soil + 5

e soil

meter test

meter test

al., 1973

g/cm2 and

ure (4.4).

is equal to

s shown in

e swelling

crystalline

his agrees

El-Sharif

100

5% lime

t

3

d

o

n

g

e

s

f,

0


163

4.5 Laboratory model test

The test experiments are conducted on square model footings (F1 and F4) of 3 cm side

dimensions and 1 cm thick as well as circular footings (F2 and F3) of 3 cm diameter

and 1 cm thick. The model footings were inserted in a laboratory model container of

inside dimensions 49 cm by 25 cm, and 25 cm in height. To prevent tilting of the

model footing upon loading and during the course of the test, the model footing was

attached to a cylindrical solid steel rod (1.30 cm in diameter), which was screwed into

the model footing’s center of gravity. This rod was greased and passed through a

circular hole (1.4cm in diameter) in a steel plate that was fixed to the sides of the

metal box container. The details of the different parts of the mould and the test model

setup are shown in Figure (4.5).

4.6 Test procedure

The footing heave is measured using dial gauge with sensitivity of 0.02 mm per

division. The specimens are subjected to water leakage at foundation level. The

specimen thickness is equal to 2.67 footing width for untreated and treated expansive

soils. Thickness of soil replacement is equal to 0.67 footing width. The test procedure

is as following; i) eight soil specimens are prepared in the testing mould from the

expansive soil used, ii) for each test, a specific weight is placed in the mould in four

equal layers (the replacement cushion is compacted to a density equal to the expansive

soil density), iii) the specimens are compacted to thickness equal 8 cm for both

untreated and treated soils, iv) the location of footings are shown in Figure (4.6), v)

the initial immediate settlement corresponding to the footing stress 0.5 kg/cm2 is

recorded, vi) water is added to the soil surface and allowed to flow downward through

the soil specimen, vii) the footing heave at different time intervals is recorded, viii) the

test is terminated when the rate of heave become nearly zero and, ix) at the end of

each test, the moisture content profile is determined.


164

Figure (4.5): Testing mould details.


165

4.7 Laboratory test program

The suggested experimental program consists of four groups. The main objective of

this experimental program is to determine the effect of footing stress, cushion type,

polyethylene sheet location and water leakage space on the heave- time relationship

and moisture content profile. Three cushion types are used (medium sand cushion,

medium sand +5% hydrated lime cushion, and expansive soil +5% hydrated lime

cushion). Properties of hydrated lime are shown in Table (4.4) and the suggested

experimental program is presented in Table (4.5).

As the plastic sheet is used as a preserving technique to prevent water leakage to soil,

the main aim of using plastic sheet is measuring the differential heave between

footings in case of deterioration of this plastic sheet. Accordingly it was suggested to

make a hole in the plastic sheet used at different spaces and allow water to leak

through this hole and then measure the induced differential heave between different

footings, Figure (4.6). The location of plastic sheet was varied to see the effectiveness

of replacement soil layer in distributing the moisture in case of plastic sheet at top of

replacement layer and in case of plastic sheet at bottom of replacement soil layer.

Table (4.4): Properties of medium sand-lime mixes

A summary of the laboratory test program is shown in the following flow chart

Lime content (%) 0 5

Max. dry density γdmax. (kN/m3) 16.67 19.77

Optimum Moisture Content O.M.C. (%) 6.8 7.85

Degree of improvement of sand compaction (%) - 18.6

Coefficient of permeability K (cm/sec) 0.00175 0.00162

Permeability degree (Hausmann, 1990) Med.

permeability Med.

permeability


166


167

Figure (4.6): A schematic cross-section in the used mould


168

Tabl

e (4

.5):

Lab

orat

ory

mod

elin

g te

st p

rogr

am

Leakage

space from

footing (F1)

‐ ‐ ‐ 0 0 0

10 cm

20 cm

Location

of

polyethylene

sheet

‐ ‐

No sheet

Top of sand

Bottom

of

sand

Bottom

of

sand

Footing

stress q

(kg/ cm2)

F1 & F2=0,

F3 & F4=0.5

F1 & F2=0,

F3 & F4=0.5

F1 & F2=0 ,

F3 & F4=0.5

F1 & F2=0 ,

F3 & F4=0.5

Thickn

ess of

treated

layer(cm

)

‐ ‐

2 cm

2 cm

2 cm

2 cm

2 cm

Thickn

ess of

expa

nsive

soil layer

8 cm

8 cm

6 cm

6 cm

6 cm

6 cm

6 cm

Cushion Type

‐

Untreated

Med

ium Sand

Med

ium

Sand

+5% lime

Clay cushion

‐5%

Lime

Stabilizatio

n

Med

ium Sand

Med

ium Sand

Stud

y Factor

Untreated

expansive

soil with

differen

t stress

Treatin

g expansive

soil using differen

t

cushions

Treatin

g expansive

soil using sand

cushion and plastic

sheet (differen

t locatio

ns)

Treatin

g expansive

soil using sand

cushion and plastic

sheet (differen

t Leakage spaces)

Test

No.

i.1

ii.1

ii.2

ii.3

ii.4

iii.1

iii.2

iii.3

iv.1

iv.2

iv.3

Group

No.

G1

G2

G3

G4


169

4.8 Test results and discussions

4.8.1 Measured heave results

The results of measured footings heave versus wetted time for different groups are

shown in Appendix (A). The final moisture content beneath footings versus depth is

shown in Appendix (B).

4.8.1.1 Untreated expansive soil

The measured footings heave of untreated expansive soil versus wetted time for

variable footing stress are shown in Figure (4.7). The Final measured footing heave

and final moisture content are summarized in Table (4.6) and Table (4.7), respectively.

The differential heave between footings are presented in Figure (4.8) and summarized

in Table (4.8).

The results show that;

(a) The heave of footings increase as the time increases until a certain time, then it

remains constant, Figure (4.7).

(b) The final measured heave below loaded footings is less than unloaded footings,

Figure (4.7).

(c) The final moisture content of expansive soil used is approximately lies between the

liquid limit, Table (4.7).

(d) The final moisture content below loaded footings is less than unloaded footings,

Table (4.7).

(e) The differential heave between footings (∆) increases to a certain time then, it

decreases until it becomes nearly constant for different loading conditions, Figure (4.8)

and Table (4.8). The ratio (∆max/ ∆final) is equal to 1.3 and 1.4 for loaded and unloaded

footings respectively.


170

(f) No significant difference for the differential heave between loaded and loaded

footings and that between unloaded and unloaded footings for untreated expansive

soil, Figure (4.8) and Table (4.8). And this can be attributed to the small difference in

footings stresses.

Figure (4.7): Footings heave versus logarithmic time for variable footing stress

Figure (4.8): Differential heave between footings versus log time.

F1

F2F3

F4

0

5

10

15

20

25

30

1 10 100 1000 10000 100000

∆h (m

m)

Time (min)

Untreated expansive soil

q1=0.5 kg/cm2q2=0.0q3=0.0q4=0.5 kg/cm2

F1 F2 F3 F4

Loaded (0.5 Kg/cm2) Loaded (0.5 Kg/cm2)

Untreated Expansive soil

0.00

5.00

10.00

15.00

20.00

1 10 100 1000 10000 100000

∆ (m

m)

Time (min)


Loaded (F1‐F4)

Unloaded (F2‐F3)

F1 F2 F3 F4




171

Table (4.6): Final measured footings heave (∆Hf ) untreated expansive soil:

Method of Treatment

Description Results

Footing number F1 F2 F3 F4

Footing stress (kg/cm2) 0.5 0.0 0.0 0.5

Untreated ∆Hf (cm) 1.82 2.62 2.82 2.02

Tf (min) 18720 18720 18720 18720

Table (4.7): Moisture content beneath footings for untreated expansive soil:

Method of Treatment

Description Moisture Content (Wc %)



Untreated

Top of swelling layer at depth(cm) 0 83.00 84.00 82.00 84.00

Middle of swelling layer at depth(cm) 4 82.00 83.00 81.00 83.00

Bottom of swelling layer at depth(cm) 8 75.00 75.00 74.86 74.21

Table (4.8): Differential heave between footings (∆) for untreated expansive soil:

Method of Treatment Description Results

Diff. heave between F1 ‐ F4 F2 ‐ F3

Untreated

∆max(cm) 2.60 2.80

Time at ∆max(min) 2880 2880

∆final(cm) 2.00 2.00

Time at ∆final(min) 18720 18720

∆max/ ∆final 1.30 1.40


172

4.8.1.2 Treated expansive soil using various treatment techniques

The measured footings heave for various treatment techniques versus wetted time are

shown in Figure (4.9) for loaded and unloaded footings. The Final measured footing

heave and final moisture content are summarized in Table (4.9) and Table (4.10)

respectively. The differential heave between footings are presented in Figure (4.10)

and summarized in Table (4.11).


a) The final heave of footings in case of treated expansive soil with medium sand

cushion is nearly equal to that of treated expansive soil with medium sand

cushion +5% lime and expansive soil cushion with 5% lime, Figure (4.9) and

Table (4.9).

b) The final heave of unloaded footings in case of untreated expansive soil is

reduced by about 30% in case of using medium sand cushion, medium sand

cushion with 5% lime or expansive soil cushion with 5% lime as treatment

techniques, Figure (4.9) and Table (4.9).

c) The final heave of loaded footings in case of untreated expansive soil is reduced

by about 25% in case of using medium sand cushion, medium sand cushion with

5% lime or expansive soil cushion with 5% lime as treatment techniques, Figure

(4.9) and Table (4.9).

d) A slight change in the final moisture content of expansive soil takes place for

various treatment techniques and they all lies between liquid limit and plastic

limit, Table (4.10).

e) The differential heave between footings increases slightly to a certain time then,

it may slightly decreases until it becomes nearly constant for different loading

conditions and various treatment techniques, Figure (4.10) and Table (4.11).

In case of treated expansive soil using sand cushion, the ratio (∆max/ ∆final) is

equal to 1.22 for loaded footings and 1.10 for unloaded footings, while in case


173

of treated expansive soil using sand +5% lime cushion, the ratio (∆max/ ∆final) is

equal to 1.22 for loaded footings and 1.5 for unloaded footings.

Accordingly the change in (∆max/ ∆final) is not significant for the two cases. Also

these results show that the sand and sand +5% lime cushion are permeable and

the relative movement between footings is very small.

Figure (4.9a): Footing heave versus log time for different treatment techniques

Figure (4.9b): Footing heave versus log time for different treatment techniques

0

5

10

15

20

25

30

1 10 100 1000 10000 100000

∆h (m

m)

Time (min)

Footing (1)

Untreated

Medium sand

Medium sand+5% lime

Expansive soil+5% lime

N.B; * Footing stress= 0.5 Kg/cm2

* Thickness of replacement soil= 2 cm

0

5

10

15

20

25

30

1 10 100 1000 10000 100000

∆h (m

m)

Time (min)

Footing (4)

Untreated

Medium sand

Medium sand+5% lime





174

Figure (4.9c): Footing heave versus log time for different treatment techniques

Figure (4.9d): Footing heave versus log time for different treatment techniques

0

5

10

15

20

25

30

1 10 100 1000 10000 100000

∆h (m

m)

Time (min)

Footing (2)

Untreated

Medium sand

Medium sand+5% lime


N.B; * Footing stress= 0.0* Thickness of replacement soil= 2 cm

0

5

10

15

20

25

30

1 10 100 1000 10000 100000

∆h (m

m)

Time (min)

Footing (3)

Untreated

Medium sand

Medium sand+5% lime




175

Figure (4.10a): Differential heave between loaded footings versus log time for

different treatment techniques, q=0.5 Kg/cm2

Figure (4.10b): Differential heave between unloaded footings versus log time for

different treatment techniques, q=0.0

0.00

5.00

10.00

15.00

20.00

1 10 100 1000 10000 100000

∆ (m

m)

Time (min)

Loaded Footings (F1 ‐ F4)

Untreated

Medium sand

Medium sand+5%lime

Expansive soil+5%lime

0.00

5.00

10.00

15.00

20.00

1 10 100 1000 10000 100000

∆ (m

m)

Time (min)

Un Loaded Footings (F2 ‐ F3)

Untreated

Medium sand

Medium sand+5%lime

Expansive soil+5%lime


176

Table (4.9): Final measured footings heave (∆Hf) results:

Method of Treatment

Description Heave Values (cm)



Untreated

∆Hf(cm) 1.82 2.62 2.82 2.02

Tf (min) 18720 18720 18720 18720

Medium Sand (replacement soil)

without plastic sheet

∆Hf(cm) 1.61 1.92 2.02 1.52

Tf (min) 23040 23040 23040 23040

Medium Sand + 5% lime (replacement soil)


∆Hf(cm) 1.52 2 2.1 1.61

Tf (min) 23040 23040 23040 23040

Expansive soil + 5% lime (replacement soil)


∆Hf(cm) 1.6 1.95 2.00 1.5

Tf (min) 20160 20160 20160 20160


177

Table (4.10): Moisture content beneath footings for various treatment techniques:

Method of Treatment




Untreated

Top of swelling layer at depth(cm)

0 83.00 84.00 82.00 84.00

Middle of swelling layer at depth(cm)

4 82.00 83.00 81.00 83.00

Bottom of swelling layer at depth(cm)

8 75.00 75.00 74.86 74.21

Medium Sand (replacement soil)



2 63.93 74.04 76.00 68.04


5 57.92 73.00 75.00 66.00


8 57.00 71.00 72.00 65.00

Medium Sand + 5% lime (replacement soil) without plastic

sheet


2 65.00 74.04 76.00 68.04


5 63.00 75.00 77.00 66.00


8 62.00 73.00 75.00 65.00

Expansive soil + 5% lime (replacement soil) without plastic

sheet


2 65.00 75.00 73.00 64.00


5 63.00 73.04 71.09 61.81


8 62.00 72.00 70.00 61.00


178

Table (4.11): Differential footings heave (∆) for various treatment techniques:



Untreated

∆max(cm) 2.60 2.80


∆final(cm) 2.00 2.00


∆max / ∆final 1.30 1.40

Medium Sand (replacement soil) without plastic sheet

∆max(cm) 1.10 1.10


∆final(cm) 0.90 1.00


∆max / ∆final 1.22 1.10

Medium Sand + 5% lime (replacement soil) without

plastic sheet

∆max(cm) 1.10 1.50


∆final(cm) 0.90 1.00


∆max / ∆final 1.22 1.50

Expansive soil + 5% lime (replacement soil) without

plastic sheet

∆max(cm) 1.10 2.40


∆final(cm) 0.90 1.00


∆max / ∆final 1.22 2.40


179

4.8.1.3 Treated expansive soil using various horizontal barrier locations

The measured footings heave for various horizontal barrier locations versus wetted

time are shown in Figure (4.11) for variable footing stress. The Final measured footing

heave and final moisture content are summarized in Table (4.12) and Table (4.13)

respectively. The differential heave between footings are presented in Figure (4.12)

and summarized in Table (4.14).


a) The final heave of footings for different sheet locations reaches approximately

the same value at the end of the experiment for treated expansive soil with

medium sand cushion, Figure (4.11) and Table (4.12). Although the plastic

sheet used on top of sand cushion can be subjected to damage at any point but

the flow of water seeping to the expansive soil is uniformly distributed resulting

in uniform heave and eliminating differential movement between footings,

Figure (4.11) and Table (4.12).

b) On the other side, using of plastic sheet at bottom of sand cushion concentrates

the leakage at definite point which results in high differential movement

between footings as shown in Figures (4.12a and 4.12b).

c) The moisture content is higher near the leakage area than far areas, which causes

these areas to heave faster than other zones. And this causes differential heave

between footings, Figure (4.12) and Table (4.13).

d) Also on comparing the results of differential heave between footing in case of

using plastic sheet at bottom of sand cushion and at the top of sand cushion for

the same leakage point, it was found that when the differential heave between

footings reaches the peak value (in case of plastic sheet at bottom of sand

cushion) it starts to decrease again until it reaches the permanent (constant)

differential heave value, this final differential heave value is nearly the same

value reached in case of using the plastic sheet at top of sand cushion, as shown

in Figure (4.12) and Table (4.14).This means the following;


180

i) In case of treated expansive soil using sand cushion without horizontal

barrier, the ratio (∆max/ ∆final) is equal to 1.22 for loaded footings and 1.10

for unloaded footings.

ii) In case of treated expansive soil using sand cushion and horizontal barrier

at the top of cushion, the ratio (∆max/ ∆final) is equal to 2.20 for loaded

footings and 3.86 for unloaded footings.

iii) In case of treated expansive soil using sand cushion and horizontal barrier

at the bottom of cushion, the ratio (∆max/ ∆final) is equal to 22.29 for

loaded footings and 20.37 for unloaded footings.

Accordingly the using of plastic sheet at bottom of sand cushion results in

a very high differential heave than using the plastic sheet at top of sand

cushion, on which it could affect the stability of buildings severely.

iv) The differential heave values using plastic sheet at the bottom of sand

cushion takes a long time to be reached than the differential heave values

using plastic sheet at top, accordingly leakage can happens but it could

takes long time until the resulted heave or problem is detected.

v) If water infiltrates to expansive soil due to leakage from drainage utilities

or water pipes near buildings, the value of differential heave depends on

the leakage location in expansive soil. Accordingly when leakage

happens below the replacement layer this can be another form- similar to

plastic sheet at bottom of sand cushion- for the harmful effects due to

concentrating the leakage at definite points in the expansive soil which

causes high differential movements between footings as discussed in the

case study (Chapter 3). Therefore it is recommended to install any water

sources or drainage utilities at the top or through replacement layer to

help in uniform distribution of seeping water through soil.

e) The rate of increase of differential heave incase of plastic sheet at bottom of

sand cushion is very high than that of plastic sheet at bottom of sand cushion


181

and this causes very big movement in short time for buildings, and this explains

the high terror that possess the defected building inhabitants (Chapter 3).

Figure (4.11a): Footing heave versus log Time for various plastic horizontal barrier

locations

Figure (4.11b): Footing heave versus log Time for various plastic horizontal barrier

locations

0

5

10

15

20

25

1 10 100 1000 10000 100000

∆h (m

m)

Time (min)

Footing (1)

no sheet

sheet at replacement top

sheet at replacement bottom



0

5

10

15

20

25

1 10 100 1000 10000 100000

∆h (m

m)

Time (min)

Footing (4)

no sheet






182

Figure (4.11c): Footing heave versus log Time for various plastic horizontal barrier

locations

Figure (4.11d): Footing heave versus log Time for various plastic horizontal barrier

locations

0

5

10

15

20

25

1 10 100 1000 10000 100000

∆h (m

m)

Time (min)

Footing (2)

no sheet




0

5

10

15

20

25

1 10 100 1000 10000 100000

∆h (m

m)

Time (min)

Footing (3)

no sheet





183


various plastic horizontal barrier locations, q=0.5 Kg/cm2


various plastic horizontal barrier locations, q=0.0

0.00

5.00

10.00

15.00

20.00

1 10 100 1000 10000 100000

∆ (m

m)

Time (min)


no sheet



0.00

5.00

10.00

15.00

20.00

1 10 100 1000 10000 100000

∆ (m

m)

Time (min)

Unloaded Footings (F2 ‐ F3)

no sheet




184


Method of Treatment





∆Hf(cm) 1.61 1.92 2.02 1.52

Tf (min) 23040 23040 23040 23040

Medium Sand (replacement soil), plastic sheet at top of

sand

∆Hf(cm) 1.6 1.95 2.00 1.5

Tf (min) 20160 20160 20160 20160

Medium Sand (replacement soil) plastic sheet below sand

∆Hf(cm) 1.61 2.004 1.95 1.54

Tf (min) 40320 46080 51840 59040

Table (4.13): Moisture content beneath footings for various horizontal barrier locations:

Method of Treatment








Medium Sand (replacement soil), plastic

sheet at top of sand




Medium Sand (replacement soil) plastic sheet below

sand





185

Table (4.14): Differential footings heave (∆) for various horizontal barrier locations:




∆max(cm) 1.10 1.10


∆final(cm) 0.90 1.00


∆max / ∆final 1.22 1.10

Medium Sand (replacement soil), plastic sheet at top of

sand

∆max(cm) 2.20 2.02


∆final(cm) 1.00 0.52


∆max / ∆final 2.20 3.86

Medium Sand (replacement soil) plastic sheet below sand

∆max(cm) 15.60 11.00


∆final(cm) 0.70 0.54

Time at ∆final(min) 59040 59040 ∆max / ∆final 22.29 20.37

4.8.1.4 Treated expansive soil using various leakage spacing:

The measured footings heave versus wetted time is shown in Figure (4.13) for various

leakage spacing. The Final measured footing heave and final moisture content are

summarized in Table (4.15) and Table (4.16) respectively. The differential heave

between footings are presented in Figure (4.14) and summarized in Table (4.17).


a) The final heave of footings for various leakage spaces reaches the same values

at the end of the experiment for treated expansive soil with medium sand

cushion, Figure (4.13). This result means that the expansive soil zones don’t

start heaving until moisture reaches these zones, as the footings far from leakage

points starts heaving later than the nearer ones and so it reaches its final heave

values after it. Accordingly this causes differential heave of footings during the

moisture distribution and heaving time.


186

b) The differential heave between footings increases until it reaches the peak value

after a certain time then it starts to decrease again, as shown in Figure (4.14).

c) The maximum differential heave between footings increases as leakage space

decreases and the ratio (∆max / ∆final) is equal to 22.29 for loaded footings and

20.37 for unloaded footings in case of leakage beneath F1. While in case of

leakage 10 cm far from F1, the ratio (∆max/ ∆final) is equal to 14.71 for loaded

footings and 8.21 for unloaded footings.

d) In case of leakage 20 cm far from F1, the ratio (∆max/ ∆final) is equal to 22.29

for loaded footings and 20.37 for unloaded footings.

This means that, when the leakage point is beneath footing F1, it gives

maximum differential heave between F1 and F4. As the leakage point moves

away from F1 towards F4, the differential heave decreases until it becomes

zero, then it reverses into the other direction and starts to increase again,

Figure (4.14).

Figure (4.13a): Heave of footing 1 versus logarithmic time for various leakage

spaces

0

5

10

15

20

25

1 10 100 1000 10000 100000

∆h (m

m)

Time (min)

Footing (1)

Leakage below F1

10 cm from F1

20 cm from F1

F1 F2 F3 F4

Loaded Loaded

Unloaded Unloaded

Swelling soil

Medium SandPoint of Leakage

20cm

10cm




187

Figure (4.13b): Heave of footing 4 versus logarithmic time for various leakage

spaces

Figure (4.13c): Heave of footing 2 versus logarithmic time for various leakage

spaces

0

5

10

15

20

25

1 10 100 1000 10000 100000

∆h (m

m)

Time (min)

Footing (4)

Leakage below F1

10 cm from F1

20 cm from F1

F1 F2 F3 F4

Loaded Loaded

Unloaded Unloaded

Swelling soil


20cm

10cm



0

5

10

15

20

25

1 10 100 1000 10000 100000

∆h (m

m)

Time (min)

Footing (2)

Leakage below F1

10 cm from F1

20 cm from F1

F1 F2 F3 F4

Loaded Loaded

Unloaded Unloaded

Swelling soil


20cm

10cm



188

Figure (4.13d): Heave of footing 3 versus logarithmic time for various leakage

spaces


various leakage spacing, q=0.5 Kg/cm2

0

5

10

15

20

25

1 10 100 1000 10000 100000

∆h (m

m)

Time (min)

Footing (3)

Leakage below F1

10 cm from F1

20 cm from F1

F1 F2 F3 F4

Loaded Loaded

Unloaded Unloaded

Swelling soil


20cm

10cm


0.00

5.00

10.00

15.00

20.00

1 10 100 1000 10000 100000

∆ (m

m)

Time (min)


Leakage below F1

10 cm from F1

20 cm from F1


189


various leakage spacing, q=0.0


Method of Treatment




Medium Sand (replacement soil), plastic sheet below sand,

leakage below F1

∆Hf(cm) 1.61 2.004 1.95 1.54

Tf (min) 40320 46080 51840 59040


leakage 10 cm from F1

∆Hf(cm) 1.61 2.11 20.04 15.4

Tf (min) 48960 38880 48960 67680


leakage 20 cm from F1

∆Hf(cm) 1.54 2.00 2.11 1.61

Tf (min) 66240 47520 41760 51840

0.00

5.00

10.00

15.00

20.00

1 10 100 1000 10000 100000

∆ (m

m)

Time (min)

Unloaded Footings (F2 ‐ F3)

Leakage below F1

10 cm from F1

20 cm from F1


190

Table (4.16): Moisture content beneath footings for various leakage spacing:

Method of Treatment




Medium Sand (replacement soil), plastic sheet below sand, leakage below

F1




Medium Sand (replacement soil), plastic sheet below sand, leakage 10 cm

from F1





from F1




Table (4.17): Differential footings heave (∆) for various leakage spacing:

Group Test No. Method of Treatment Description Results


G4

iv‐1

Medium Sand (replacement soil), plastic sheet below

sand, leakage below F1

∆max(cm) 15.60 11.00


∆final(cm) 0.70 0.54


∆max / ∆final 22.29 20.37

iv‐2


from F1

∆max(cm) 10.30 8.70


∆final(cm) 0.70 1.06


∆max / ∆final 14.71 8.21

iv‐3


from F1

∆max(cm) 11.30 6.00


∆final(cm) 0.70 1.06


∆max / ∆final 16.14 5.66


191

4.8.2 Predicted heave results

The final heave of footings is predicted using two semi-empirical heave methods. The

first method used is the Stress Changes method (according to Rama et al., 1988). The

other method is the Suction Changes method (according to Hamberg, 1985). The

predicted final heave and the ratio between measured from laboratory tests and

predicted final heave values for different study factors are summarized in Table (4.18

to 4.21).

The predicted heave results of the laboratory model test show that:

a) The predicted heave values using Suction Changes method (according to

Hamberg, 1985) gives higher values than that of the Stress Changes method

(according to Rama et al., 1988).

b) The ratio between the predicted heave of footings using Suction Changes

method and that using Stress Changes method ranged between ( 3.58 to 5.16 )

for loaded footings, and (3.62 to 4.82) for unloaded footings.

c) The predicted footings heave using Suction Changes method (according to

Hamberg, 1985) ranged from 2.40 to 3.86 times the measured heave in case of

loaded footings while it ranged from 2.19 to 2.70 times the measured heave in

case of unloaded footings.

This means that the predicted heave using Suction Changes method is about 3

times the measured heave and this agrees with Abdel-Moatty, 1999.

d) The predicted footings heave using Stress Changes method (according to Rama

et al., 1988) ranged from 0.67 to 0.74 times the measured heave in case of

loaded footings while it ranged from 0.52 to 0.58 times the measured heave in

case of unloaded footings.


192

This means that the predicted heave using Stress Changes method is about half

the measured heave.

Table (4.18): The final predicted footings heave for untreated expansive soil:

Treatment method




Untreated

∆H(cm), Measured 1.82 2.62 2.82 2.02

∆H(cm), Hamberg, 1985 7.26 7.33 7.19 7.30

∆H(cm), Rama et al., 1988 1.41 1.52 1.52 1.41

∆HHamberg / ∆Hmeasured 3.99 2.80 2.55 3.61

∆HRama / ∆Hmeasured 0.78 0.58 0.54 0.70

∆HHamberg / ∆HRama 5.14 4.81 4.72 5.16


193

Table (4.19): The final predicted footings heave for various treatment techniques:

Treatment method




Untreated

∆H(cm), Measured 1.82 2.62 2.82 2.02

∆H(cm), Hamberg, 1985 7.26 7.33 7.19 7.30

∆H(cm), Rama et al., 1988 1.41 1.52 1.52 1.41


∆HRama / ∆Hmeasured 0.78 0.58 0.54 0.70

∆HHamberg / ∆HRama 5.14 4.81 4.72 5.16

Medium Sand (replacement

soil)

∆H(cm), Measured 1.61 1.92 2.02 1.52

∆H(cm), Hamberg, 1986 3.99 4.92 5.04 4.47

∆H(cm), Rama et al., 1989 1.06 1.10 1.10 1.06


∆HRama / ∆Hmeasured 0.66 0.57 0.55 0.70

∆HHamberg / ∆HRama 3.76 4.47 4.58 4.21

Medium Sand + 5% lime

(replacement soil)

∆H(cm), Measured 1.52 2 2.1 1.61

∆H(cm), Hamberg, 1989 4.25 5.02 5.16 4.47

∆H(cm), Rama et al., 1992 1.06 1.10 1.10 1.06


∆HRama / ∆Hmeasured 0.70 0.55 0.52 0.66

∆HHamberg / ∆HRama 4.01 4.55 4.68 4.21

Expansive soil +5% lime

(replacement soil)

∆H(cm), Measured 1.6 1.95 2.00 1.5

∆H(cm), Hamberg, 1990 4.25 4.97 4.83 4.18

∆H(cm), Rama et al., 1993 1.06 1.10 1.10 1.06


∆HRama / ∆Hmeasured 0.66 0.56 0.55 0.71

∆HHamberg / ∆HRama 4.01 4.51 4.38 3.94


194

Table (4.20): The final predicted footings heave for various horizontal barrier

locations:

Treatment method





∆H(cm), Measured 1.61 1.92 2.02 1.52

∆H(cm), Hamberg, 1986 3.99 4.92 5.04 4.47

∆H(cm), Rama et al., 1989 1.06 1.10 1.10 1.06


∆HRama / ∆Hmeasured 0.66 0.57 0.55 0.70

∆HHamberg / ∆HRama 3.76 4.47 4.58 4.21


soil), plastic sheet at top of sand

∆H(cm), Measured 1.6 1.95 2.00 1.5

∆H(cm), Hamberg, 1986 3.80 4.85 4.67 3.57

∆H(cm), Rama et al., 1989 1.06 1.10 1.10 1.06


∆HRama / ∆Hmeasured 0.66 0.56 0.55 0.71

∆HHamberg / ∆HRama 3.58 4.40 4.24 3.37


soil) plastic sheet below sand

∆H(cm), Measured 1.61 2.004 1.95 1.54

∆H(cm), Hamberg, 1987 4.71 4.39 3.99 3.70

∆H(cm), Rama et al., 1990 1.06 1.10 1.10 1.06


∆HRama / ∆Hmeasured 0.66 0.55 0.56 0.69

∆HHamberg / ∆HRama 4.44 3.99 3.62 3.49


195

Table (4.21): The final predicted footings heave various leakage spacing:

Treatment method





soil), plastic sheet below sand,

leakage below F1

∆H(cm), Measured 1.61 2.004 1.95 1.54

∆H(cm), Hamberg, 1987 4.71 4.39 3.99 3.70

∆H(cm), Rama et al., 1990 1.06 1.10 1.10 1.06


∆HRama / ∆Hmeasured 0.66 0.55 0.56 0.69

∆HHamberg / ∆HRama 4.44 3.99 3.62 3.49


soil), plastic sheet below sand, leakage 10 cm

from F1

∆H(cm), Measured 1.61 2.11 2.004 1.54

∆H(cm), Hamberg, 1987 4.45 4.92 4.63 4.04

∆H(cm), Rama et al., 1990 1.06 1.10 1.10 1.06


∆HRama / ∆Hmeasured 0.66 0.52 0.55 0.69

∆HHamberg / ∆HRama 4.20 4.47 4.21 3.81


soil), plastic sheet below sand, leakage 20 cm

from F1

∆H(cm), Measured 1.54 2.004 2.11 1.61

∆H(cm), Hamberg, 1988 4.20 4.68 4.89 4.50

∆H(cm), Rama et al., 1991 1.06 1.10 1.10 1.06


∆HRama / ∆Hmeasured 0.69 0.55 0.52 0.66

∆HHamberg / ∆HRama 3.96 4.25 4.44 4.25


196

Chapter 5

SUMMARY AND CONCLUSIONS

5.1 Summary

Several research projects have been conducted in Egypt in recent years to investigate the

fundamental mechanisms of heaving of treated and untreated expansive soils and the

influence of geological and climatic conditions on swelling potential. A review for all

the available literature related to the topic foundation on the expansive soil and

employed treatment techniques used for damping the swelling phenomenon of

expansive soil were presented. Also, the methods used to evaluate the heave

associated with wetting and the swelling characteristics before and after treatment.

A case study of a severely cracked reinforced concrete building constructed on treated

expansive soil using medium sand replacement at site in the middle Mokattam plateau

was presented. Geological formation was summarized and geotechnical properties

were investigated for the studied area which represents the northeast extension of the

new communities around the Greater Cairo. The building was founded on shallow

footings and plain concrete raft resting directly on treated expansive clay using sand

replacement. The field investigation revealed that the damage was caused by

differential heaving as a result of water leakage below replacement soil layer. The

movement of the building was simulated depending on six survey readings during a

time period reached about four years nearly after the building inclination. Also the

horizontal displacement components were obtained for the building (using two semi

empirical methods) and compared with the measured values to decide the best and

more suitable method which can be used in predicting footings heave.

A Laboratory testing program was preformed to evaluate and analyze the swelling

behavior and differential heave induced incase of some usually using treatment

techniques for expansive soil. The geotechnical properties of the expansive soil used


197

were determined using direct and indirect methods and classified by different

empirical and semi empirical methods. The used treatment techniques were medium

sand, medium sand + 5% lime, clay + 5% lime stabilization and plastic horizontal

barrier. The effect of footings stresses, type of treatment technique, horizontal barrier

location and leakage space below footings were taken in to consideration. The heave

movement, moisture content distribution and differential heave movements were

monitored for different loading cases and different treatment techniques. The final

predicted heave of footings is predicted using two semi-empirical heave methods. The

predicted heave values were summarized and compared with those measured in the

laboratory model.

5.2 Conclusions

5.2.1 Case study

The geotechnical properties of the expansive soil layers indicated that the existing

layers are cohesive and highly plastic due to the presence of clay minerals (active to

very active) and with high to very high swelling potential. The predicted heave for the

expansive clay layer according to Hamberg, 1985 is about two times the values

obtained using Rama et al., 1988 equation.

All the vertical displacement readings were emphasizing that the building is

undergoing settlement. This means that at the time of investigation the building has

completed its heave, and all the deformation happening are due to water dissipation

from soil. The building inclination was in the longitudinal direction, while in the

transverse direction the relative displacement was minor to make problems to the

building in this direction. The displacement of settlement points in the last three

readings tends to decrease and it stops in others, which means that the building tends

to return to its initial mode and began to gains its stability, and this can be a sign in

starting the repair works.


198

In case of longitudinal direction, the direction of the horizontal displacement

component for the building agree with that obtained from Rama et al, 1988 and the

predicted value is about 4 times the measured value, while according to Hamberg,

1985 the predicted value is about 0.17 times the measured value. In case of transverse

direction, the building horizontal displacement according to Hamberg, 1985 is

approximately equal to that measured value.

Although the expansive soil layer were treated with sand replacement but it failed in

preventing the harmful effects due to the differential heaving of soil, and this can be

attributed to the presence of water seeping from a nearby drainage utility below the

replacement layer.

The case study emphasized the importance of site management for buildings on

expansive soil sites, although the causes of the house distortion were out of the footing

designer’s control, but less attention to maintain a reasonably uniform state of subsoil

moisture around the buildings could have decreased the existing defects. And this was

clear with the presence of water leaking from a nearby drainage utility below the

replacement soil level, and over the long term, the moisture infiltration has

exacerbated swelling damage to the structure as we have.

Finally this research project has helped to identify the expansive soils and associated

problems in the area and to increase the sensitivity of the proposed measures, which

can help in mitigating the structural damages originating from the behavior of

expansive soils. This awareness can lead to a very positive development in terms of

ensuring the durability of the properties - for a problem occasionally repeated - in this

area. Also it may helps in increasing the potential of improving the safety of the

communities by assisting homeowners and designers in promoting proper design,

positive construction and maintenance altitudes.


199

5.2.2 Experimental results

The expansive soil used in the laboratory model is active to very active due to the

presence of clay minerals (illites) and high to very high swelling potential.

5.2.2.1 Measured results

(a) All the used treatment techniques in the laboratory model test succeed in

decreasing the final heave with nearly the same value.

(b) The using of expansive soil layer stabilized with 5% lime succeeds in the earlier

times but with passing of time it gives higher heave values in little time and this

can cause problems.

(c) The heave of footings increase as the time increases until a certain time, then it

remains constant.

(d) The final heave and final moisture content distribution below loaded footings is

less than unloaded footings.

(e) The final moisture content of soil specimen at the end of the test lies between

liquid limit and plastic limit.

(a) The differential heave between footings increases to a certain time then, it

decreases until it becomes nearly constant for different loading conditions and

various treatment techniques.

(f) The final heave of footings in case of treated expansive soil with medium sand

cushion is nearly equal to that treated with medium sand cushion +5% lime and

expansive soil cushion with 5% lime.

(g) The final heave of unloaded footings for untreated expansive soil is reduced by

about 30% in case of using various treatment techniques.

(h) The final heave of loaded footings for untreated expansive soil is reduced by

about 25% in case of using various treatment techniques.

(i) In case of plastic horizontal sheet at the top of replacement layer; the

replacement layer succeeds in distributing the moisture uniformly over the

expansive soil layer, and this result in uniform heave of footings with time.


200

While in case of water seeping from the plastic sheet used at bottom of sand

cushion, this concentrates the leakage at definite point which results in high

differential movement between footings.

(j) The moisture content is higher near the water leakage area than far areas, which

causes these areas to heave faster than other zones. And this causes differential

heave between footings.

(k) The final heave and maximum differential heave values using plastic sheet at

the bottom of sand cushion takes a long time to be reached than that using

plastic sheet at top, Accordingly leakage can happens but it could takes long

time until the resulted heave or problem is detected, also when it is detected the

rate of change of differential heave is very high in a very short time which can

cause terror for any building inhabitants, also this give uncertain feeling for the

inspecting engineer and always conflict happens between specialists about the

type of differential movement and kind of cracks; is it due to heave or

settlement? Until this sudden heave happens to cut this uncertainty.

5.2.2.2 Predicted heave

(a) The predicted heave values using Suction Changes method (according to

Hamberg, 1985) gives higher values than Stress Changes method values

(according to Rama et al., 1988).

(b) The predicted footings heave using Suction Changes method (according to

Hamberg, 1985) ranged from 2.40 to 3.86 times the measured heave in case

of loaded footings while it ranged from 2.19 to 2.70 times the measured

heave in case of unloaded footings.

This means that the predicted heave using Suction Changes method is about

3 times the measured heave and this agrees with Abdel-Moatty, 1999.

(c) The predicted footings heave using Stress Changes method (according to

Rama et al., 1988) ranged from 0.67 to 0.74 times the measured heave in


201

case of loaded footings while it ranged from 0.52 to 0.58 times the measured

heave in case of unloaded footings.

This means that the predicted heave using Stress Changes method is about

half the measured heave.

5.3 Recommendations

In view of the results obtained in this study, the following recommendations can

be stated:

Any ground drainage utilities or water sources shall be placed at top or through the

replacement layer- if there -to help in distributing moisture uniformly through the soil.

Suggest structure systems able to induce equal stresses beneath various footings for

the same building to decrease the effect of differential heave between footings due to

various loading conditions.

In case of using plastic horizontal barriers, due care shall be applied during the barrier

installation to protect it from further deterioration during installation or in the future.

Also the plastic horizontal barrier should be placed at the top of sand cushion incase of

treating expansive soil with replacement layer in order to eliminate the inducing of

differential heave due to water leakage from the sheet.

5.4 Recommended Future Studies

Develop analytical and finite element models that may help practitioner engineers to

adequately design civil infrastructure on this type of soil in case of using horizontal

moisture barriers. This rational design method should be able to quantify the heave or

subsidence of the soil associated with the suction changes during water diffusion, as

well as the contact pressures on soil-structure interfaces. Continue the work to take the

effect of settlement of samples after stopping the water leakage source.


202

REFERENCES

1. Abdel-Hady, H. (2007). “Properties of treated expansive soil”. 12th Int.

Colloquium on Structural and Geotech. Eng., Faculty of Engineering, Ain

Shams University, Struct. Eng. Deprt., GTE-40, 8 pp.

2. Abdel-Moatty, N. A., (1999). “Geotechnical properties of some swelling

deposits in new communities around the Greater Cairo”. M.Sc. thesis, Faculty

of Engineering, Ain Shams University, 176pp.

3. Abdullah, W. S., (2002). “Bidimensional Swell Effect on Accuracy of Footing

Heave Prediction,” Geotechnical Testing Journal, GTJODJ, Vol. 25, No. 2,

June, pp. 177–186.

4. Abouleid, A. F. and Reyad, M. M. (1985)."Study of the behavior of isolated

footings on expansive soils”. Proc. of the 11th Int. Conf. on Soil Mech. and

Found. Eng., San Francisco, pp. 965-967.

5. Ahmed, A. A. (1985), "Study on the behavior of some Egyptian expansive

soils” M.Sc. thesis, Faculty of Engineering, Ain Shams University.

6. Ahmed, A. A. (2000),”Response behavior of shallow foundation on expansive

soil due to wetting” Sci. Bull. Fac. Eng. Ain Shams University, v. 35, No. 4,

Dec. 31, pp 1-18.

7. Aitchison, G. D. and Martin, R. (1973). “ A membrane oedometer for

complex stress-path studies in expansive clays”, Proc. 3rd Int. Conf on

Expansive Soils, Haifa, Israel, Vol 2, Jul ,pp 83-88.

8. Aitchison, G. D., Peter, P., and Power, R. F. (1977). ‘‘Stress-deformation

mechanisms in clays as determinants of design requirements for small and large

structures—An evaluation through case histories in Adelaide, of the relative


203

significance of several mechanisms.’’ Proc., Symp. On How to Live with

Geotechnical Risks, pp 26–33. (Quoted by Li and Cameron, 2002).

9. Al-Rawas, A. A., Taha, R., Nelson, J. D., Al- Shab, T. B., and Al-Siyabi, H.,

(2002), “A Comparative Evaluation of Various Additives Used in the

Stabilization of Expansive Soils,” Geotechnical Testing Journal, GTJODJ, Vol.

25, No. 2, June, pp. 199–209.

10. Altmeyer, W. T. (1955). ‘‘Discussion of engineering properties of expansive

clays.’’ Proc., ASCE, 81 (separate no. 658), 17–19.

11. Awad, T. (1993), “Effect of expansive soil on imposed stress of footing and

supper structure”. Ph. D. Thesis Ain Shams University, 239 p.

12. Awad, T. (2005), “Properties of expansive soil treated with sand lime cushion”.

J. Egyptian Geotech. Soc., Vol. 16, pp. 1-12.

13. Awad, T. and Abdel-Hady, H. (2005). “Movement of laboratory footing

model resting on treated expansive soil”. 11th Int. Colloquium on Structural and

Geotech. Eng., Faculty of Engineering, Ain Shams University, Struct. Eng.

Deprt., GE-30, 10 pp.

14. Anderson, J.U., Elfadil, F.K., and O’Connor, G.A., (1973), “Factors affecting

the coefficient of linear extensibility in Vertisols”, Soil Science Society of

America Journal, 37,298-299.

15. ASTM (American Society for Testing and Materials) (1988). “Standard Test

Methods for Expansion Index of Soils Designation D4829-88”, Annual Book of

ASTM Standards, Section 4, Construction, Vol.04.08, Soil and Rock (I),

American Society for Testing and Materials, Philadelphia, 1995, pp. 896-899.


204

16. ASTM, (1993), “Standard Specification for Fly Ash and Raw or Calcined

Natural Pozzolan for Use as a Mineral Admixture in Portland Cement

Concrete”, Annual Book of ASTM Standards, C 618-93, Vol. 04-02, pp. 310-

312.

17. ASTM, (1993), “Standard Practice for Characterizing Fly Ash for Use in Soil

Stabilization”, Annual Book of ASTM Standards, D 5239-92, Vol. 04- 08, pp.

1277-1279.

18. Bailey, S.W. , (1980). “Summary of recommendation of AIPEA nomenclature

committee”. Clays Clay Min. 28:73-78.

19. Basma, A. A., Al-Rawas, A. A., Al-Saadi, S. N., and Al-Zadjali, T. F.,

(1998), “Stabilization of Expansive Clays in Oman,” Environmental and

Engineering Geoscience, Vol. IV, No. 4, pp. 503–510. (Quoted by Al-Rawas et

al., 2002).

20. Bekkouche A., Djedid A., Aissa Mamoune S.M. (2001), “Identification and

prediction of the swelling behavior of some soils from the Tlemcen region of

Algeria” , Bulletin des laboratories des Ponts et Chaussées - 233 - July-August -

Ref. 4375 - PP. 69-77.

21. Brackley, I. J. A. (1975). “Swell Under Load”. Proc. 6th Reg. Conf. for Africa

on SMFE, 1, pp. 71-79. Durban.

22. Brackley, I.J.A., Steffen, R., and Kirsten (1983). “An Empirical Equation for

the Prediction of Clay Heave”. Proceedings 7th Asian Regional Conference on

Soil Mechanics and Foundation Engineering, Volume 1, p8-14.


205

23. Burland, J. B., (1962)."The Estimation of Field Effective Stresses and the

Prediction of Total Heave Using a Revised Method of Analyzing the Double

Oedometer Test," The Civil Engineer in South Africa, Vol 4, No.7, Jul,pp 133-

137.

24. Carter, M., and S.P. Bentley. , (1991). “Correlation of soil properties”.

Pentech Press, London. (Quoted by Lucian, 2006).

25. Casagrande, A. (1948). “Classification and identification of soils”. Trans. Am.

Soc. Civil Eng. 113:901- 931.

26. Çetiner, S. I., (2004), ““Stabilization of expansive soils by cayirhan fly ash and

desulphogypsum” M.S., Department of Civil Engineering, the Middle East

Technical University, January, 107 pages.

27. Chen, F.H. 1988. Foundations on Expansive Soils. Elsevier Science Publishers,

Amsterdam, the Netherlands, 463 pp.

28. Chu, S. C. and Kao, H. S., (1993), “A Study of Engineering Properties of a

Clay Modified by Fly Ash and Slag,” Proceedings, Fly Ash for Soil

Improvement, American Society of Civil Engineers, Geotechnical Special

Publication, No. 36, pp. 89–99.

29. Collins, R. J. and Ciesielski, S. K., (1994), “Production and Use of Industrial

Wastes, Synthesis of Highway Practice 199—Recycling and Use of Waste

Materials and By-Products in Highway Construction”, Transportation Research

Board, National Research Council, Washington, D.C., Chapter 3. (Quoted by

Al-Rawas, et al., 2002).

30. Colmenares, J.E. (2002). ”Suction and volume changes of compacted sand-

bentonite mixtures”. MSc Dissertation. Imperial College of Science,

Technology and Medicine. University of London.


206

31. Croney, D., Coleman, J.D. and Black W.P.M (1958). “Movement and

distribution of water in soil relation to highway design and performance”.

National Research Council, Highway Research Board, Special Rep. 40,

Washington, D.C., pp. 226–252. (Quoted by Abdel-Moatty, 1999).

32. Dawson, R.F. (1953). “Movement of small houses erected on an expansive clay

soil”. Proc. 3rd Intl. Conf. Soil Mech. Found. Eng. 1:346-350.

33. Day, R. W. (1997). ‘‘Soil-related damage to tilt-up structures.’’ Pract. Period.

Struct. Des. Constr., 2(2), pp 55–60. (Quoted by Li and Cameron, 2002).

34. Didier, G., Lareal, P. and Gielly J. (1973), “Prediction of swelling potential

and swelling pressure of soils”. Proc. 8th Int. Conf. SMFE, Vol. 2.2, pp. 67-72.

35. Diyaljee, V. A., and Wiens, B. (1995). “The use of edge membranes in airport

pavement rehabilitation.” Proc., 2nd Int. Congress on Unsaturated Soils, Paris,

959–966.

36. Eades, J.L. and Grim, R.E. (1966). “A quick test to determine lime

requirements for lime stabilization”. Highway Research Record 139, Highway

Research Board-National Research Council, Washington D.C., 99. 61-72.

(Quoted by Hardcastle, 2003).

37. El-Khoraibi, M.C. , Tarek, M.F. and Hussein, S.A.(1991).”Interpretation of

swelling pressure using soil suction characteristics.” Proc. 2nd Int. Eng. Conf.

Al-Azhar, pp. 174-185.

38. El-Nahhas, F., Mostafa, A.R. and Abdel-Tawab, S., (1990), "Geotechnical

characteristics of the lime stone formations of Gebel Mokattam area" 1st

Alexandria Conf. on Structural and Geotech. Eng., Vol. 1, pp. 9-19.


207

39. El-Nahhas, F., Awad, T. and Abd-Moatty, N., (1998), "Prediction of ground

heave due to swelling of multi layered soil deposits" 8th Int. Colloquium on

Structural and Geotech. Eng., Faculty of Engineering, Ain Shams

University,Struct. Eng. Deprt., 12 pp.

40. El-Hoseiny, K. E., Youssef, A. A. and Samaan, E. H. (2005). "Improving of

expansive soil behavior using different techniques". 11 th Int. Colloquium on

Structural and Geotech. Eng., Ain Shams Univ., Faculty of Eng., Struct. Eng.

Dept., GE-15. 10 pp.

41. El-Kadi, F. and Awad, T. (1998), "Foundation failure due to swelling of soil

deposit -case study". 8 th Int. Colloquium on Structural and Geotech. Eng., Ain

Shams Univ., Faculty of Eng., Struct. Eng. Dept., pp. 108-125.

42. El-Sabaie, A. R., (1992) "Behaviour of footings resting on expansive soils

using model tests" M. Sc. Thesis, Ain Shams University, 188p.

43. El-Sohby, M.A., and Rabba, E.A., (1981), “Some Factors Affecting Swelling

of Expansive Clayey soils”, Geotechnical Engineering Vol. 12, pp. 19-39.

44. El-Sohby, M.A., Rabba, E.A., and Aboutaha, M. M., (1995), “On prediction

of swelling and collapsing potential of arid soils”, Proc. XI African Reg. Conf.

on Soil Mech. and Found. Eng. Cairo, vol. 2, no. pp. 122-144.

45. El-Sohby, M.A., and Mazen, O., (1987), “On prediction of Swelling pressure

and deformational behavior of of Expansive Soils”, Proc. Of the 9th Reg. Conf.

for Africa.

46. Fang, H. Y., (1991), Foundation Engineering Handbook, 2nd ed., Van Nostrand

Reinhold Publishing Company. (Quoted by Al-Rawas, et al., 2002).


208

47. Fargher, P. J., Woodburn, J. A. and Selby J. (1979). “Footings and

foundations for small buildings in arid climates”. Inst. of Eng. Aust., S. A. Div.

and Univ. Adelaide Adult Ec. Dep., Adelaide, Australia. (Quoted by Fredlund,

and Rahardjo, 1993).

48. Firth, R. A., (1971), "Personal communication and discussion as a postgraduate

student". (Quoted by Awad, 1993)

49. Fredlund, D. G.,(1979). ”Second Canadian Geotechnical Colloquium

Appropriate Concepts and Technology for Unsaturated Soils,” Can. Geotech. J.,

vol. 16, no. pp. 121-139. (Quoted by Fredlund, and Rahardjo, 1993).

50. Fredlund, D. G.,(1983).”Prediction of ground movement in swelling clay “. 3

Est Annual Soil Mech. And Found. Eng. Conf. of Minnesota, Minneapolis, Feb.

(Quoted by Nelson, and Miller, 1992).

51. Fredlund, D. G., and Rahardjo, H. (1993). “Soil mechanics for unsaturated

soils“, Wiley, New York.

52. Gill, W.R. and C.A. Reaves. (1957). “Relationships of Atterberg limits and

Cation-exchange capacity to some physical properties of soil”. Soil Sci. Soc.

Am. Proc. 21:491-494.

53. Gromko, G. J., (1974), “Review of Expansive Soils,” Journal of the

Geotechnical Engineering Division, American Society for Civil Engineers,

ASCE 100, GT6, pp. 667–687.

54. Grossman, R.B., B.R. Brasher, D.P. Franzmeier, and J.L. Walker., (1968).

‘Linear extensibility as calculated from natural-clod bulk density

measurements’. Soil Sci. Soc. Am. Proc. 32:570-573.


209

55. Hamberg, D.J., (1985), “A simplified method for predicting heave in expansive

soils”, M.S. thesis, Colorado State University, Fort Collins, CO, USA. (Quoted

by Covar and Lytton, 2001).

56. Hamilton, J. J. (1977), Foundations on swelling or shrinking sub-soils,

Canadian Building Digest, http://irc.nrc-cnrc.gc.ca/pubs/cbd/cbd184_e.html,

accessed on 11th September 2007.

57. Hammitt, G. M. and Ahlvin, R. G. (1973).”Membranes and encapsulation of

soils for control of swelling”, Proc. of workshop Proceedings of Workshop on

Expansive Clay and Shale in Highway Design and Construction ,D. R. Lamb

and S. J. Hanna, Federal Highway Administration, Denver, CO., pp. 80-95.

(Quoted by Nelson and Miller, 1992).

58. Hardcastle, J.H., (2003)."Evaluation and treatment of expansive volcanic

soils”. US95. Final Report to Idaho Transportation Department, Department of

civil Engineering, University of Idaho, Moscow, 88 pp.

59. Hisham, H. H. and Safwat, A. S. (1998). "Effect of using sea water on

engineering properties of lime treated expansive soil". J. Egyptian Geotech.

Soc., Vol. 9, pp. 35-63.

60. Holt, J.H., (1969). A study of physicochemical, mineralogical and engineering

index properties of fine grained soils in relation to their expansive

characteristics. Ph.D. Dissertation, Texas A&M Univ., College Station, TX.

(Quoted by Hardcastle, 2003).

61. Holtz, Wesley G. and Harold J. Gibbs. (1956). Engineering properties of

expansive clays. Trans. ASCE 121:641-677.

62. Holtz, W. G., (1959). “Expansive clays- properties and problems”. Quart.

Colorado School of Mines, 54(4), 89-117.


210

63. Hunter, D. (1988). ‘‘Lime-induced heave in sulfate-bearing clay soils.’’ J.

Geotech. Eng., 114(2), 150–167.

64. Ingles, O.G. and Metcalf, J.B., (1972), Soil Stabilization, Butterworths,

Sydney. (Quoted by YEŞİLBAŞ , 2004).

65. Jennings, J. E. B. and Knight, K. (1975), "The Prediction of Total Heave from

the Double Oedometer Test," Symposium on Expansive Clays, The South

African Institute of Civil Engineers, , pp 13-19.

66. Jennings, J. E. and Kerrich, J. E., (1962)."The Heaving of Buildings and the

Associated Economic Consequences, with Particular Reference to the Orange

Free State Goldfields," Civil Engineer in South Africa, Vol 4, No. 11, Nov, pp

221-248. (Quoted by Nelson and Miller, 1992).

67. Jennings, J. E., Firtu, R. A., Ralph, T. K. (1973), "An Improved Method for

Predicting Heave Using the Oedometer Test," Proceedings, Third International

Research and Engineering Conference on Expansive Clay Soils, Haifi, Israel,

Vol 2, Aug, pp 149-154.

68. Johnson, L. D. (1977). ‘‘Evaluation of laboratory suction tests for prediction of

heave in foundation soils.’’ Technical Rep. S-77-7, U.S. Army Engineers

Waterways Experiment Station, Vicksburg, Miss.

69. Komornik, J. and David, A. (1969). “Prediction of swelling potential for

compacted clays”. Journal of Soil Mech. and Found. Eng. Div. ASCE, Vol. 95,

No. 1, pp 209-225.

70. Ladd, C. C. and Lambe, T. W., (1961) "The Identification and Behavior of

Compacted Expansive Clays," Proceedings, Fifth International Conference on

Soil Mechanics and Foundation Engineering, Parish, Vol 1, pp 201-205.


211

71. Lee, J. J., and Kocherhans, J. G. (1973). ‘‘Soil stabilization by use of

moisture barriers.’’ Proc., 3rd Int. Conf. on Expansive Soils, 295–300.

72. Li, J. and Cameron D. A., (2002). " Case Study of Courtyard House Damaged

by Expansive Soils” Journal of Performance of Constructed Facilities, Vol. 16,

No. 4, November 1, pp 169–175.

73. Little, D. N. (2000). “Evaluation of structural properties of lime stabilized soils

and aggregates”. Vol. 3: Mixture design and testing protocol for lime-stabilized

soils, National Lime Association, Arlington, Va. (Quoted by Petry and Little,

2002).

74. Little, D. N., and Petry, T. M. (1992). “Review of Stabilization of Clays and

Expansive Soils in Pavements and Lightly Loaded Structures—History,

Practice, and Future”. Journal of Materials in Civil Engineering, Vol. 14, No. 6 ,

PP. 447–460.

75. Little, D. N., Males, E. H., Prusinski, J. R., and Stewart, B. (2000).

‘‘Cementitious stabilization.’’ 79th Millenium Rep. Series, Transportation

Research Board, Washington, D.C.

76. Lucian, C. (2006), “Geotechnical Aspects of Buildings on Expansive Soils in

Kibaha, Tanzania Preliminary Study”, Licentiate Thesis, Division of Soil and

Rock Mechanics, KTH, Sweden.

77. Lytton, R. L. (1977). ‘‘Foundations in expansive soils.’’ In numerical methods

of geotechnical engineering, chapt. 13, Desai C.S. and Christian J.T. (Eds.) ,

New York: McGraw-Hill, pp. 427–458. (Quoted by Fredlund and Rahardjo,

1993).


212

78. Marie, N., El-Ashaal, A., and Abdel-Motaleb, A.(2000).”Improvement of

expansive soils in waterways using different replacement cushions”. J. Egyptian

Geotech. Soc., Vol. 11, pp.1-10.

79. McCallister, L.D. and Petry T.M., (1988), “Physical Property Changes in a

Lime-Treated Expansive Clay Caused by Leaching”, Transportation Research

Record 1295, pp. 37-44. (Quoted by YEŞİLBAŞ , 2004).

80. McKeen, R.G. (1981).“ Design of Airport Pavements for Expansive Soils:

Final Report , DOT/FAA/RD-81/25, U.S. Dept. of Transportation, Federal

Aviation Administration, Systems Research and Development Service,

Washington, DC. (Quoted by Abdel-Moatty, 1999).

81. McKeen, R.G. (1992). “A model for predicting expansive soil behavior”.

Proceedings, Seventh International Conference on Expansive Soils, Dallas,

Texas, pp. 1-6.

82. McKeen, R.G. and Hamberg, D.I. (1981). “Characterization of expansive

soils”. Shales and Swelling soils, Transportation Research Record 790, National

Academy of Sciences, Washington, D.C., pp. 73-78. (Quoted by Covar. and

Lytton, 2001).

83. Mitchell, J. K. ,(1976). “Fundamentals of soil behavior”, 1st Ed., Wiley, New

York.(Quoted by Abdel-Moatty,1999).

84. Mitchell, J.K. (1986). “Practical problems from surprising soil behavior”.

Journal of Geotechnical Engineering, Vol. 112, No. 3, American Society of

Civil Engineers, pp. 259-289. (Quoted by Hardcastle, 2003).

85. Mitchell, J.K. ,(1993).” Fundamentals of Soil Behavior”. 2nd. Edition. John

Wiley & Sons, Inc., New York, 473 pp. (Quoted by Hardcastle, 2003).


213

86. Mohan, D. and Rao, B. G. (1965).”Moisture variation and performance of

foundations in black cotton soils in India”. Moisture equilibrium and moisture

changes in soils beneath covered areas. Symposium, Soil Mechanics section,

Commonwealth scientific and industrial research organization . Butterworth’s,

Australia, pp. 175-183. (Quoted by Nelson and Miller, 1992).

87. Morgan, S. and Halff, A. H., (1984), “Oil Sludge Solidification Using Cement

Kiln Dust,” Journal of Environmental Engineering, American Society for Civil

Engineers, ASCE, Vol. 10, No. 5.

88. Moussa, A. A., Abed-Miguid, M. A., Okdah, S. M., and Heikal, A. H.

(1985). "Effect of sand cushion on swelling and swelling pressure of silty clay”.

Proc. of the 11th Int. Conf. on Soil Mech. and Found. Eng., San Francisco, pp.

1023-1026.

89. Moustafa, A.R., El-Nahhas, F., and Abdel Tawab, S., (1991), “Engineering

geology of Mokattam city and vicinity, eastern Greater Cairo, Egypt”:

Engineering Geology, v. 31, p. 327-344.

90. Najder J. and Werno M. (1973). “Protection of Buildings on Expansive Clays.

Proceedings 3-rd International Conference on Expansive Soils, Vol. I, Haifa, 1:

325-334.

91. Nayak , N.V. , and Christensen, D. (1974). “Swelling Characteristics of

compacted expansive soils”. Journal of Clays and Clay Minerals, Vol. No.4,

pp.251-261. (Quoted by Abdel-Moatty, 1999).

92. Nelson, J. D., and Miller, D. J. ,(1992). Expansive soils, problems and practice

in foundation and pavement engineering, Wiley, New York.


214

93. Olson, H.W., Krosley, L., Nelson, K., Chabrillat, S., Goetz, A.F.H and Noe,

D.C., (2000). Mineralogy swelling potential relationships for expansive shales.

Advances in Unsaturated Geotechnics ,Geotechnical Special Publication No. 99,

American Society of Civil Engineers, New York, pp. 361-371.

94. O’Neill, M. W., and Ghazzaly, O. I. (1977). “Swell potential related to

building performance.” J. Geotech. Engrg. Div., 103(12), 1363–1379.

95. Parker, J.C., D.F. Amos, and D.L. Kaster. (1977). “An evaluation of several

methods of estimating soil volume change”. Soil Sci. Soc. Am. J. 41:1059-1064.

96. Parker, J.C., D.F. Amos, and L.W. Zelazny. (1982). “Water adsorption and

swelling of clay minerals in soil systems”. Soil Sci. Soc. Am. J. 46:450-456.

97. Pearring, J.R., (1963), “A study of basic mineralogical, physical-chemical, and

engineering index properties of laterite soils”. Ph.D. Dissertation, Texas A&M

Univ., College Station, TX, USA. (Quoted by Covar and Lytton, 2001).

98. Peck, R.B., W.E. Hanson and T.H. Thornburn (1974), “Foundation

Engineering”, Second edition., Wiley, New York.

99. Pengelly, A.D., Boehm, D.W., Rector, E., and Welsh, J.P. (1997).

“Engineering experience with in-situ modification of collapsible and expansive

soils”. Unsaturated Soil Engineering Practice, Geotechnical Special Publication

No. 68, American Society of Civil Engineers, New York, pp. 277-298.

100. Perrin, L.L. (1992). “Expansion of lime-treated clays containing sulfates”.

Proceedings, Seventh International Conference on Expansive Soils, Dallas,

Texas, pp. 409-414.


215

101. Petry, T.M. and Armstrong, J.C. (2001). “Stabilization of expansive clay

soils”. Transportation Research Record 1219, Transportation Research Board-

National Research Council, Washington, D.C., pp. (Quoted by Hardcastle,

,2003).

102. Petry, T. M., and Little, D. N. (2002). ‘‘Review of Stabilization of Clays and

Expansive Soils in Pavements and Lightly Loaded Structures-History, Practice,

and Future’’ Journ. of Mat. in Civil Eng. , Nov.- Dec., pp. 447-460.

103. Pidgeon, J. T. (1987), “Prediction of differential heave for design of

foundations in expansive soil areas”. 9th African Regional Conference on Soil

Mechanics and Foundation Engineering, Lagos, Nigeria.

104. Porter, A. A., and Nelson, J. D. (1980).”Strain controlled testing of soils”.

Proc. 4th Int. Conf. Expansive Soils, ASCE and Int. Soc. Soil Mech. Found.

Eng., Denver, June: pp.34-44.

105. Puppala, A.J. and Musenda, C. (2000). “Effects of fiber reinforcement on

strength and volume change in expansive clay”. Transportation Research

Record 17336, Transportation Research Board-National Research Council,

Washington, D.C., pp. 134-140. (Quoted by Hardcastle, 2003).

106. Ralph, T. K., and Nagar, N. , (1972).”The prediction of heave from laboratory

tests”. Univ. of Witwatersrand, Johannesburg. (Quoted by Nelson and Miller,

1992).

107. Rama, R., Rahadjio, H., and Fredlund, D.G. (1988). “Closed from heave

solutions for expansive soils”. Journal of Geot. Eng. ASCE, Vol. 144, No. 5, pp.

573-589.


216

108. Raman, V., (1967) "Identification of Expansive Soils from the Plasticity Index

and the Shrinkage Index Data," The Indian Engineer, Calcutta, Vol 11, No.1,

Jan, pp 17-22. (Quoted by Snethen et al., 1975).

109. Ranganatham, B. V. and Satyanarayana, B., (1965) “A Rational Method of

Predicting Swelling Potential for Compacted Expansive Clays," Proceedings,

Sixth International Conference on Soil Mechanics and Foundation Engineering,

Vol 1, pp 92-96.

110. Richard, Sulliven and Bramlette, (1969). “Predicting heave buildings on

unsaturated clay”. Int. Conf. Texas.

111. Rogers, C. D. and Glendinning, S., (2000), “Lime Requirement for

Stabilization,” Transportation Research Record, No. 1721, Transportation

Research Board, National Research Council, Washington, D.C., pp. 9–18.

(Quoted by Al-Rawas, et al., 2002).

112. Rojas, E., Romo, M. P., Garnica, P., and Cervantes, R. (2006). “Analysis of

deep moisture barriers in expansive soils. Part II: Water flow formulation and

implementation.” Int. J. Geomech., 6(5), PP. 319–327.

113. Rollings, M.P. and Rolling, R.S., (1996), Geotechnical Materials in

Construction, McGraw-Hill, New York. (Quoted by YEŞİLBAŞ, 2004).

114. Satyanaryana, B., (1969) “Behavior of treated expansive soil" Proceedings,

second International Research and Engineering Conference on Expansive Clay

Soils, Texas, pp 308-316.

115. Schneider, G.L., and Poor, A.P.(1974). ” The prediction of soil heave and

swell pressures developed by expansive clay. Res. Rep. TR. 9.74,Construction

Res. Center, Univ. of Texas, Arlington.


217

116. Schreiner, H.D. (1988). “Identification and classification of expansive soils”. p.

23-29. In C.V.J. Varma (ed.) Proc. 6th Intl. Conf. on Expansive Soils. New

Dehli, India. (Quoted by Thomas, 1998).

117. Seed, H. B., Mitchell, J. K. and Chan, C. K. (1960), “The strength of

compacted cohesive soils”. Proceedings, ASCE Research Conference on

Cohesive soils, Boulder, American Society of Civil Engineers, New York, pp

877-964.

118. Seed, H. B., Mitchell, J. K., and Chan, C. K. (1962a). ‘‘Study of swell and

swell pressure characteristics of compacted clays.’’ Highway Research Board

Bulletin 313, National Research Council, Washington, D.C., 12–39.

119. Seed, H.B., Woodward, R.J., Jr., and Lundgren, R., (1962b), “Prediction of

swelling potential for compacted clays”. Journal of Soil Mech. Found. Div.,

Am. Soc. Civil Eng. 88 (SM3), 53-87.

120. Sivapullaiah, P. V., Sridharan, A., Prashanth, J. P., and Raju, K. V. B.,

(1997) “Role of Fabric and Bonding in the Lime Stabilization of Soils,”

Proceedings of the 4th International Conference on Civil Engineering, Sharif

University of Technology, Tehran, Iran, Vol. 2, 4–6 May, pp. 345–352.

121. Skempton, A.W., (1953). ‘‘The colloidal activity of clays.’’ Proc., 3rd Int.

Conf. on Soil Mechanics, Foundation Engineering, Vol. 1, 57–61.

122. Snethen, D. R., Townsend, F. C., Johnson, L. D., Patrick, D. M., and

Vedros, P. J. ,(1975). ‘‘A review of the engineering experiences with expansive

soils in highway sub-grades.’’ Rep. No. FHWA-RD-75-48, Interim Report to

FHWA, Washington, D.C.


218

123. Snethen, D.R. (1979a). “Technical Guidelines for Expansive Soils in Highway

Sub-grades”. Report No. FHWA-RD-79-51. U.S. Army corps of Engineers

Waterways Experiment Station, Vicksburg, MS.

124. Snethen, D. R. (1979b). ‘‘An evaluation of methodology for predicting and

minimization of detrimental volume change of expansive soils in highway sub-

grades.’’ Rep. No. FHWA-RD-79-49, Final Rep., Washington,D.C.

125. Snethen, D.R., (1980).”Characterization of Expansive Soils Using Soil Suction

Data", 4th International Conference on Expansive Soils, pp. 54-75,

126. Snethen, D. R. (1984). ‘‘Evaluation of expedient methods for identification and

classification of potentially expansive soils.’’ Proc., 5th Int. Conf. on Expansive

Soils, 22–26.

127. Soil Survey Staff. (1996). Soil survey laboratory methods manual. Soil Survey

Invest. Rpt. No. 42, Ver. 3. National Soil Survey Center, Lincoln, NE. (Quoted

by Thomas, 1998).

128. Sowers, G.B., Sowers, G.F., (1970). “Introductory Soil Mechanics and

Foundations”, 3rd ed. McMillan, New York. (Quoted by Yilmaz, 2006)

129. Steinberg, M. (2000). “Expansive soils and the geomembrane remedy”.

Advances in Unsaturated Geotechnics, Geotechnical Special Publication 99,

American Society of Civil Engineers, Reston, Virginia, pp. 456-466.

130. Sullivan, R. A. and McClelland, B., (1969), “Predicting Heave of Buildings on

Unsaturated Clay,” Second International Research and Engineering Conference

on Expansive Clay Soils, Texas A&M University, College Station, TX, pp.

404–420.


219

131. Tensar Earth Technologies, Inc. (2001). “Strategies for Sub-grade

Stabilization: A Comparison of Tensar Biaxial Geogrid vs. Lime”. Tensar Earth

Technologies, Inc., Atlanta, Georgia. 2pp. (Quoted by Hardcastle, 2003).

132. TFHRC (Transportation and Federal Highways Research Council), (2003),

“Coal Fly Ash-User Guideline-Embankment or Fill”

http://www.tfhrc.gov/hnr20/recycle/waste/cfa54.htm.

133. Thomas, P. J. ,(1998). “Quantifying Properties and Variability of Expansive

Soils in Selected Map Units”, Ph.D. Thesis, Faculty of the Virginia, 79pp.

134. Thomas, P.J., Baker J.C., and Zelazny L.W., (2000), “An expansive soil

index for predicting shrink-swell potential”, Soil Science Society of America

Journal, 64, 268- 274. (Quoted by Kariuki, 2003).

135. TM–ARMY (1983), “Foundations in expansive soils”, Publications of the

Headquarters, United States Army Corps of Engineers. TM 5-818-7 0098

CEMP, USA.

136. Transportation Research Board. (1987). “Lime Stabilization: Reactions,

Properties, Design, and Construction”. State of the Art Report 5, Transportation

Research Board-National Research Council, Washington, D.C., 59 pp. (Quoted

by Hardcastle, 2003).

137. U.S. Department of Agriculture, U.S.D.A., Soil Conservation Service, (1971).

“Guide for Interpreting Engineering Uses of Soils,” Washington, D.C., 87 pp.

138. Uniform Building Code, (1968). UBC Standard No. 29-2. (Quoted by Yilmaz,

, 2006)


220

139. Van der Merwe, D.H. (1964). “The prediction of heave from the plasticity

index and percentage clay fraction of soils”. The Civil Engineer in South Africa,

South African Institution of Civil Engineers, Vol. 6, pp. 103-116.

140. Vassilev, S.V., Menendez, R., Alvarez, D., Somoano, M.D. and Tarazona,

M.R.M., (2003), “Phase-Mineral and Chemical Composition of Coal Fly Ashes

as a Basis for their Multicomponent Utilization. 1. Characterization of Feed

Coals and Fly Ashes”, Fuel, Vol. 82, pp. 1793-1811. (Quoted by Çetiner, 2004).

141. Vijayvergiya, V. N., and Ghazzaly, O. I.,(1973), “Prediction of swelling

potential for natural clays”, 3rd Int. Conf. on Expansive Soils, Haifa, pp. 227-

236.

142. Wayne, A.C., Mohamed, A.O., and El-Fatih, M.A., (1984), “Construction on

Expansive Soils in Sudan”, Journal of Construction Engineering and

Management, Vol. 110, pp. 359-374. (Quoted by YEŞİLBAŞ, 2004).

143. Wibawa, B., (2003), “Swelling Clay Stabilized with Lime” Civil Engineering

Research, pp. 47-48. (Quoted by YEŞİLBAŞ, 2004).

144. Weston, D.J., (1980). ‘Expansive road treatment for Southern Africa”.

Proceeding of the 4th International Conference on Expansive Soils, Denver, vol.

1, pp. 339–360.

145. Woodward-Clyde-Sherard and Associates, (1968). "Final report of field

survey: Remedial methods applied to houses damaged by high volume change

soils” Federal Housing Administration, FHA Contract H-799.94.

146. Yesilbas G., (2004), “Stabilization of expansive soils using aggregate waste,

rock powder and lime” M.S., Department of Civil Engineering, the Middle East

Technical University, April, 112 pages.


221

147. Yevnining, A. and Zaslavsky, D. (1970),”Some factors affecting compacted

clay swelling”. Can. Geot. Journ., Vol. 7 , No. 79, pp. 79-91.

148. Yilmaz, I., (2006). “Indirect estimation of the swelling percent and a new

classification of soils depending on liquid limit and cation exchange capacity”,

Engineering Geology Volume 85, Issues 3-4, 21 June 2006, Pages 295-301.

149. Yoshida, R. T., Fredlund, D. G., and Hamilton, J. J. (1983). “The Prediction

of total heave of aslab on grade floor on reginea clay”. Can. Geotech. J. Vol. 20,

No. 1,pp. 69-81. Yousif, M.S.M. (2000),” Slope Stability of the Middle Eocene

Rocks of Gebel Mokattam” ICEHM, Cairo University, Egypt, September , pp

14- 32.

150. Zacharias, G. and Ranganathan, B. (1972). “Swelling and swelling

characteristics of synthetic clays”. Proc. Of Symposium on strength and

deformation behavior of soils, Banglore, Vol. 1, pp. 129-134. (Quoted by

Abdel-Moatty, 1999).

151. Zaman, M., Laguros, J. G., and Sayah, A., (1992), “Soil Stabilization Using

Cement Kiln Dust,” Proceedings of the 7th International Conference on

Expansive Soils, Dallas, TX, Vol. 1, 3–5 August, pp. 347–351.

152. Zapata, C.E., Houston, W.N., Houston, S.L. and Walsh, K.D. (2000). “Soil

water characteristic curve variability”. Advances in Unsaturated Geotechnics,

Geotechnical Special Publication 99, American Society of Civil Engineers,

Reston, Virginia, pp. 84-124.


222

Appendix A

Measured laboratory footing heave – logarithmic time relationships:

Figure (A.1): Footings heave versus logarithmic time for variable footing stress

Group 1 and Group 2

Figure (A.2): Footings heave versus logarithmic time, Group 2

0

5

10

15

20

25

30

1 10 100 1000 10000 100000

∆h (m

m)

Time (min)


q1=0.5 kg/cm2q2=0.0q3=0.0q4=0.5 kg/cm2

F1 F2 F3 F4



0

5

10

15

20

25

1 10 100 1000 10000 100000

∆h (m

m)

Time (min)

Treated expansive soil using medium sand

q1=0.5 kg/cm2q2=0.0q3=0.0q4=0.5 kg/cm2

F1 F2 F3 F4


Expansive soil

Medium Sand


223



0

5

10

15

20

25

1 10 100 1000 10000 100000

∆h (m

m)

Time (min)

Treated expansive soil using medium sand ‐ 5% lime

q1=0.5 kg/cm2

q2=0.0

F1 F2 F3 F4


Expansive soil

Medium Sand ‐ 5 % lime

0

5

10

15

20

25

1 10 100 1000 10000 100000

∆h (m

m)

Time (min)

Treated expansive soil layer using 5% lime ‐ expansive soil

q1=0.5 kg/cm2 q2=0.0

q3=0.0 q4=0.5 kg/cm2

F1 F2 F3 F4


Expansive soil

Expansive soil ‐ 5 % lime


224

Figure (A.5): Footings heave versus log time leakage below F1, Group 3

Figure (A.6): Footings heave versus log time leakage below F1, Group 3 and

Group 4

0

5

10

15

20

25

1 10 100 1000 10000 100000

∆h (m

m)

Time (min)

Treated expansive soil using medium sand and plastic horizontal barrier at top of sand

q1=0.5 kg/cm2 q2=0.0

q3=0.0 q4=0.5 kg/cm2

F1 F2 F3 F4


Expansive soil

Medium Sand

0

5

10

15

20

25

1 10 100 1000 10000 100000

∆h (m

m)

Time (min)

Treated expansive soil using medium sand and plastic horizontal barrier at bottom of sand

q1=0.5 kg/cm2 q2=0.0q3=0.0 q4=0.5 kg/cm2

F1 F2 F3 F4


Expansive soil

Medium Sand


225

Figure (A.7): Footings heave versus log time leakage 10cm from F1, Group 4

Figure (A.8): Footings heave versus log time leakage 20cm from F1, Group 4

0

5

10

15

20

25

1 10 100 1000 10000 100000

∆h (m

m)

Time (min)


q1=0.5 kg/cm2 q2=0.0

q3=0.0 q4=0.5 kg/cm2

F1 F2 F3 F4


Expansive soil

Medium Sand

10 cm

0

5

10

15

20

25

1 10 100 1000 10000 100000

∆h (m

m)

Time (min)


q1=0.5 kg/cm2q2=0.0q3=0.0

F1 F2 F3 F4


Expansive soil

Medium Sand

20 cm


226

Appendix B

Final moisture distribution of expansive soil:

Figure (B.1): Wc (%) versus depth for variable footing stress Group 1 and

Group 2

Figure (B.2): Wc (%) versus depth, Group 2

S.L P.L L.L0

1

2

3

4

5

6

7

8

0 10 20 30 40 50 60 70 80 90

Dep

th (cm)

Wc (%)


q1=0.5 kg/cm2q2=0.0q3=0.0q4=0.5 kg/cm2

Wco = 4 %

F1 F2 F3 F4



S.L P.L L.L2

3

4

5

6

7

8

0 10 20 30 40 50 60 70 80 90

Dep

th (cm)

Wc (%)Treated expansive soil using medium sand

q1=0.5 kg/cm2q2=0.0q3=0.0q4=0.5 kg/cm2

Wco = 4 %

F1 F2 F3 F4


Expansive soil

Medium Sand


227



S.L P.L L.L2

3

4

5

6

7

8

0 10 20 30 40 50 60 70 80 90

Dep

th (cm)

Wc (%)

Treated expansive soil using medium sand + 5% lime

q1=0.5 kg/cm2

q2=0.0

q3=0.0

q4=0.5 kg/cm2 Wco = 4 %

F1 F2 F3 F4


Expansive soil

Medium Sand ‐ 5 % lime

S.L P.L L.L2

3

4

5

6

7

8

0 10 20 30 40 50 60 70 80 90

Dep

th (cm)

Wc (%)Treated expansive soil layer using 5% lime ‐ expansive soil

q1=0.5 kg/cm2q2=0.0q3=0.0q4=0.5 kg/cm2 Wco = 4 %

F1 F2 F3 F4


Expansive soil

Expansive soil ‐ 5 % lime


228

Figure (B.5): Wc (%) versus depth leakage below F1, Group 3

Figure (B.6): Wc (%) versus depth leakage below F1, Group 3 and Group 4

S.L P.L L.L2

3

4

5

6

7

8

0 10 20 30 40 50 60 70 80 90

Dep

th (cm)

Wc (%)

Treated expansive soil using medium sand and plastic horizontal barrier at top of sand

q1=0.5 kg/cm2

q2=0.0

q3=0.0

q4=0.5 kg/cm2

Wco = 4 %

F1 F2 F3 F4


Expansive soil

Medium Sand

S.L P.L L.L2

3

4

5

6

7

8

0 10 20 30 40 50 60 70 80 90

Dep

th (cm)

Wc (%)


q1=0.5 kg/cm2

q2=0.0

q3=0.0

q4=0.5 kg/cm2Wco = 4 %

F1 F2 F3 F4


Expansive soil

Medium Sand


229

Figure (B.7): Wc (%) versus depth leakage 10cm from F1, Group 4

Figure (B.8): Wc (%) versus depth leakage 10cm from F1, Group 4

S.L P.L L.L2

3

4

5

6

7

8

0 10 20 30 40 50 60 70 80 90Dep

th (cm)

Wc (%)


q1=0.5 kg/cm2

q2=0.0

q3=0.0

q4=0.5 kg/cm2 Wco = 4 %

F1 F2 F3 F4


Expansive soil

Medium Sand

10 cm

S.L P.L L.L2

3

4

5

6

7

8

0 10 20 30 40 50 60 70 80 90

Dep

th (cm)

Wc (%)


q1=0.5 kg/cm2

q2=0.0

q3=0.0

q4=0.5 kg/cm2 Wco = 4 %

F1 F2 F3 F4


Expansive soil

Medium Sand

20 cm

لجنة الممتحنين

محمد يوسف عبد اللطيف يوسف : اسم الباحث

تقييم أداء بعض نظم معالجة التربة اإلنتفاشية: عنوان الرسالة

الماجستير: الدرجة

التوقيع االسم

خديجة إبراهيم عبد الغنى . د.أ

أستاذ الهندسة الجيوتقنية واألساسات

البناء اإلسكان و مرآز بحوث

عبد المنعم أحمد موسى / د.أ


جامعة عين شمس -آلية الهندسة

على عبد الفتاح على أحمد/ د.أ


جامعة عين شمس -آلية الهندسة

سجامعة عين شم آلية الهندسة

قسم الهندسة اإلنشائية رسالة ماجستير

محمد يوسف عبد اللطيف : اسم الطالب

تقييم أداء بعض نظم معالجة التربة اإلنتفاشية :عنوان الرسالة

ماجستير :الدرجة

شرا فاإل لجنة

د على عبد الفتاح على أحمد.أ

أستاذ الهندسة الجيوتقنية واالساسات جامعة عين شمس -آلية الهندسة

هدى عبد الهادى إبراهيم. د

مدرس الهندسة الجيوتقنية واالساساتجامعة عين شمس -آلية الهندسة

٢٠٠٨/ / : تاريخ البحث

الدراسات العليا

: اجيزت الرسالة : تم اإلجازة

/ /٢٠٠٨/ / ٢٠٠٨

: موافقة مجلس الجامعة : موافقة مجلس الكلية

/ /٢٠٠٨/ / ٢٠٠٨

التعريف بالباحث

محمد يوسف عبد اللطيف يوسف: اإلسم

١٩٨٣/ ٦/ ٢٣: تاريخ الميالد

جمهورية مصر العربية –اإلسكندرية : محل الميالد

بكالوريوس هندسة مدنية :الدرجة الجامعية األولى

مدنى إنشاءات : التخصص

٢٠٠٥يونيو : اريخ المنح ت

جامعة عين شمس -آلية الهندسة : ةنحاالم الجهة

جامعة عين شمس -الهندسة معيد بكلية: الوظيفة

سجامعة عين شم آلية الهندسة

يةئقسم الهندسة اإلنشا

محمد يوسف عبد اللطيف: ملخص الرسالة المقدمة من

تقييم أداء بعض نظم معالجة التربة اإلنتفاشية: عنوان الرسالة

أستاذ الهندسة الجيوتقنية واالساسات أحمد على د على عبد الفتاح.أ :إشراف تحت

مدرس الهندسة الجيوتقنية واالساسات إبراهيم هدى عبد الهادى. د

٢٠٠٥/ ١٠/١٠: تاريخ التسجيل

:تاريخ المناقشة

ملخص البحثبعض هذه العوامل يتعلق بالظروف . السطحية المرتكزة على تربة إنتفاشية معالجة بعدة عوامل يتأثر السلوك اإلنتفاشى لألساسات

بينما تتعلق العوامل األخرى بطرق المعالجة المستخدمة لتقليل الجهد ، المحيطة المتحكمة فى التكوين الطبقى وآيفية تحرك المياهخالل العقود األخيرة ظهرت بر آافية لمعالجة التربة اإلنتفاشية إال أنهتعت معالجةالوبالرغم من ان بعض طرق .اإلنتفاشى للتربة

.العديد من المشاآل للمنشآت ذات األساسات السطحية التى تم تنفيذها على تربة إنتفاشية معالجة

سلوك اإلنتفاشى ولذلك آان الهدف األساسي من هذا البحث هو دراسة مدى فاعلية بعض طرق المعالجة شائعة اإلستخدام لتقليل الدراسة حالة ميل مبنى سكنى بالهضبة الوسطى تم و. مع دراسة العوامل التى تؤدى إلى حدوث حرآة نسبية بين القواعد لتربةل

وعلى الرغم من أن التربة أسفل المبنى آانت معالجة بتربة إحالل إال أنها فشلت فى منع األضرار الناتجة عن اإلنتفاش .للمقطمتم توصيف التتابع الطبقى قد و. من مصدر صرف قريب أسفل تربة اإلحالل وجود تسرب للمياه لىرجع ذلك إوي، النسبى للتربة

للهضبة الوسطى للمقطم وتعريف وتصنيف التربة وتعيين الخواص اإلنتفاشية لها وتوصيف الحالة اإلنشائية لمبنى تعرض للميل مساحية ست رصدات وتم تحليل . نتيجة تسرب المياه الى تربة التأسيس المعالجة بتربة إحالل من الرمل بالهضبة الوسطى للمقطم

باستخدام طريقتينلمبنى ل األفقيةة حرآال مرآبات آما تم تعيين. تصل إلى أربع سنين تقريبازمنية المبنى على مدى فترةلحرآة وذلك لتحديد أفضل و أنسب الطرق التى يمكن استخدامها للتنبؤ بالحرآة ، ومقارنتها بالقيم المرصودة على الطبيعة تجريبيتينشبه

.اإلنتفاشية لألساسات

قواعد السطحية المرتكزة على تربة إنتفاشية معالجة لنماذج من الم وضع برنامج إختبار معملى لتعيين السلوك اإلنتفاشى وقد تاستخدام الطرق المباشرة وغير المباشرة فى تعيين الخواص اإلنتفاشية أيضا وتم. بطرق مختلفة بإستخدام نموذج معملى آبير

الطين المخلوط مع جير أو% ٥والرمل المخلوط مع أربة اإلنتفاشية باستخدام اإلحالل بالرمل وقد تمت معالجة الت.للتربة المختبرةقياس الحرآة اإلنتفاشية للقواعد السطحية المرتكزة وتم. اثيلين البولىعازلة من أفقية شرائح المضاف إليهالرمل جير أو% ٥

. بإستخدام المعادالت التجريبية وشبه التجريبية التى تعتمد على الخواص اإلنتفاشية للتربة بها تنبؤعلى التربة المعالجة معمليا وال .وقد أخذ فى اإلعتبار تأثير أحمال القواعد ومنسوب شرائح البولى اثيلين وبعد مصدر تسرب المياه عن القواعد المنفصلة

وأظهر وجود الشرائح . النهائية لإلنتفاش تقريبٌا بنفس النسبة وقد أثبتت جميع طرق المعالجة المستخدمة نجاحٌا فى تقليل القيمةاألفقية العازلة أعلى منسوب تربة اإلحالل تميزٌا من حيث توزيع الرطوبة بصورة منتظمة فى حالة حدوث تسرب للمياه من

أدى ، اإلحالل أسفل منسوبلة عازشرائح أفقية ولكن عند استخدام . وهذا التوزيع نتج عنه انتفاش منتظم أسفل القواعد، خاللهوقد .تسرب المياه من الغشاء إلى ترآيز الرطوبة فى نقطة محدودة من التربة أنتج عنها حدوث حرآة نسبية عالية بين القواعد

األنظمة المقترحةأسهم هذا البحث فى التعرف على التربة اإلنتفاشية وبعض المشاآل المصاحبة لها وآذلك زيادة فاعلية .ناتجة عن سلوك التربة اإلنتفاشيةالتى من شأنها أن تقلل األضرار اإلنشائية الو تهاجاللمع

جامعة عين شمس آلية الهندسة

تقييم أداء بعض نظم معالجة التربة اإلنتفاشية إعداد

محمد يوسف عبد اللطيف يوسف/ م

٢٠٠٥هندسة مدنية عام بكالوريوس جامعة عين شمس

رسالة مقدمة

للحصول على درجة الماجستير فى الهندسة المدنية )الهندسة اإلنشائية شعبة(

ف إشراتحت

د على عبد الفتاح على أحمد.أ

أستاذ الهندسة الجيوتقنية واالساسات جامعة عين شمس -آلية الهندسة

هدى عبد الهادى إبراهيم. د

مدرس الهندسة الجيوتقنية واالساساتجامعة عين شمس -آلية الهندسة

٢٠٠٨

Date post:	24-Feb-2023
Category:	Documents
Upload:	shams
View:	0 times
Download:	0 times

SWELLING PERFORMANCE OF SOME EXPANSIVE SOIL TREATMENT TECHNIQUES BY

Documents