MODELING STUDIES TO ASSESS LONG TERM SETTLEMENT

OF LIGHT WEIGHT AGGREGATE EMBANKMENT

by

SEEMA KALLA

Presented to the Faculty of the Graduate School of

The University of Texas at Arlington in Partial Fulfillment
of the Requirements
for the Degree of

MASTER OF SCIENCE IN CIVIL ENGINEERING

THE UNIVERSITY OF TEXAS AT ARLINGTON

December 2010

Copyright by Seema Kalla 2010

All Rights Reserved

ACKNOWLEDGEMENTS

I would like to express my gratitude to Dr. Anand J. Puppala for his valuable
guidance, constant support and encouragement throughout my Masters. I would also like to
thank my dissertation committee Dr. Laureano Hoyos and Dr. Melanie L. Sattler for readily
accepting to serve in my committee and for their valuable suggestions. I specially thank Dr.
Sireesh, Ekarut and Dr. Chittoori for their support throughout my thesis. I also thank the staff of
civil and environmental engineering department in UTA.
I would also like to thank my family and friends for being so understanding and supportive
of me during this journey. I am also very grateful to my beloved parents, husband and my
siblings for their unending support and faith in me.

November 10, 2010

iii

ABSTRACT
MODELING STUDIES TO ASSESS LONG TERM SETTLEMENT OF LIGHT WEIGHT
AGGREGATE EMBANKMENT

Seema Kalla, M.S

The University of Texas at Arlington, 2010

Supervising Professor: Anand J. Puppala

Expanded clay and shale (ECS) is a low weight material derived from clay and
shale, and has been used in large quantities throughout the world. Very few studies have been
performed to investigate the behavior of ECS and how it can reduce the settlements of
embankments when used as a fill material. This thesis presents research findings of numerical
modeling to assess the long term settlements of ECS embankment on a soft clay deposit and
supports both pavement and traffic loading. Observed settlements near the pavement of a light
weight ECS embankment built on soft clay indicated that this material is adequate in reducing
settlements when compared to a normal fill embankment.
Laboratory investigations were conducted on both embankment materials and these
results showed that normal clayey fill with low plasticity have high compressible properties when
compared to the ECS granular material. Numerical modeling followed by the hyperbolic
formulation showed that the ECS embankment and normal embankment on soft clay could
provide adequate information to interpret long term settlements. Analyses showed that the ECS
embankment is efficient in minimizing the long term settlements when compared to normal fill
embankment.
iv

TABLE OF CONTENTS
ACKNOWLEDGEMENTS ................................................................................................................iii
ABSTRACT ..................................................................................................................................... iv
LIST OF ILLUSTRATIONS............................................................................................................. viii
LIST OF TABLES ............................................................................................................................ xi
Chapter

Page
1. INTRODUCTION...... ..................................... 1
1.1 General ............................................................................................................. 1
1.2 Research Objective .......................................................................................... 1
1.3 Thesis Organization ......................................................................................... 2
2. LITERATURE REVIEW .................................................................................................. 4
2.1 Factors Causing Settlement ............................................................................. 4
2.2 Remedial Measures to Reduce Settlement ..................................................... 6
2.2.1 Light-Weight Expanded Polystyrene, EPS....................................... 7
2.2.2 Light Weight Fill from a Mixture of Fly Ash and Bottom Ash............ 8
2.2.3 Lightweight Tire Scrap ................................................................... 10
2.2.4 Lightweight Aggregate ................................................................... 12
2.2.4.1 Lightweight Expanded Clay Aggregate .......................... 13
2.2.4.2 LWA from Mining Residues, Heavy Metal
Sludge, and Incinerator Fly Ash ................................... 14
2.2.4.3 ECS as Lightweight Aggregate ...................................... 15
2.3 Numerical Modeling ....................................................................................... 17
2.3.1 Modeling for Settlements ............................................................... 18
2.3.2 Difference between Finite Element Method
And Finite Difference Method ........................................................ 19
vi

2.4 Instrumentation............................................................................................... 20
2.4.1 Characteristics and Instrumentation Objective .............................. 20
2.4.1.1 Analytical Assessment ..................................................... 21
2.4.1.2 Prediction of Future Performance .................................... 21
2.4.1.3 Legal Evaluation............................................................... 21
2.4.1.4 Development and Verification of Future Design .............. 22
2.4.2 Different Type of Instrumentation Used in the
Field Monitoring .............................................................................. 22
2.4.2.1 Inclinometers ................................................................. 22
2.5 Surface Elevation Surveying .......................................................................... 23
2.6 Visual Monitoring ............................................................................................ 23
2.7 Summary ........................................................................................................ 23
3. EXPERIMENTAL STUDIES ......................................................................................... 24
3.1 Quality Control for Laboratory Testing ........................................................... 24
3.1.1 Tests Conducted ............................................................................ 24
3.1.1.1 Sieve Analysis and Hydrometer Analysis .................... 24
3.1.1.2 Atterberg Limits ............................................................ 26
3.1.1.3 Compaction Test .......................................................... 28
3.1.1.4 Direct Shear Test ......................................................... 29
3.1.1.5 Consolidation Test ....................................................... 31
3.1.1.6 Vertical Swell Tests ...................................................... 34
3.1.2 Selection of Design Properties ....................................................... 35
3.2 Summary ........................................................................................................ 37
4. NUMERICAL MODELING ............................................................................................ 38
4.1 Modeling Methods .......................................................................................... 38
4.2 Finite Element Modeling Procedure ............................................................... 39
vii

4.2.1 Choice of Constitutive Model and Material Properties ................... 39

4.2.2 Mesh Generation ............................................................................ 43
4.2.3 Initial and Boundary Conditions ..................................................... 44
4.2.4 Initial Stresses ................................................................................ 45
4.2.5 Calculation ..................................................................................... 46
4.2.6 Results of Finite Element Analysis ................................................. 47
4.2.7 Vertical Stresses in Finite Element Analysis .................................. 50
4.3 Finite Difference Modeling Procedure ............................................................ 52
4.3.1 Grid Generation .............................................................................. 53
4.3.2 Boundary and Initial Conditions ..................................................... 54
4.3.3 Choice of Constitutive Model and Material Properties .................. 55
4.3.4 Results in Finite Difference Analysis.............................................. 58
4.3.5 Vertical Stresses in Finite Difference Analysis .............................. 62
4.4 Summary ........................................................................................................ 64
5. ANALYSIS OF FIELD RESULTS AND COMPARISONS WITH
NUMERICAL RESULTS .............................................................................................. 65
5.1 Analysis of Field Results ................................................................................ 66
5.1.1 Instrumentation Data ...................................................................... 66
5.2 Pavement Surface Profiling............................................................................ 71
5.3 Hyperbolic Formulation .................................................................................. 72
5.3.1 Long Term Settlements Using Field Elevation Data with
Hyperbolic Formulation ................................................................... 72
5.4 Comparison of Field Results with Hyperbolic Formulation ............................ 73
5.5 Comparison of Hyperbolic Formulation Results with Numerical Model
Results ........................................................................................................... 74
5.5.1 Vertical Displacement at Pavement Surface.................................. 77
5.6 Summary ........................................................................................................ 79
viii

6. SUMMARY, CONCLUSIONS AND FUTURE RESEARCH RECOMMENDATIONS ... 81

6.1 Introduction..................................................................................................... 81
6.2 Summary and Conclusions ............................................................................ 81
6.2 Future Research Recommendations ............................................................. 82
REFERENCES ............................................................................................................................... 83
BIOGRAPHICAL INFORMATION .................................................................................................. 93

ix

LIST OF ILLUSTRATIONS

Figure

Page

2.1 Longitudinal Profile of EPS Embankment at Lokkeberg Bridge,

Swedish Border - Fredyland and Aaboe, 2001 .......................................................................... 7
2.2 Deformations in EPS Embankment at Lokkeberg - Fredyland and Aaboe, 2001 ...................... 8
2.3 Disposal pond of fly and bottom ash mixture generated at Wabash
River Power Plant -Sungmin Yoon et al. 2009 ........................................................................... 9
2.4 Cross section of an embankment at State Road 641, Terre Haute, Ind.
- Sungmin Yoon et al. 2009 ........................................................................................................ 9
2.5 Scrap tire beads - An Deng and Jin-Rong Feng, 2009 ............................................................ 10
2.6 Cross section through embankment for the Portland Jetport Interchange
- Humphrey et al., 1998 ............................................................................................................ 11
2.7 LECA Granular Arioz et al., 2008 .......................................................................................... 14
2.8 Schematic of lightweight embankment system to reduce settlement ...................................... 14
2.9 Ternary plot of LWA - Huang et al., 2007................................................................................. 15
2.10 ECS Lightweight Aggregate - Saride et al. 2008 ................................................................... 17
2.11 Discretization methods (a) FDM (b) FEM............................................................................... 20
2.12 Details of inclinometer casing and assembling procedure - Slope Indicator, 1997 ............... 22
3.1 Particle Size Distribution for ECS ............................................................................................. 25
3.2 Plasticity Chart for Normal Fill .................................................................................................. 27
3.3 Compaction Curve for Normal Fill ............................................................................................ 29
3.4 Direct Shear Test results for ECS material .............................................................................. 30
3.5 Direct Shear Test results for Normal Fill material .................................................................... 30
3.6 e log (p) graph for Normal Fill material.................................................................................. 32
3.7 e log (p) graph for Normal Fill material ................................................................................. 33
x

3.8 e log (p) graph for ECS material .......................................................................................... 33

3.9 Free Swell test graph for Normal Fill ........................................................................................ 34
4.1 Geometry of embankment model in PLAXIS ........................................................................... 41
4.2 Typical Mesh Generation in PLAXIS ........................................................................................ 44
4.3 Active Water Level in PLAXIS .................................................................................................. 44
4.4 POP and OCR profile illustration - PLAXIS 8, User Manual .................................................... 45
4.5 Calculation steps using FEM program ..................................................................................... 46
4.6 Settlement vs. Time plot in the pavement in FEM using MCM for Embankment .................... 47
4.7 Settlement contours at the end of pore pressure dissipation in ECS
using MCM for Embankment.................................................................................................... 48
4.8 Settlement contours at the end of pore pressure dissipation in Normal Fill
using MCM for Embankment.................................................................................................... 48
4.9 Settlement vs. Time plot in the pavement in FEM using SSM
for Embankment ....................................................................................................................... 49
4.10 Settlement contours at the end of pore pressure dissipation in ECS using
SSM for Embankment ............................................................................................................ 49
4.11 Settlement contours at the end of pore pressure dissipation in Normal Fill
using SSM for Embankment .................................................................................................. 50
4.12 Stress distribution in Normal Fill after full dissipation of pore pressure using
MCM for Embankment ........................................................................................................... 51
4.13 Stress distribution in ECS Fill after full dissipation of pore pressure using
MCM for Embankment ........................................................................................................... 51
4.14 Stress distribution in Normal Fill using SSM for Embankment............................................... 52
4.15 Stress distribution in ECS Fill using SSM for Embankment ................................................... 52
4.16 Fine grid generation in FDM ................................................................................................... 54
4.17 Geometry with selected Material layers and initial conditions ............................................... 55
4.18 Settlement Results near pavement in FLAC using MCM for Embankment .......................... 60
4.19 Maximum Settlement of ECS using MCM for Embankment .................................................. 61
4.20 Maximum Settlement of Normal Fill using MCM for Embankment ........................................ 61
xi

4.21 Stress distribution in ECS Fill at undrained condition using MCM for Embankment ............. 62
4.22 Stress distribution in Normal Fill in undrained condition using MCM for Embankment ......... 63
4.23 Stress distribution in Normal Fill in drained condition using MCM Embankment, FLAC ....... 63
4.24 Stress distribution in ECS Fill in drained condition using MCM for Embankment.................. 64
5.1 Satellite Imagery of SH-360 ..................................................................................................... 65
5.2 Profile View at ECS section ..................................................................................................... 66
5.3 (a) Horizontal movement seen near V2 location (b) Testing at V2 location............................. 68
5.4 (a) V1, (b) V2, (c) V3, (d) V4, Vertical Inclinometer Readings:
Cumulative displacement vs. Depth ......................................................................................... 69
5.5 Horizontal movement on ECS section ..................................................................................... 70
5.6 Settlement readings from selected survey points .................................................................... 71
5.7 Hyperbolic plot of t (days) versus t/s (day/mm) corresponding
to the settlement records .......................................................................................................... 73
5.8 Comparison of Elevation Survey data with Hyperbolic Formulation for ECS .......................... 74
5.9 Comparison of Elevation Survey data with Hyperbolic Formulation for Normal Fill ................ 74
5.10 Vertical displacement at SH 360 - Control Section ................................................................ 76
5.11 Close view of Vertical displacement at SH 360 - Control Section ......................................... 76
5.12 Comparison of predicted and measured long term settlement using different methods at
ECS Section ........................................................................................................................... 77
5.13 Comparison of predicted and measured long term settlement results using different
methods at Control Section .................................................................................................... 78

xii

LIST OF TABLES

Table

Page

3.1 Summarizes the results obtained from both Grain Size Analysis and
Atterbergs test ..................................................................................................................... 27
3.2 Compressibility Coefficients ................................................................................................. 32
3.3 General Properties for ECS - ESCSI, 2004 ......................................................................... 36
4.1 Properties of subsoil used in Finite Element Analysis ......................................................... 40
4.2 Pavement properties used in Finite Element Analysis .......................................................... 41
4.3 Embankment Material Properties Mohr Coulomb Model, MCM ........................................ 42
4.4 Embankment Material Properties Soft Soil Model, SSM.................................................... 43
4.5 Properties of subsoil at the location ...................................................................................... 58
5.1 Hyperbolic Formulation Results ............................................................................................ 79
5.2 PLAXIS, FEM Results using MCM / SSM for Embankment ................................................. 79
5.3 FLAC, FDM Results using MCM for Embankment ............................................................... 79

xiii

CHAPTER 1
INTRODUCTION
1.1 General
Settlement is a common problem on road embankments and the cause of failure in the
road embankment is due to the movement of the pavement which results in either large
settlement or sliding due to inadequate shear strength. Erosion damage may also take place
but, since this is seldom severe enough to render the pavement unstable, it cannot be classified
as failure and is not considered in this thesis. Normal fill can comprise of mix of different soils
like silty sand, clayey sand or silty clay etc which are commonly found embankment material at
various locations in Texas. Due to climatic variations, swelling or shrinkage related soil
movements commonly occur in the soils underneath the infrastructures such as pavements,
embankments and light to medium loaded residential & commercial buildings. In pavements, the
soil movement results in settlements, surface cracking and thereby creating difficult driving
conditions and also costly rehabilitation and maintenance for the TxDOT (Briaud et al., 1997).
Many companies are spending lot of time and effort in reducing pavement distress
related to subgrade swell/shrinkage. Researchers are working to find suitable ground
improvement techniques of stabilizing embankment soils over soft soil.
1.2 Research Objective
A locally available light weight aggregate (LWA) material, known as expanded clay
shale or simply ECS, was used as a fill material for the construction of the south embankment of
State Highway SH 360 in Arlington, Texas. The application of expanded clay and shale as a
stabilizing fill material beneath the pavements is to mitigate the distress caused from
consolidation settlements of underlying foundation soils and also cyclic swell-shrink movements.

This method is expected to improve the long-term performance of these pavements by reducing
the overall settlements
Hence, the main objective of the proposed research is to evaluate the effectiveness of
ECS material in stabilizing soft subsoils which are supporting transportation infrastructure
including highways and embankments. The planned tasks for the study can be summarized as
follows:
1. To review the available literature on settlements
2. Study normal fill (treated as Control section) and ECS fill (Test Section)
3. Simulate both type of fills in Finite Element Modeling (FEM using PLAXIS 2D
8.2) and Finite Difference Modeling (FDM using FLAC 2D 4.0)
4. Simulate the long term settlements at SH 360 using different constitutive
models.
5. Studying hyperbolic method for interpreting long term settlements using
measured elevation survey data
6. Compare numerical model results with hyperbolic methods prediction results
and evaluate the performance of

ECS in mitigating settlements

of

embankments
1.3 Thesis Organization
This thesis report consists of six chapters. The first chapter explains about objective of
this research. The second chapter presents detailed review of available literature addressing
different light weight material treatment and QA/QC aspects followed by description of
instrumentation of these sections with inclinometers and Elevation survey. Chapter three
explains laboratory procedures pertaining to tests conducted on control section and test section
soil specimens. The results obtained from laboratory studies, which includes strength and other
properties of both control and treated soil properties are presented. Chapter four includes
different model studies used in numerical simulation in FEM (PLAXIS 2D) and FDM (FLAC 2D).
2

Chapter five presents a comprehensive analysis of the results from the field studies. This
includes results obtained from vertical inclinometers and Elevation survey. Here it presents the
details of field monitoring data using elevation survey and studying that data using hyperbolic
method. This Chapter also presents the comparison of hyperbolic formulation with FEM and
FDM analysis results.
Summary and conclusions from this research study, significance of the findings from
laboratory, field studies and numerical modeling and future research needs are addressed in
Chapter 6.

CHAPTER 2
LITERATURE REVIEW
A comprehensive literature search was conducted to determine the settlement
problems nationwide, current knowledge of the causes of settlement, and current mitigation
techniques.
2.1 Factors Causing Settlement
Many studies are reported in the literature (Hopkins 1969, Stewart 1985, Kramer and
Sajer, 1991) that explains the mechanisms causing the settlement in embankment on the bridge
transition. A joint study performed by University of Texas at Arlington and University of Texas at
El Paso as well as TxDOT has summarized the following factors causing settlements in the
embankments:
Consolidation settlement of foundation soil: Consolidation of foundation soil under
approach embankment is

one of the most important contributing factors to bridge settlement

(Hopkins, 1969; Wahls, 1990; Dupont and Allen, 2002). It usually occurs because of heavy
traffic loads applied at the embankment surface and to the weight of the embankment itself
(Dupont and Allen, 2002). Mostly cohesive soil represents a threat instead of cohesion-less soil.
Cohesive soil like soft or high plastic clay represents more critical situation. Cohesive soils are
more susceptible to high lateral and vertical movement. Settlements of soil are divided in to
three different phases: Initial, primary and secondary consolidation (Hopkins 1969). The initial
settlement is a short term settlement when load is applied on the surface of the soil (Hopkins
1969). The primary settlement mainly contributes to the total settlement of the soils (Hopkins
1973). There is a gradual escape of water due to the compression of loaded soil. Secondary
phase occurs as a result of change in void ratio of the loaded soil after dissipation of excess
pore pressure (Hopkins 1969).
4

 Poor compaction: Poor compaction of the embankment results in low density, high
deformation and unnecessary maintenance cost (Lenke, 2006). Compaction mechanically
increases the density, stability and reduces water seepage, swelling and contraction.
Poor drainage and soil erosion: The dysfunctional, damaged and blocked drainage systems
cause erosion and void development which can cause serious settlement problems
(Jayawickrama et al., 2005). The rain water can seep in to the embankment fill via cracked
pavement section or faulty joints and can soften the fill and create internal erosion.
Traffic volume: Lenke (2006) noted that the settlement was found increasing with vehicle
velocity, vehicle weight, heavy truck traffic and number of cycles of repetitive loading.
Types of Abutment: A number of researchers Schaefer and Koch 1992; Briaud et al. 1997;
Hoppe 1999; Abu-Hejleh et al. 2006 considered lateral movement of the bridge abutment and
settlement of the embankment amongst the primary problem. The lateral movement affects the
integral abutment bridges much more severely than non-integral abutment bridges. In integral
bridges, as the temperature increases, the bridge super-structure moves the abutments toward
the retained soil causing high lateral stresses, which can reach stress levels as high as the
passive pressure limit (Schaefer and Koch 1992). As the temperature decreases, the abutments
move away from the soil, creating a void between the abutment and backfill material. The
presence of the void can intensify soil erosion, increasing the size of the void under the
approach slab (Schaefer and Koch 1992; Wahls 1990, Greimann et al. 1986) reported that the
lateral movement of the bridge that occurs in integral abutment bridges also affects the pile
stresses and can potentially reduce the vertical load carrying capacity of the piles. The U-type
abutment are the closed abutments having two side walls and front wall resting on spread
footing below natural soil. The compaction of embankment fill is difficult in these abutments and
is subjected to higher lateral movement. A great number of cantilever and stub abutments have
been constructed in Texas. Stub abutments are economical but have maintenance problems
(TxDOT Bridge Design Manual, 2001). The settlement at beginning of the bridge has been
5

noticed on stub while cantilever abutment has shown good performance but has constructional
problem and high construction cost.

2.2 Remedial Measures to Reduce Settlement

Construction of structures like embankments and retaining walls over soft soils was
always a challenging ground engineering problem due to high settlements experienced by these
structures. The ground improvement techniques including stone columns, grouting, dewatering,
preloading, vertical drains and in situ mixing of admixtures like lime and cement have been used
to improve bearing capacity and reduce settlements of soft soils (Wahls, 1990; Bergado and
Patawaran, 2000; Barksdale and Bachus, 1983; Porbaha, 2000).
Wang and Huang, 2006, compared four ground and embankment treatment techniques
which included embankment reinforcement, light weight embankment, cement deep mixing piles
and separating walls, for uneven settlement, horizontal displacement and stress distribution and
found that light weight embankment has more effectiveness among them. There are different
types of lightweight geomaterials available in the Market.
Miki (1996) specifies the following applications of the lightweight geomaterials:
Reducing residual settlement of embankment built on soft ground
Minimizing differential settlement between approach embankment and structure
Minimizing deformation when constructing near adjacent structure
Faster construction period
Low maintenance infrastructure
The following section discusses different types of lightweight embankment fill materials.

2.2.1 Light-Weight Expanded Polystyrene, EPS

Polysterene also known as EPS is a foamed plastic

and is frequently utilized as a

lightweight packaging material and as a thermal insulating material is used in construction

applications (Thompsett et al., 1995).
A case study near Oslo in 1972, used expanded polystyrene to raise the road level of a
highly settled road. On top of the expanded polystyrene, a new 500 mm pavement was laid.
During a period of next 12 years, only 80mm settlements were recorded and almost zero
settlement since then (Thompsett et al., 1995).
Another example where the EPS material was used for an embankment is reported on
the Lokkerberg Bridge near Swedish border. A miniscule 6 cm deformation was observed after
12 years and majority of that occurred during construction period (Frydenlund et al., 2001).
Figure 2.1 shows the longitudinal profile of the EPS embankments of the Lokkerberg Bridge and
Figure 2.2 presents the settlement information for the same bridge.

Figure 2.1 Longitudinal Profile of EPS Embankment at Lokkeberg bridge, Swedish Border Fredyland and Aaboe, 2001

The use of expanded polystyrene blocks for lightweight fill has now become standard
practice in road construction and road widening projects in Europe, Asia and North America.

Figure 2.2 Deformations in EPS Embankment at Lokkeberg - Fredyland and Aaboe, 2001

Expanded polystyrene is resistant to common inorganic acids and alkalis but not
resistant to organic solvents including petrol and diesel oil. For protecting the EPS various
methods are used like clay fills, polyurethane coatings and impermeable membranes.
(Thompsett et al., 1995).
The success rate of EPS as an embankment material is inconsistent. There have been
five reported failures and the general causes have been due to water fluctuations, buoyancy
and fires.
2.2.2 Light Weight Fill from a Mixture of Fly Ash and Bottom Ash
Fly ash and bottom ash are coal combustion by-products. These two by-products are
generated in large quantities throughout the world. In USA, about million tons of coal
combustion by-products are produced every year. Figure 2.3 illustrates the disposal method of
Fly and bottom ash mixture. To reduce the environmental impact of unused Fly ash, different
applications have been identified. One of the applications is to use it as a fill material in
embankments (Patelunas 1986; Srivastava and Collins 1989).
8

Figure 2.3 Disposal pond of fly and bottom ash mixture generated at Wabash River Power Plant
Sungmin Yoon et al. 2009

A case study in Terre Haute, Indiana was conducted to demonstrate the effectiveness
of ash mixture as fill material in embankment (Sungmin Yoon et al., 2009). The ash mixture was
classified as sandy silt (ML) by USCS. After the embankment was constructed it was monitored
for a period of one year. Figure 2.4 shows a cross sectional view of the demonstration
embankment.

Figure 2.4 Cross section of an embankment at State Road 641,Terre Haute, Indiana - Sungmin
Yoon et al. 2009

The study showed a maximum settlement of 80 mm at the bottom of the embankment

and the settlement started to stabilize in 5 months after the embankment was constructed.. At
the top of the embankment, a differential settlement of about 5 mm was observed while
negligible lateral movements were observed at shoulders and toe of the embankment. Sungmin
Yoon et al., concluded that the fly and bottom ash mixture used in the construction of the
embankment is a viable alternative to conventional fill materials.
Environmental impact studies showed that the leaching of trace metals from fly and
bottom ash is a main concern as it can contaminate the ground-water (Alleman et al., 1996).
Longer term studies are necessary for more definite conclusions on the environmental impact of
ash embankment construction.
2.2.3 Lightweight Tire Scrap
Heavy traffic volume across the world has an unintended by product- scrap tires and
researchers are putting lots of effort in finding beneficial reuse of this material. Tuncer and
Bosscher, 1994, studied the reuse of discarded tires as a fill and drainage material and
observed that the tire chips constrained deformation modulus in elastic range is 1/100 that of
pure sand, have high hydraulic conductivity and improves the frictional respose of sand in
mixtures. Figure 2.5 presents a photograph of typical tire scrap.

Figure 2.5 Scrap tire beads - An Deng and Jin-Rong Feng, 2009

10

Aderinlewo and Okine 2008, observed that the tire scrapes have useful properties like
high permeability, low thermal conductivity and is lightweight but also has the disadvantages of
large settlement, potential for internal heating, and leachate effects.
In a case study in Portland, Maine the tire shreds were used as lightweight fill for
construction of two 9.8-m high highway embankments (Humphrey et al. 1998). Tire shreds were
chosen because they were cheaper and reused the 1.2 million scrap tires. Wick drains were
used to accelerate consolidation of the foundation soils.
The tire shreds were placed on top of the geotextile on the prepared base. After placing
the shreds, geotextile separator was placed on the sides and top of the tire shred zone and then
the surrounding soil cover. Low-permeability soil was placed on the outside and top of the fill to
limit inflow of air and water. The embankment was topped with granular soil and temporary
surcharge.

Figure 2.6 Cross section through embankment for the Portland Jetport Interchange - Humphrey
et al., 1998

At each tire shred layer settlement plates were installed to monitor settlement. The
study concluded that the tire shreds had a high factor of safety against slope instability and they
are more economical alternative as a low weight fill material.

11

2.2.4 Lightweight Aggregate

In the decades following WWII, USA has enjoyed strong social, economic, industrial,
and urban growth, which has naturally led to the building of new, modern communication
infrastructures between centers of development and on their outskirts. This development boom
led to building of roads and railway lines over weak geotechnical formations like soft soil
deposits. This prompted substantial research in material science related with lightweight
aggregates.
Lightweight Aggregate (LWA) is described as a range of special use aggregates that
have an apparent specific gravity considerably below ordinary sand and gravel (ACI-213, 1967).
Lightweight aggregates range from extremely light materials used for insulative and nonstructural concrete (Vermiculite and Perlite) to expanded clay shale (ECS) used for high
performance structural concrete. However the strength of this material is considerably low but
has very good insulation properties, high elasticity, absorbs sounds and other vibrations (ESCSI
1994). Use of LWA is environmental friendly as they have lower transportation requirements
and use raw materials that have limited structural applications in their natural state. This
minimizes demands on finite resources such as quality natural sands, stones and gravels (Ries
and Holm, 1993).
There are several types of LWAs, which include granulated / foamed blast-furnace slag,
lightweight expanded clay aggregate (LECA), expanded clay or shale (ECS), furnace bottom
ash (FBA), pulverized fuel ash (PFA), or the less common pumice (a volcanic material) and
vermiculite. However, this study focuses on expanded clay and shale, which will be referred to
as lightweight aggregate (LWA). The advantages of LWAs over conventional aggregates
include high internal angle of friction, increased thermal insulating capacity, freeze-thaw
resistance, and low unit weight. Lightweight aggregate fills are approximately half the weight of
normal fills. It has been effectively used to solve numerous geotechnical engineering problems
and to convert unstable soft soil into usable land (Holm and Valsangkar, 1993).
12

Several states use LWA in asphalt surface treatments. The advantages of using LWA in
asphalt surface treatments are high skid resistance, extension of roadway life, increased
freeze/thaw resistance, increased resistance to stripping, and considerable reduction of
windshield, headlight, and paint damage from flying stones. One disadvantage mentioned is
that LWAs can roll out of the asphalt under constant breaking or turning conditions (Gallaway,
1997). Texas was the only state that appeared to have conducted considerable research into
LWA for highway uses. Moore (1970) investigated the use of LWAs in flexible bases. The study
developed the pressure-slaking test and the modified pressure-slaking test for testing the
aggregates suitability. The test consists of underwater pressure-cooking followed by severe
agitation. The study concludes that a synthetic aggregate is acceptable for use in Texas flexible
bases when the loss in gradation from the pressure-slaking test is less than 10 percent or less
than 4 percent in the modified pressure-slaking test. In addition, the study concluded that
synthetic aggregates suitable for flexible bases in Texas could be produced from clay in rotary
kilns (Moore and William, 1970).
Studies were conducted in the Nordic countries concerning LWA use in road
construction. However, many papers have not been translated in to English. According to one
Finnish/Swedish research group, expanded clay (Exclay, a type of lightweight aggregate) has
two main road-related uses, either for frost insulation purpose in frost susceptible soils or for
lightening of road embankments on soft soils to prevent settlement (Gustavsson et al., 2002).
2.2.4.1 Lightweight Expanded Clay Aggregate
Due to abundance of clay which is a raw material for lightweight expanded clay
aggregate (LECA), its use is becoming very popular around the world (Rossignolo at al., 2003).

13

Figure 2.7 LECA Granular - Arioz et al., 2008

LECA is produced by baking natural clay in a kiln at a temperature of1000-1200 C
which introduces swelling in it due to the expanding gases inside the mass (Mladenovic, et al.
2004). The high porosity contributes to the light weight nature of LECA.
Excellent frost insulation property of Expanded clay LWA makes it highly suitable to use
in road pavements which have frost susceptible soils. The other important benefits of LECA are
use as light weight fill for stability improvement and reduction of settlement.

Figure 2.8 Schematic of lightweight embankment system to reduce settlement

2.2.4.2 LWA from Mining Residues, Heavy Metal Sludge, and Incinerator Fly Ash
Huang et al, 2006, experimented with the mining residues, heavy metal sludge and fly
ash to create a new type of LWA. The composition of the mining residues which is extracted
from a mine in Hualien, Taiwan resembles that of shale. The fly ash used is from the municipal
14

waste incinerator in Taipei, Taiwan. Heavy metal sludge, was taken from a printed circuit board
factory in Taiwan. These sludge materials comprise heavy metal hazardous wastes including Cr
and Cu (Huang et al, 2006).
The study illustrated the manufacturing process which included raw material
preparation, drying, pulverizing, mixing, pelletization, sintering and cooling (Huang et al, 2006).
The analysis of composition showed that the chemical composition of the pellets matches the
chemical composition of expandable LWA as shown in Figure 2.9.

Figure 2.9 Ternary plot of LWA - Huang et al., 2006

0

The study concluded that above sintering temperature of 1150 C, these aggregates
become non-hazardous. This new method of utilizing hazardous waste products for creating
LWA not only combines the benefits of light weight aggregates but also provide a safe
ecological disposal of these waste products.
2.2.4.3 ECS as Lightweight Aggregate
Expanded clay and shale lightweight aggregate due to its property of high strength and
low unit weight makes it a reliable and economical geotechnical solution for landfill leachate
drainage systems, high-rise building, underground conduits and pipelines for insulation or in
unstable soil, insulating backfill and insulating road base, retaining walls and embankments
(Holm and Valsangkar 1993 & 2001)..
15

The benefits of ECS can be summarized as (www.escsi.org):

Reduces dead loads
Reduces lateral forces
Reduces overturning forces
Provides a high angle of internal friction.
Free draining
Water and acid insoluble
High insulation value
Chemically inert, and environmental friendly
High strength and durability
Readily available
Controlled gradations
Easy to handle and install
Cost effective
ECS is prepared by expanding select minerals in a rotary kiln at temperatures over
1000 C (Holm and Ooi, 2003). The production and ra w material selection processes are strictly
controlled to ensure a uniform, high quality product that is structurally efficient, durable and
inert, yet also lightweight and insulative (Holm and Valsangkar, 1993 & 2001).

16

Figure 2.10 ECS Lightweight Aggregate - Saride et al. 2008

The initial cost of ECS lightweight aggregate is usually higher than a comparable
normal weight aggregate but the cost offsets when the net savings from long term use is taken
into perspective. The cost savings are obtained in the following areas: labor, lower dead loads,
better fire resistance resulting in reduced concrete thickness, and less reinforcing required in
building frames, girders, piers, and footings (Holm and Valsangkar, 1993 & 2001). ECS is not
only economical in long run but also a environmentally sustainable material. Taking these
benefits into consideration, ECS has found a huge favor in construction industry today.
2.3 Numerical Modeling
A numerical simulation/modeling of different construction processes is important to
ascertain the safety of current and planned construction (Zimmerman et al., 2005). Numerical
modeling software offers a unified and generic framework for such computations such as
analysis of deformations, stresses, thermal behavior and continuous stability assessment in
soils, rocks and structures, including soil structure interaction. The softwares have evolved and
now allows us to perform the modeling in 2D and/or 3D. All the stages of construction i.e. from

17

the initial stage to the final construction can be simulated in a single environment (Zimmerman
et al., 2005).
According to Joaquin and Sherwin (2005) there are three important steps in the
computational modeling of any physical process: (i) problem definition, (ii) mathematical model,
and (iii) computer simulation. The first step is to define the problem in terms of a set of relevant
measurable parameters. The second step is to represent the idealization of the physical reality
by a mathematical model. These mathematical models use numerical time-stepping procedure
to obtain the models behavior over time. After establishing the suitable boundary and initial
conditions, the third and final step is to proceed to its solution using computer simulation.
(Joaquin and Sherwin, 2005).
2.3.1 Modeling for Settlements
The settlement of embankments on soft soils is traditionally an important geotechnical
problem and has been extensively studied by a large number of researchers. Excessive
settlement can render the highways unserviceable, increase the cost of maintenance, and even
be detrimental to the stability of embankment. Therefore it's very important to accurately
evaluate the embankment settlement during the highway construction design (Murray, 1971).
The soft soil properties of high compressibility and low permeability introduces large
embankment settlement over time as excess pore pressure dissipates embankment loading is
applied. (Biot, 1941; Bowles, 1979; Borges, 2004).
For predicting the behavior of embankment on soft ground, one of the key point is to
simulate the consolidation process. The consolidation rate is mainly influenced by the
foundation soil permeability. The permeability of soft ground varies during the loading and
consolidation process, and significant changes occur before and after the soil yields (Tavenas
et al. 1978, 1980).
Available methods for calculating embankment settlement are the layer-wise
summation method (LSM), empirical formulation method, finite element method (FEM), etc.
18

(Qian and Yin, 1996; Wang, 2004). Of these methods, FEM is a very powerful numerical tool for
solving complicated 2D or 3D consolidation settlement problems. It can handle arbitrary
boundary conditions, different loading schemes and it considers the coupling effects of loading
and soil consolidation (Zienkiewicz and Taylor, 1989; Taiebat and Carter, 2001; Wang, 2004).
For modeling long term settlements in this case study, both FEM and FDM method
have been used. PLAXIS software utilizes FEM while FLAC software utilizes FDM.
2.3.2 Differences between Finite Element Method and Finite Difference Methods
A finite difference method (FDM) discretization is based upon the differential form of the
Partial Difference Equation to be solved. It utilizes a point-wise approximation to a solution. The
domain is discretized into a grid of hexahedral cells or nodes. The solution will be obtained at
each nodal point. Although FDM is easy to implement and the compute time for each step is
fast, however the number of steps required for convergence is high. The other disadvantage is
that the domain is not accurately represented if the domain is discontinuous or non-rectangular
in shape (Fausett, 2003). Fig 2.11 (a) illustrates the FDM mesh.
A finite element method (FEM) discretization is based upon a piecewise representation
of the solution in terms of specified basis functions. In FEM the discretization is not restricted to
a grid of hexahedral cells or nodes, instead, a solution is approximated using interconnecting
sub-regions or elements. These elements are typically simple geometrical figures as illustrated
in Fig 2.11 (b). This flexibility of construction of elements in FEM allows it to accurately model
the complex geometries. The downside is that FEM is difficult to implement but the opinions
vary on this (Zienkiewicz and Cheung, 1965).

19

(a)

(b)

Figure 2.11 Discretization methods (a) FDM (b) FEM

2.4 Instrumentation
The main objective of this section is to present a brief review on instruments used in
this study. Instruments provide actual field data and helps in evaluating the movement and
failures under actual field conditions. The instrumentation was used in this study to compare the
performance of the LWA vis-a-vis normal fill. The choices of instruments depend on many
factors like geological understanding of the area, subsurface material and the groundwater
depth. The other factors include the transportation facility, the rate and magnitude of movement
and type of movement i.e. horizontal or vertical movement or both. For large and fast
movements relatively crude instrumentation can be used but if the movement is slow and small
instrumentation accuracy and the repeatability of its measurement takes precedence (Machan
and Bennett, 2008)..
2.4.1 Characteristics and Instrumentation Details
Field instrumentation is very important in geotechnical engineering and therefore,
geotechnical engineers should have proper knowledge of instrumentation. But instrumentation
is not answer to everything so its use must be prudent. The wrong type of instruments or wrong
placement of instruments can provide wrong or confusing results which can divert the attention
from real problems (US Army Corps of Engineer, EM 1110-2-1908).
20

The objectives of a geotechnical instrumentation plan are grouped into four categories:
analytical assessment; prediction of future performance; legal evaluation; and development and
verification of future designs.
2.4.1.1 Analytical Assessment
The following objectives are met during the analytical assessment (US Army Corps of
Engineer, EM 1110-2-1908).
(1) Verification of design parameters Observed data from instrumentation can be
used to not only verify the selected design parameters but also used to modify and refine the
future designs.
(2) Verification of design assumptions and construction techniques The data obtained
from satisfactory actual performance of new or modified design can help its chances of
acceptance.
(3) Analysis of adverse events Causes of various types of failures and deformations
at the project site can be uncovered from the precious instrumentation data.
(4) Verification of apparent satisfactory performance The instrumentation data of
satisfactory performance as well as adverse events can prove to be valuable records for future
design purposes and development of new and innovative technology.
2.4.1.2 Prediction of Future Performance
Instrumentation data analysis can be used for the predictions such as continued
satisfactory performance or indicating a sign of potential future distress. The instrumentation
data recorded during and after events like rainfall can be very useful for future performance (US
Army Corps of Engineer, EM 1110-2-1908).
2.4.1.3 Legal Evaluation
Instrumentation data can be used in determining causes or extent of adverse events so
that the merit of various legal claims can be evaluated (US Army Corps of Engineer, EM 11102-1908).
21

2.4.1.4 Development and Verification of Future Designs

Analysis of the performance of existing normal fill and LWA fill embankments and
instrumentation data generated during operation, can be used to further the construction
technology. Instrumentation data from existing embankment construction projects can provide
safer and economical design for future construction of embankments (US Army Corps of
Engineer, EM 1110-2-1908).
2.4.2 Different Types of Instruments Used in the Field Monitoring
There are various field instruments like Extensometers, Inclinometers, Piezometers,
Soil Stain Meter, Pressure cells and settlement cells to measure soil movements, settlements,
strains, deformations, pore pressures in the soil. They also warn us about the design
assumptions suitable for the given field conditions. The following sections present the details of
instruments used in the current study and the installation procedures followed.
2.4.2.1 Inclinometers
Inclinometers are used to measure the displacement normal to the axis of a pipe. The
inclinometer probe which contains a gravity sensing transducer is inserted into the casing pipe.
This transducer is designed to measure inclination with respect to the vertical axis. The casing
pipe may be installed either in the borehole or in fill, in most cases in a near vertical alignment,
to provide horizontal deformation. Due to the sensitive nature of the probes, they should be
handled carefully during storage and transportation (Dunnicliff, 2003).

Figure 2.12 Details of inclinometer casing and assembling procedure - Slope Indicator, 1997

22

2.5 Surface Elevation Surveying

The surface surveying consists of establishing benchmark control stations and survey
points at various locations of interest. The survey points are periodically measured to monitor
for potential deformations at test section and control sections. For this technique, the control
stations are established on stable ground. Surface monitoring is accurate to a resolution of
approximately 0.25 in. (6 mm). However, long term deformation trends can be difficult to
establish, due to personnel turnover, susceptibility to damage of control points, etc. This method
provides essentially provides the relative movement of the pavement surface relative to its
initial position (i.e. settlement or heave).
2.6 Visual Monitoring
This approach consists of occasional on-site inspections. Field inspection is a good
technique to identify signs of distress or change. However, visual observations are typically not
accurate method in detecting small movements or long-term settlements. Periodic good
resolution photographs may be a good idea for comparison.
2.7 Summary
This chapter discusses various causes of settlement, its remedial measures, different
types of fill materials and finally the instrumentation used in determining the horizontal and
vertical movement. It is observed that among all the different types of embankment fill materials,
ECS is the most suitable for Texas since it is easily produced and very economical. For this
study Vertical Inclinometers were used to determine horizontal movement and theodolite for
vertical surface movement.

23

CHAPTER 3
EXPERIMENTAL STUDIES
Laboratory testing is an essential aspect of geotechnical engineering. Laboratory test
are conducted using ASTM standard procedures. Material properties like cohesion unit weight,
internal friction angle etc can be determined from different laboratory tests. Visual observation
and field test results are the primary source of information before doing laboratory testing.
Laboratory test results provide a clear understanding and knowledge of soil properties.
3.1 Quality Control for Laboratory Testing
Quality control for Lab Testing involves proper handling of samples while transporting
and storage, Its important to maintain quality otherwise it can result in misleading test results. .
In addition, the following guidelines given by Mayne, et al., (2002) for laboratory testing of soils
was followed.
3.1.1 Tests Conducted
To study the behavior of LWA and Normal fill, Laboratory tests namely, sieve and
hydrometer test, Atterberg limits tests, proctors compaction tests, direct shear tests, Swell
tests, and Consolidation tests were performed.
The soil used for various tests were collected from site and then wrapped in a plastic
bags and transported to the laboratory.
3.1.1.1 Sieve Analysis and Hydrometer Analysis
Sieve analysis was done as per ASTM D22. There are two different procedures for dry
and wet sieving. For our study, the sample collected was in the form of hard lumps and the
sample was pulverized to very fine particles before doing dry sieve analysis. For normal fill
material, first dry sieving was performed, where the sample was oven dried and allowed to cool.
An oven dried sample of 1000 grams passing through 4.75, 2.0, 1.8, 0.6, 0.425, 0.25, 0.125,
24

0.075 mm sieves was taken. The pan was attached at the bottom of the sieve stack. The
sample was poured on the top sieve and stirred for about 10 minutes. Soil retained at each
sieve was measured and weighed. The weights of the sample on all the sieves were added to
compare it with initial sample weight. The difference shouldnt be more than 1%.
Secondly, wet sieving was performed where oven dried sample was kept soaked in a
tap water for about 2 hours. The sample was transferred to 200 sieve. The sample was washed
thoroughly, discarding the material passing no. 200 sieve. The retained material collected from
no. 200 sieve was oven dried and weighed it after it has cooled. Difference between dry weight
before and after washing was around 50% recorded.
Hydrometer analysis was performed on the same sample which was finer than no. 200
sieve size. The lower limit of the particle size determined by this procedure is about 0.001mm.
Hydrometer analysis test procedure was adopted as per ASTM D-422.
For ECS, the material had less than 5% fines, so wet sieve analysis was not done. An
oven dry sample of 500 grams passing through 4.75, 2.0, 1.8, 0.6, 0.425, 0.25, 0.125, 0.075mm
sieves was taken. Soils retained on various sieves were calculated and a particle distribution
was obtained. Figure 3.1 shows the particle distribution curve for the ECS materials.

Figure 3.1 Particle Size Distributions for ECS

25

3.1.1.2 Atterberg Limits

For classification of soil, Atterberg Limits tests (Liquid Limit and Plastic Limit) were
performed on the collected sample. The liquid limit test of a soil was performed using
casagrande liquid limit apparatus. About 250 grams of air dried sample passing sieve no. 40
was used for both liquid and plastic limit test. Distilled water was preferred instead of tap water
to avoid ion exchange between soil and water impurities, which may affect the soil plasticity.
Place about 50 grams of soil paste in a cup, level off with the spatula the top surface
symmetrically to give maximum depth of 1 cm. The grooving tool was used to straight groove
through the soil paste along the diameter through the center of the hinge. The handle was
turned at a rate of 2 revolutions per second and counted the number of blows until the two parts
of the soil come in contact at the bottom of the groove. About 15 grams of soil was transferred
in the container to determine the water content by oven drying. The test was repeated at least 3
to 4 times. The flow curve was plotted to represent the number of blows on logarithmic scale
and corresponding moisture content.
The plastic limit test was done to determine plasticity of the soil. Here 30 grams of soil
passing sieve no. 40 was used. Distilled water was used to mix water thoroughly in to the soil.
10 grams of plastic soil mass was used to form a ball and then roll in to thread with the fingers
on the ground glass plate. Keep on rolling until the thread starts to crumble at a diameter of
3mm. The crumbled thread was kept in a container for moisture content determination. The
processes were repeated 2 more times with fresh sample and the average of three moisture
contents was obtained to calculate plastic limit. After determining liquid limit and plastic limit, the
plasticity index was calculated and plotted on the plasticity index chart to know the type of soil.
Figure 3.2 presents the test results obtained for normal fill.

26

Table 3.1 Summarizes the results obtained from both Grain Size Analysis and Atterbergs test
Material

Liquid

Plastic

Plasticity

USCS

Type

Gravel

Sand

Silt

Clay

Limit

Limit

Index

Classification

0.2

44

40

15.8

35

14

21

CL

0.2

41

35

23.8

36

15

21

CL

NF-1
(North
Side)
NF-2
(North
Side)
ECS

65% Coarse Sand

(South

25% Medium Sand

Side)

10% Fine Sand

Figure 3.2 Plasticity Chart for Normal Fill

27

3.1.1.3 Compaction Test

Compaction characteristics were determined according to ASTM D698 (Standard
Procter Test). Laboratory compaction tests are used to determine the relation between water
content and dry weight and to find the maximum dry unit weight and optimum water content.
For each sample the required amount of sample was air dried and weighted and the
mass of material required is around 3kg passing no. 4 sieve. A suitable amount of water was
added to the dry soil to obtain the desired moisture content. The soil sample was evenly
distributed so that the mold is about half full. The sample was compacted by applying 25 blows
of the rammer dropped from a controlled height of 300mm and the rammer was positioned
properly before releasing. The soil sample is evenly compacted in to 3 layers following the same
procedure. The extension collar and the base plate were removed carefully and weight of soil
and mold was taken. The soil was extracted from the mold using extractor and immediately
three representative samples were taken from the soil to determine the water content. The
process was repeated to obtain more compaction points. Figure 3.3 shows the compaction test
data comprising of optimum moisture content and dry densities for the normal fill materials.

28

Figure 3.3 Compaction Curve for Normal Fill

3.1.1.4 Direct Shear Test
A series of direct shear tests were performed on ECS material according to ASTM D
3080 method. ECS samples were well compacted and were placed in a 2.5 inch shear box.
Normal stresses of 50, 100 and 200kPa were applied on the samples and then the samples
were sheared. Normal stresses were applied with the help of a loading ram. Shearing was
applied with the help of a horizontal ram. This set up was connected to the computer where all
the graphs were plotted between shear stress and shear displacement. Figure 3.4 and 3.5
shows the graph between measured shear stress and applied normal stress. The calculated
peak friction angle of the ECS is 49.5 and the cohesion is 75 kPa. The same for normal fill are
85 kPa and 18 degrees respectively.

29

Figure 3.4 Direct Shear Test results for ECS material

Figure 3.5 Direct Shear Test results for Normal Fill Material
30

3.1.1.5 Consolidation Tests

Consolidation tests on the two materials, Normal fill and ECS were performed. The test
was conducted according to ASTM D2435. The soil sample is kept inside a stiff steel ring which
blocks its lateral expansion and two porous stones allow the water drainage. The sample is fully
submerged in water and remains fully saturated during the test. Sample was prepared in a 2.5
inch diameter and 1 inch in height ring and was then placed for testing in an oedometer. An
initial sitting pressure of 35kPa was applied on the sample and then the sample was
consolidated for 24 hours. After 24 hours of consolidation, the initial sitting was applied such
that it shows no swelling. Dial gauge reading was noted under initial sitting pressure. The first
increment of the load 1000psf was applied and the readings were recorded at certain time
intervals for 60 minutes. The particular sequences of loading and unloading of applied pressure
are 1000, 2000, 4000, 8000, 16000, 32000, 16000, 8000, 4000 and 1000 psf. This range of
applied pressure completely covers the effective stresses that are needed for settlement
calculations. This range encompasses the smallest and largest effective stresses in the field.
This experimental result shows the response of different materials subjected to loadingunloading cycles. Here two materials ECS and Normal fill, behaves differently during loading
and unloading phases.
Figure 3.6, 3.7 and 3.8 shows the graphs between void ratio versus logarithm of
effective pressure for the ECS material and the Normal fill. Compression index Cc for the ECS
material was found to be 0.14. The lower compression index value for the ECS material was
expected as its a granular fill material which is around 0.02 to 0.05. Table 2 gives the values for
the compression and recompression indices for the both materials. From the table, Cc for
Normal fill material has a Cc value of 0.12 which indicates that it is low plastic soil and with high
compressibility which may result in high settlements on the pavement surface and can cause
undulations.

31

Table 3.2 Compressibility Coefficients

Compressibility Coefficients

Normal Fill

ECS

Compression index, Cc

0.12

0.14

Swelling Index, Cs

0.033

0.009

Figure 3.6 e log (p) graph for Normal Fill material

32

Figure 3.7 e log (p) graph for Normal Fill material

Figure 3.8 e log (p) graph for ECS material

33

3.1.1.6 Vertical Swell Tests

One-Dimensional Free Swell Test measures the amount of heave in the vertical
direction of a laterally confined specimen in a rigid ring. The swelling characteristics of these
materials are important to understand the swelling related movements due to climate
fluctuations. To understand the swelling nature of both of the materials, vertical free swell tests
were conducted on both the ECS and normal fill samples as per as per the ASTM D4546. A
conventional oedometer steel ring of size 64 mm (2.5 in.) in diameter and 25 mm (1 in.) in
height was used. The inner face of the consolidation ring is lubricated to minimize the friction
during free swell.
Readings were taken for 24 hours and the swell strain were calculated. The magnitude
of the vertical free swell strain for the normal fill was found to be 2.5% and for the swell strain, it
was observed to be negligible for the ECS material. This confirms that the normal fill soil is a
moderate swelling material whereas the ECS is a non-swelling material.

Figure 3.9 Free Swell test graph for Normal Fill

34

3.1.2 Selection of Design Properties

The lab test results cannot be trusted completely before doing a comprehensive
assessment of quality and consistency of the data. Only if the data appears to be consistent
with the expectations it should be accepted else the inconsistencies and poor data should be
identified and eliminated and then retesting should be done (Geotechnical Design Manual,
2010).
Past knowledge and experience of the area surrounding the test site also comes in
handy, since it helps the geotechnical designer to make educated engineering judgement while
selecting an average, typical or design value for the selected property.
Normal Fill Material. The normal fill materials are likely to be clay with low plasticity (CL)
and contain enough fines to be moderately moisture sensitive. Local material is not all weather
material.
Direct shear strength testing on local material that meets gradation requirements results
in undrained friction angles of 18 degrees and cohesion value of 85kPa.
The common normal fill material can be of CL, ML, SM, GM, SP-SM, GP-GM with unit
3

weight of 17 to 21kN/m and a friction angle between 15 to 38 degrees. Normal fill with fines
content may sometimes be modeled as having an apparent cohesion value from 5 to 95kPa. If a
cohesion value is used, the friction angle should be reduced so as not to increase the overall
strength of the material. For long term analysis, the normal fill material should be modeled with
less cohesive strength.
Low Weight Aggregate. These materials are often used as fill material for
3

embankments. For design purposes, typical values of 6 to 11kN/m for the unit weight and
internal angles of friction of about 40 to 50 degrees as an undrained parameter should be used.
ECS aggregates provide a practical, reliable and economical geotechnical solution (DeMerchant
and Valsanger, 2002). Table 3.3 shows the general engineering properties of ECS (after
ESCSI, 2004).
35

Table 3.3 General Properties for ECS (ESCSI, 2004)

Aggregate

Measuring

Test

Commonly

Typical for

Typical

property

method

method

specifications

ECS

design values

for ECS

aggregates

for ordinary
fills

Soundness

Magnesium

AASHTO

Loss

Sulphate

T 104

Abrasion

Los Angeles

ASTM

Resistance

Abrasion

C 131

Compacted

Density Test

ASTM

Bulk Density
Stability

<30 %

<6 %

<6 %

<40 %

20 40%

10 45%

<70 lb/ft

40 65 lb/ft

100-130lb/ft

35 - 45

30 - 38

D 698
Direct Shear

ASTM D

According to

Test &

3080 &

project

Triaxial (CD)

Corps of

sand &

engineers

gravel)

(fine sand-

EM 1110-21906
Loose Bulk

Loose

ASTM C 29

Density

Dry <50 lb/ft

Saturated<65

Dry 30-50
3

lb/ft

89-105
3

lb/ft

lb/ft
pH

pH meter

AASHTO T

5 10

7 10

5 - 10

289

Soft Soils Material.

Soft soils are characterized by very low strength, very high

compressibility and having very important time-consolidation effects. These soils pose special
challenges for the design of engineering transportation projects and therefore require careful
consideration regarding settlement and stability of embankments.

36

Vertical settlement is also a major concern while constructing embankments on soft

soils. Estimation of settlement and the time period for it to occur can be derived from laboratory
consolidation tests. Secondary compression must always be evaluated when estimating longterm settlement. Compression index values based on vertical strain (Cc) typically range from 0.1
to 0.3 for organic silts and clays, and are generally above 0.3 to 0.4 for soft clay. The coefficient
of secondary compression (C) is typically equal to 0.06 x Cc for soft clay respectively
(Geotechnical Design Manual, 2010).
3.2 Summary
This chapter discusses different types of engineering tests to ascertain the properties of
ECS and Normal fill materials. Typical tests that were performed are sieve analysis and
hydrometer test, Atterberg limits test, compaction test, direct shear test, consolidation test and
swell test. The quality and consistency of the laboratory test data was reviewed and determined
if the results are consistent with expectations.

37

CHAPTER 4
NUMERICAL MODELING
The use of numerical modeling to predict soil deformation and stresses has been
practiced for years. During the numerical analysis, detailed site-specific properties of the road
and embankment systems were incorporated to simulate complex construction sequences. The
main field of application of constitutive models is the execution of numerical calculations by
means of appropriate methods such as Finite element or Finite difference methods.
4.1 Modeling Methods
Numerical modeling using linear model has a benefit of fast estimate of the material
response but the downside is its limited accuracy. Therefore the application of linear model is
limited to cases where the stress or deformation states of a soil mass are of interest. For getting
reliable description of the soil behavior it is necessary to employ nonlinear models (GEO5 Theoretical Manual, 2010).
The non-linear models can be divided into two groups. The first group of models
originates from the Mohr-Coulomb failure criterion. The Mohr Coulomb model belongs to this
group. According to Coulomb C.A., 1776, the MohrCoulomb failure criterion represents the
linear envelope that is obtained from a plot of the shear strength of a material versus the applied
normal stress. The second group of material models is represented by the Modified Cam-Clay
model. This model is based on the concept of critical state of soil (GEO5 - Theoretical Manual,
2010). Accurately modeling the constitutive behavior of the soil is difficult because of the
complexity involved in the selection of design parameters and the soil properties. Simplified
non-linear models such as Mohr-Coulomb, or more advanced such as Modified Cam-Clay and
the Hardening Soil Model can be used with some degree of accuracy (Ng and Lings, 1995). .

38

4.2 Finite Element Modeling Procedure

Numerical modeling enables the designer to study the effects of embankment loading,
surcharge and the soil behavior in various conditions without resorting to simplified
assumptions. Two dimensional finite element modeling of the different fill embankments is
performed using PLAXIS 2D software.
A parametric study was performed to arrive at the critical parameters that define the
behavior of the system. The parametric study comprised of all the elements which had an
influence on the behavior of the system. The various aspects like lateral movements and the
total settlements were studied. The standard units length (m), force (N) and time (days) were
used. The geometry was drawn using geometric lines and standard fixities were then used to
define the boundary conditions.
The properties of different soil material sets were created and assigned to material
model. After the model was created and material models were assigned, finite mesh was
generated using different mesh settings. The following sections present the details of the finite
element model including the choice of material models, finite element mesh, and boundary and
loading conditions that were adopted to simulate field conditions and obtain settlements of the
embankments.
4.2.1 Choice of Constitutive Model and Material Properties
In the material set, type of material and type of model from material model box can be
selected. In order to simulate the behavior of the soil, a suitable model and appropriate material
parameters must be assigned to the geometry. In PLAXIS, soil properties are collected in
material data sets and the various datasets are stored in a material database. The material
properties used in the model for different material types are presented in Table 4.1 and Table
4.2.
The selected project is a bridge embankment constructed on soft soil, Arlington, Texas.
The embankment is found on a 5.5m thick soft clay layer followed by 2.5m of well graded sand

39

layer, followed by sand stone. Typically, soft clay has drained shear strength value ranges from
40 to 50kPa (Bowles 2000). Hence, the soft clay layer over sand was modeled with an effective
cohesion of 45kPa. The embankment section included 4 in. thick hot-mix asphalt concrete
(HMAC) and 11 in. of reinforced concrete pavement working platform above the existing
ground.
A plate element was used to simulate the pavement layer, which is placed on top of the
embankment. The properties of the pavement layer are obtained from the literature and they are
entered in a material set as a Youngs modulus value of 30GPa and a thickness of 0.35 m (for
road). The material properties of subsoil and pavement layers which are used in the current
modeling analysis are listed in Tables 4.1 and 4.2.
Table 4.1 Properties of subsoils used in Finite Element Analysis
Sub-Soil Properties

Soft Clay

Sand (well graded)

Sandstone

Drained

Drained

Un-drained

Compression Index, Cc

0.34

Swelling Index, Cs

0.043

1.3

Young Modulus, E (MPa)

2.0E+7

2.13E+10

Poisson Ratio,

0.25

0.2

35 - 48

18.85

17.65

26.47

45

27200

Friction Angle, (deg.)

33

28

Dilation Angle, (deg.)

Initial Void Ratio, e0

POP (kPa)
3

Unit weight, (kN/m )

Cohesion, c (kPa)

Model

Soft-Soil Model

Mohr-Coulomb

40

Hardening Soil Model

Table 4.2 Pavement properties used in Finite Element Analysis

Pavement Properties

Concrete

EA

1.116E+11

EI

1.179E+09

0.15

The numerical model has been analyzed with 2D plain strain model considering only
half of the section due to the symmetry of the problem. A total width of 52m has been used
which starts from the center of the embankment. The geometry of an embankment has been
made using geometric line option with approximate dimensions as per the field dimensions. The
LWA and Normal fill embankments with 1:2 (V:H) slopes were modeled in two dimensional plain
strain method of PLAXIS. A maximum traffic load of 40kN was used for loading on the
embankment.

Figure 4.1 Geometry of embankment model in PLAXIS

Two cases were investigated in the finite element analysis to evaluate the influence of
the following:
1. Vertical settlement on control section and test section using Mohr Coulomb Model
(MCM) for embankment fill
41

2. Vertical settlement on control section and test section using Cam-Clay Model or Soft-Soil
Model (SSM) for embankment fill.
Here the actual settlement behavior of the embankment over soft soil is expected to be
between these two cases.
The soft clay, the sand and the embankment fill were modeled as elastic-perfectly
plastic materials. No deformation below the sand layer was assumed. Mohr-Coulomb failure
envelope was used as the failure criterion in first case and critical state based soft soil (similar
to CAMCLAY) failure criterion was used in the second case. The material properties of the
embankment materials for Mohr-Coulomb Criteria and Soft-Soil Criteria are provided in Table
4.3 and Table 4.4, respectively. The elastic modulus adopted for the normal fill was typically
between 7 to 21MPa (Bowles 2000).

Table 4.3 Embankment Material Properties Mohr Coulomb Model, MCM

Embankment Properties

ECS (Undrained)

Normal Fill (Undrained)

E (MPa)

50

21

0.25

0.35

K (MPa)

33.33

23.3

G (MPa)

20

7.7

Unit weight (kN/m )

7.85

17.2

c(kPa)

75

85

(deg.)

49.5

18

(deg.)

18

Model

Mohr-Coulomb (MCM)

Mohr-Coulomb (MCM)

42

Table 4.4 Embankment Material Properties Soft Soil Model, SSM

Embankment Properties

ECS (Drained)

Normal Fill (Drained)

Cc

0.05

0.12

Cs

0.0

0.033

e0

1.0

0.5

E (MPa)

50

21

0.25

0.35

Unit weight (kN/m )

7.85

17.2

POP

58

52

Model

Soft Soil Model (SSM)

Soft Soil Model (SSM)

4.2.2 Mesh Generation

Two types of triangular elements are used in the PLAXIS, 6 noded triangular elements
and 15 noded triangular elements. Advantages of higher order triangular elements is that they
provide better representation of the description of continuous strain and stress variations and
also provides good description of a continuous displacement field with relatively few elements.
The disadvantages of higher order elements is that the failure loads may be dependent on the
mesh and makes poor description of discontinuous stress and strain. In PLAXIS, the program
automatically creates unstructured mesh as there is no possibility of making a so-called
structured mesh. The mesh size cannot be set explicitly. The mesh is generated based on
random seeds. The mesh size may be changed globally by means of global coarseness and
locally by means of local coarseness. Figure 4.2 presents a typical mesh generated for the
current analysis.

43

Figure 4.2 Typical Mesh Generation in PLAXIS

4.2.3 Initial and Boundary Conditions
3

In the initial conditions, water unit weight is set to 10kN/m . The water pressure is fully
hydrostatic and is based on a general phreatic level. In addition to phreatic level, boundary
condition for consolidation analysis can be additional input. The lines of consolidation need to
be selected in vertical direction that means vertical boundaries must be closed to restrain the
horizontal flow and no free outflow is allowed at that boundary.

Figure 4.3 Active Water Level in PLAXIS

44

The water conditions can also be specified in the Geometry configuration mode using
phreatic level by generating pore pressure using phreatic level. In the analysis, constant ground
water level has been considered.
4.2.4 Initial Stresses
In initial stresses which are effective stresses, Over-Consolidation Ratio (OCR) and
Pre-Overburden Pressure (POP) are used in analysis when using cam-clay model. Initial
stresses are developed by the POP procedure (PLAXIS 8, User Manual). It is also possible to
specify the initial stress state using the Pre-Overburden Pressure (POP) as an alternative to the
over consolidation ratio. The Pre-Overburden Pressure is defined by:



   

(4.1)

Figure 4.4 POP and OCR profile illustrations - PLAXIS 8, User Manual

Once the geometry of the model was developed, finite element model is complete.
Initial situation and initial stress state should be stated before calculation. This was done in the
initial conditions part of the input program. When using Mohr Coulomb model, the analysis
require the generation of the initial stresses by means of Ko procedure. Ko procedure can be
used to calculate initial stresses. The suggested Ko procedure is based on Jakys formula (1Ko* Sin).

45

4.2.5 Calculation
After the generation of Phreatic level and initial stresses, the input is complete and
calculations can be generated. These calculations are generally used to define the different
phases of embankment construction. Figure 4.5 presents a snapshot of different phases of the
construction process as implemented in FEM program.
The actual construction sequence was not reported in the reference and it was
simulated by adding embankment fill in three layers with equal thickness (3m). Four loads
coming in contact with road through tires were used to simulate the traffic loading. The loading
consisted of 40kN having uniform maximum vertical contact stress over the contact area with
ratio of 1:0.85 (width=19.27 cm, length=16.38 cm) placed on top of the pavement.
In the modeling analysis, the embankment construction consists of three phases, each
taking 30 days. After the construction phase, consolidation period of 60 days was introduced to
allow excess pore pressure to dissipate. The consolidation option in FEM software allows fully
automatic time stepping procedure that takes the critical time step into account. Pavement
construction and traffic loading were also taken into the consolidation analysis with different
time intervals. The last phase in consolidation analysis was selecting minimum pore pressure
2

where the default value of 1kN/m was used for the pore pressure.
To calculate the global safety factor for the road embankment, the phi-c reduction
option available in the PLAXIS was selected and used in the next phase.

Figure 4.5 Calculation Steps using FEM program

46

4.2.6 Results of Finite Element Analysis

On evaluating the total displacement, it can be seen that the failure mechanism is
developing with excess pore pressure distribution. The settlement at the pavement surface and
embankment were increasing considerably after the end of the construction of embankment and
pavement. This is due to the dissipation of excess pore pressure in soft soil layer which causes
consolidation in soils.
The vertical displacement (or settlement) contours for normal fill and ECS fill from the
numerical analysis (using Mohr coulomb and Cam-Clay model for embankment) over 30 years
of time are presented in Figure 4.6 and Figure 4.9 respectively. Figure 4.7, 4.8 and Figure 4.10,
4.11 shows maximum settlement that can occur after full dissipation of pore pressure at both
embankment locations.

Figure 4.6 Settlement vs. Time plot in the pavement in FEM using MCM for Embankment

47

Figure 4.7 Settlement contours at the end of pore pressure dissipation in ECS using MCM for
Embankment

Figure 4.8 Settlement contours at the end of pore pressure dissipation in Normal Fill using MCM
for Embankment

48

Figure 4.9 Settlement vs. Time plot in the pavement in FEM using SSM for Embankment

Figure 4.10 Settlement contours at the end of pore pressure dissipation in ECS using SSM for
Embankment

49

Figure 4.11 Settlement contours at the end of pore pressure dissipation in Normal Fill using
SSM for Embankment

4.2.7 Vertical Stress in Finite Element Analysis

The vertical stress contours from the numerical analyses for the two cases are also
presented in Figures 4.12, 4.13, 4.14 and 4.15. It is shown that more stress concentration
occurs at bottom of the normal fill as the loads are transmitted from the top to the bottom. The
degree of stress concentration is slightly reduced when ECS materials are used instead of
normal fill. In addition, the ECS carry fewer loads than the normal fill due to the low weight and
high strength properties.
From Figure 4.12 and 4.13, it has been observed that a maximum stress of 80kPa has
been transmitted to the soft foundation soil in the case of the ECS embankment after the full
dissipation of pore water pressure; whereas, a maximum vertical stress of 150kPa is recorded
at the interface of the embankment base and soft foundation soil for the control embankment.
This is expected since the ECS is a LWA material which has approximately half the unit weight
of the normal fill materials and hence imparts lower thrust on the foundation soil.

50

Figure 4.12 Stress distribution in Normal Fill after full dissipation of pore pressure using MCM
for Embankment

Figure 4.13 Stress distribution in ECS Fill after full dissipation of pore pressure using MCM for
Embankment

From Figure 4.14 and 4.15, it has been observed that a maximum stress of 50kPa has
been transmitted to the soft foundation soil in the case of the ECS embankment; whereas, a
maximum vertical stress from 150kPa is recorded at the interface of the embankment base and
soft foundation soil interface in the case of the control embankment.
51

Figure 4.14 Stress distribution in Normal Fill using SSM for Embankment

Figure 4.15 Stress distribution in ECS Fill using SSM for Embankment

4.3 Finite Difference Modeling Procedure

The finite difference method (FDM) is perhaps the oldest numerical technique used for
the solution of sets of differential equations, given initial values and/or boundary values (Desai
and Christian 1977).

For FDM analysis, the computer code FLAC Fast Lagrangian Analysis

of Continua (FLAC, 2000) was selected because of its flexibility and wide acceptance. It easily
incorporates key factors and phenomena which affects the behavior of ground conditions thus

52

making it a better choice to perform the FDM. The FDM code is an explicit two-dimensional
finite difference program that performs a Lagrangian analysis. Here, explicit means it uses a
time stepping procedure to solve the problem without forming the stiffness matrix. The
Lagrangian formulation enables the grid to move and deform with the material it represents
since the incremental displacements are added to the coordinates (FLAC 4.0 Manual, 2000).
FLAC is a very good tool for assessing the effect of many different material properties.
It is also a very robust tool because it can handle any constitutive model with no adjustment to
the solution algorithm (FLAC 4.0 Manual, 2000).
The use of FLAC in problem solving for static mechanical analysis in geotechnical
engineering by considering special aspects that is considered in model creating and solution.
Grid generation
Boundary and Initial Conditions
Choice of constitutive model and material properties
Interpretation of results
4.3.1 Grid Generation
In FLAC, grid generation is limited to simple and regular shaped regions. There is
a Fish function which can be used to generate user defined grids with varying zones. The
advantage of using FISH function is that it can easily adjust grid boundary and zone density
using SET command.
The FLAC grid is configured by specifying the command CONFIG axisymmetry at the
beginning of the data file for analysis. The structural element formulation only applies for planestrain or plane-stress analysis. The first step in FLAC is to select system units which are in
meters, kilograms and seconds and then the advanced constitutive modeling option is selected.
In this software, the factor of safety interface option is available only for Mohr Coulomb Model in
FLAC. Adjust total stress and ground water analysis can also be selected in FLAC simply by
specifying the CONFIG ats, gw command at the beginning of the data file.

53

The model used in our case study takes advantage of half symmetry. The size of the
model is 52 meters wide and 26 meters in height. Here the grid is composed of several zones
consist of constant height but variable zone width. A reasonably fine grid should be selected to
ensure that the settlement contours will be well-defined as it develops. It is best to use the finest
grid possible when studying problems as shown in Figure 14.16.

Figure 4.16 Fine grid generation in FDM

The basic grid used is named as Slope inside GENERATE command. The zones along
the slope face are all quadrilateral-shaped. The grid is divided into different regions using mark
command to define different soil types in different regions.
4.3.2 Boundary and Initial Conditions
If the geometry and loading in the embankment system is symmetrical, then we can
make mechanical boundary conditions correspond to roller boundaries along the symmetry line
and fixed displacements in the x and y direction, using the command Fix x and y at the model
2

base. The acceleration of gravity is set to be 9.81 m/sec (positive means to set it downwards).
There is an in-situ state of stress in the ground before starting any excavation or
construction. An initial condition specified in the FLAC grid simulates that in-situ stress. This is
an important step as it ultimately affects the subsequent behavior of the model. Ideally,

54

information about the initial state comes from field measurements but, when these are not
available, the model can be run for a range of possible conditions. A set of stresses is installed
in the grid and then FLAC is run until an equilibrium state is obtained (FLAC 4.0 Manual, 2000).
It is more difficult to give the initial stresses when materials of different densities are present. A
layered system with a free surface, enclosed in a box with roller side boundaries and fixed base
are used in analysis as shown in Figure 4.17. The ground water flow option in FLAC can be
used to find phreatic surface by establishing initial pore pressure boundary condition before
mechanical response. We apply pore pressure to raise the water level but in our case study, the
constant water table is adopted in the middle of the sandstone layer

Figure 4.17 Geometry with selected Material layers and initial conditions

4.3.3 Choice of Constitutive Model and Material Properties

There are ten built-in material models in FLAC. Here we will discuss the selected one
which has been considered for analysis.
(1) Elastic, isotropic.
(2) Mohr-Coulomb plasticity.
(3) Modified Cam-clay.

55

The elastic, isotropic model is valid for homogeneous, isotropic, continuous materials
that exhibit linear stress-strain behavior. The elastic, transversely-isotropic model is appropriate
for elastic materials that exhibit a well-defined elastic anisotropy (FLAC 4.0 Manual, 2000).
The Mohr-Coulomb plasticity model is used for materials that yield when subjected to
shear loading, but the yield stress depends on the major and minor principal stresses only; the
intermediate principal stress has no effect on yield. The Mohr-Coulomb model is applicable for
most general engineering studies. Also, Mohr-Coulomb parameters for cohesion and friction
angle are usually more-readily-available than other properties for geo-engineering materials.
Mohr-Coulomb models are the most computationally-efficient plasticity models (FLAC 4.0
Manual, 2000).
The modified Cam-clay model accounts for the influence of volume change on
deformability and on resistance to failure. In the Cam-clay model, models, tangential bulk and
shear moduli are functions of plastic volumetric deformation. Few points on cam clay model are
summarized below (FLAC 4.0 Manual, 2000):
1. The elastic deformation is nonlinear, with the elastic moduli depending on mean stress.
2. Shear failure is affected by the occurrence of plastic volumetric deformation; the material can
harden or soften, depending on the degree of preconsolidation.
3. As shear loading increases, the material evolves toward a critical state at which unlimited
shear strain occurs with no accompanying change in specific volume or stress.
4. There is no resistance to tensile mean stress.
These models assume an isotropic material behavior in the elastic range described by
two elastic constants bulk modulus (K) and shear modulus (G). The elastic constants, K and
G, are used in FLAC rather than Youngs modulus, E, and Poissons ratio, , because it is
believed that bulk and shear moduli correspond to more fundamental aspects of material
behavior than do Youngs modulus and Poissons ratio (FLAC 4.0 Manual, 2000).

56

The equations to convert from (E, ) to (K, G) are

K  E/ 31  2

(4.2)

G  E/ 21  

(4.3)

Models that include groundwater flow require the bulk modulus of the water, Kw. The physical
value of Kw is 2GPa for purewater at room temperature, but the value selected should depend
on the purpose of the analysis.
Vermeer and de Borst observe that values for the dilation angle are approximately

between 0 and 20 , whether the material is soil, rock, or concrete and this property is typically
determined from direct shear tests. The permeability, k, required by FLAC is the coefficient of
the pore pressure term in Darcys law.
The modified Cam-clay model is defined by initial elastic moduli plus parameters that
prescribe the nonlinear elasticity and hardening/softening behavior. The properties are given
below and these properties are determined from the laboratory testing.
= slope of the elastic swelling line
= slope of the normal consolidation line
M = material constant
mpc = preconsolidation pressure
mv_1 = specific volume
p1 = reference pressure
v0 = initial specific volume
v = specific volume at reference pressure on normal consolidation line.
The material properties of the subgrade soil materials adopted in FLAC are provided in
Table 4.5 and these results are used in performing the analysis. The material properties of the
embankment materials are provided in FEM section in Table 4.3 as same properties are
adopted to perform analysis in FLAC.

57

Table 4.5 Properties of subsoil at the location.

Sub-Soil Properties

Soft Clay

Sand (well graded)

Sandstone

Cc

0.34

Cs

0.043

e0

1.3

G (MPa)

26E+6

13.3E+6

88.3E+8

K (MPa)

57E+6

8.0E+6

88.7E+8

0.3

0.25

0.2

mpc (KPa)

97.6

mv_1

2.3

Unit weight (kN/m )

18.85

17.65

26.47

c (kPa)

27200

(deg.)

33

28

(deg.)

Model

Cam-Clay Model

Mohr-Coulomb

Mohr-Coulomb

4.3.4 Results in Finite Difference Analysis

Since FLAC models a nonlinear system as it evolves in time, the interpretation of results
may be more difficult than with a conventional finite-element program that produces a solution
at the end of its calculation phase.
The calculation is continued using step command and update interval of plot must be
used to estimate the effects of time. The SOLVE command is used to find the equilibrium state
(FLAC 4.0 Manual, 2000).

58

FLAC models the flow of fluid through the soil which is permeable in nature. In
consolidation (a type of fluid/solid interaction) the slow dissipation of pore pressure causes
displacements to occur in the soil and thus manifest in two mechanical effects. First, changing
pore pressure causes change in effective stress, which affects the soil response. Second, the
flow of water reacts to mechanical volume change by a change in pore pressure (FLAC 4.0
Manual, 2000). In our analysis we found that the convergence to the coupled drained and
undrained analysis is very slow. In our analysis, there are certain variables that are of particular
interest such as displacements and stresses. The construction of an embankment is assumed
to occur instantaneously, pavement and traffic moving on top surface is also considered at the
same time. An undrained analysis is first conducted to evaluate the settlement in the short-term
after building of the embankment. The long-term response is then monitored after allowing
drainage. The first stage of the simulation corresponds to the short time response of the system
in which no flow is assumed to take place. The command SET flow off is specified. Loading of
40 KN by the traffic coming in contact with road through four wheels of vehicle is simulated on a
0.25m section of the models top boundary. Once the full load is applied, the model is cycled to
equilibrium using SOLVE command. During this stage, a pore pressure in soft clay beneath
embankment develops as a result of volumetric deformations, but do not dissipate.
In the second stage, fluid flow is allowed to develop by issuing the commands SET flow
on. Water then drains through the model where the pore pressure is fixed at zero, and
additional settlement takes place under the embankment. The SOLVE command is used to
perform the simulation and it requires parameters like consolidation age and steady state flow.
History command is used to track the important variables during analysis. Plots of these
histories provide an understanding of the system. When certain zones in the model reach yield
point, those zones are indicated by Plastic indicators. The maximum unbalanced force when
reaches to its equilibrium state, it represents a stable model i.e. the maximum settlement has

59

occurred in the model. Graphical output is saved in one of these formats: JPG, PCX, DXF
(AutoCAD) etc (FLAC 4.0 Manual, 2000).

Figure 4.18 Settlement Results near pavement in FLAC using MCM for Embankment

The vertical displacement (or settlement) graph for normal fill and ECS fill from the
numerical FLAC analysis using Mohr Coulomb Model in FLAC are presented in Figure 4.18,
respectively.
Displacement graph and vertical displacement contours at the end of the undrained and
drained numerical simulations are presented in Figures 4.18, 4.19 and 4.20. The vertical
displacement histories recorded at different monitoring points indicate that the settlement at the
surface of both types of an embankment near pavement increases from approximately 150mm
to 200mm as a result of drainage. Note that the displacement graphs in Figure 4.18, and vertical
displacement contours in Figures 4.19 and 4.20, correspond to the combined undrained and
drained displacements. Figure 4.19 confirms that a steady-state flow has been reached by the

60

end of the drained simulation. In Figure 4:20 confirms that a steady state flow conditions are still
in progress.

Figure 4.19 Maximum Settlement of ECS using MCM for Embankment

Figure 4.20 Maximum Settlement of Normal Fill using MCM for Embankment

61

4.3.5 Vertical Stresses in Finite Difference Analysis

From Figures 4.23 and 4.24, it has been observed that a stress distribution over whole
section remains 250kPa which means stress are uniform over the ECS embankment and
vertical stress between 200 to 300kPa is recorded at the interface of the embankment base and
soft foundation soil in the control embankment at the end of drained conditions but the stress
keeps on increasing over the depth shown in Figure 4.23. At undrained conditions shown in
Figures 4.21 and 4.22, the stress of 150kPa has been transmitted to soft soil in case of ECS
embankment whereas stress of 250kPa is recorded at the interface of embankment and soft soil
in case of control embankment.
The higher stresses distributed to the soil below normal fill is due to the higher unit
weight of the normal fill material. Due to traffic loading, ECS embankment has accommodated
more stresses within the embankment section and transferred less stresses to the foundation
soil. In drained condition shown in Figure 4.24, ECS section creates uniform distribution of
stresses within itself and also to the foundation soil. In contrast, more stresses are being
transmitted by the normal fill to the soft soil leading to large surface deformations.

Figure 4.21 Stress distribution in ECS Fill at undrained condition using MCM for Embankment

62

Figure 4.22 Stress distribution in Normal Fill in undrained condition using MCM for Embankment

Figure 4.23 Stress distribution in Normal Fill in drained condition using MCM for Embankment

63

Figure 4.24 Stress distribution in ECS Fill in drained condition using MCM for Embankment

4.4 Summary
This chapter discusses the numerical analysis methodology using FEM (PLAXIS) and
FDM (FLAC) based modeling of the present embankment section. Three different numerical
analyses were carried out using the same subsoil properties. In FEM analysis, Mohr Coulomb
and Cam-Clay model were used for embankment while in FDM analysis, only Mohr Coulomb
model was used for embankment.

64

CHAPTER 5
ANALYSIS OF FIELD RESULTS AND COMPARISONS WITH NUMERICAL RESULTS
Here two full scale test embankments were constructed on SH-360 with different type of
backfill materials. The area is underlain by 5 to 6m of soft compressible clay. Beneath the soft
clay is well graded sand and it is underlain by sandstone which is around 9 to 10m thick. The
cross sections of the bridge include two 12 ft travel lane in each direction, 8 ft shoulders on both
sides of embankment. The bridge layout is presented in Figure below.

Figure 5.1 Satellite Imagery of SH-360

This case study contains fair information on soil properties, LWA materials, and five
year field measurements of settlements done using elevation survey after construction and
horizontal movement checked monthly using inclinometer. The finite difference software
PLAXIS Version 8.0 and FLAC (Fast Lagrangian Analysis of Continua) Version 4.0, developed
by Itasca Consulting Group, Inc., were adopted for this numerical analysis. The analysis is
focused on the evaluation of the maximum settlement over different period of time. The
analyzed section taken for the modeling purpose of SH 360 is given in Figure 5.2 presented in
section 5.1.1.

65

5.1 Analysis of Field Results

Pavement in this area is typically constructed on embankments with a total height of 0.3
to 1.5m and exerts traffic load of 35 to 55kN on the ground surface. Without any ground
improvement, a sudden drop in elevation has taken place at the joint of the approach
embankment with bridge structure. Stability problem is always associated with settlement
problems where there is soft ground beneath embankment as is the case in our study location.
After construction of embankment and pavement road, readings of vertical
displacements using elevation surveys on the surface of road and horizontal movement using
inclinometers were made at certain time intervals.
5.1.1 Instrumentation Data
Four vertical inclinometers were installed in the ECS embankment materials to monitor
the horizontal deformations as shown in Figures 5.2. Lateral displacements in the fill materials
are expected due to traffic vibration, strong bonding between pavement and fill material or
erosion. The first inclinometer was installed at the center line of the two embankments near the
center of the median. The second vertical inclinometer was installed on the inside slope of the
southbound embankment. The third and forth vertical inclinometer was installed on the inside
and outer slope of the northbound embankment.

Figure 5.2 Profile View at ECS section

66

Displacement profiles are useful for determining the magnitude, depth, direction, and
rate of ground movement thus helping in providing performance statistics of light weight
embankment fill materials like ECS. . Hence, the ECS embankment section was instrumented
with four vertical inclinometers, 12m in length and 70mm in diameter. Since ECS is a granular
material, specialized field installation the inclinometers was followed. Here drilling was done
using larger diameter auger with a central annular opening. Dry Bentonite chips were laid during
removal of the auger. Water was poured into the Bentonite column and left for curing and then
reaugered with a smaller diameter auger to install the inclinometer casing. This procedure
ensured the proper installation of the inclinometers at the assigned site.
The Digitilt vertical inclinometer system consists of the inclinometer casing, a vertical
inclinometer probe, a connecting cable and a readout unit. The probe for the vertical
inclinometer consists of one accelerometer measuring tilt in the plane of the inclinometer
wheels, which track the longitudinal grooves of the casing. The other accelerometer measures
tilt in the plane perpendicular to the wheels. Inclination measurements are converted to lateral
deviations after applying the necessary corrections as required (www.slopeindicator.com).
Changes in lateral deviation, determined by comparing data from current and initial
surveys, indicate embankment movements in both directions. Plotting the cumulative changes
at each measurement interval (0.6 m in this case) yields a high resolution displacement profile.
The data collected for a period of five years after the embankment construction from four
vertical inclinometers V1, V2, V3 and V4 are presented in Figures 5.4a, b, c and d respectively.
This data was collected after the embankment construction and before the application of traffic
load.. Hence, these small deformations are attributed to the construction induced settlement..
In the ensuing months after the construction of embankment further lateral movements
were recorded due to the traffic loading and time dependent consolidation settlements of the
foundation material. Due to the granular nature of the ECS fill consolidation settlements was

67

minimal in the test section. While in the control section elevation survey confirmed that there
was much higher settlement.
From Figure 5.4a, 5.4b, 5.4c and 5.4d, it can be seen that the lateral movements
(deviations from the center line) in first, second and third of the inclinometers are within the
permissible limits except the fourth one. The first inclinometer (V1) which is located at the center
line of the median showed very less variation in its reading throughout the depth due to
symmetrical loading from either side. The fourth inclinometer (V4), located at the outside slope
of the northbound lanes, shows a slight rotation of the slope. The second and third inclinometer
(V3 and V2), located at the inside slope of the northbound and southbound lanes, shows a
slight rotation of the slope.

(a)

(b)

Figure 5.3 (a) Horizontal movement seen near V2 location (b) Testing at V2 location

The readings from the four vertical inclinometers (V1, V2, V3 and V4 for ECS material)
are presented in Figure 5.4 (a), (b), (c) and (d). Readings were taken once in month from Aug
2006-June 2007.After that the readings were taken with a gap of 2 to 5 months in-between up to
June 2009 and then again the readings were taken once in a month from July 2009 to April
2010.Again from April 2010, the readings were taken with one month of gap. The results at V1
location indicated that in north-south direction, there were very small movements but have
around 6mm from 0 to 1m depth in east-west direction. Its seen that the horizontal

68

displacement is not much concern at V1 location. Physical field observation has shown
abnormal horizontal movement near V2 location shown in Figure 5.3 which may be a future
concern but the inclinometer results at V2 is showing the movement within permissible limit.
Continuous monitoring is recommended.

(a)

(b)

(c)

(d)

Figure 5.4 (a) V1, (b) V2, (c) V3, (d) V4, Vertical Inclinometer Readings: Cumulative
displacement vs. Depth

69

At V4 vertical inclinometer, the movement is increasing from 90 to 120mm from 0 to 1m

depth. The movement is more on the top and very less below 3m, typically lateral movements
beyond 25 mm are considered problematic. Very high horizontal movement readings have been
recorded during the time duration of May 2007 to June 2010. Possible reason of movement can
be due to soil erosion, inadequate subsurface drainage systems, vibration of moving traffic and
poor construction practices. Erosion can lead to severe problems like void development, failure
of slope protection cover and loss of backfill material under the pavement and bridge. Field
observations show that the worst erosion conditions are associated with directing surface water
directly in to gaps between pavement and shoulders and directly on to and under the slope
protection. Bad construction practices like poor compaction of the fill material can lead to severe
settlement problems. Selection of the correct compaction moisture content is important in
compacting poorly graded granular backfill. This parameter can be determined from optimum
moisture content test by plotting the relationship between dry density and moisture content and
selecting a target acceptable density.

Figure 5.5 Horizontal movements on ECS section

The Figure 5.5 above illustrates horizontal displacement at test section which has been
filled with tar. Its seen that the horizontal displacement is increasing with time. At V2 and V3,

70

the movements are under permissible limits and are not considered problematic. This section is
under continuous monitoring to observe if any abnormal movements occur.
5.2 Pavement Surface Profiling
The objective of pavement surface profiling is to find the undulation on the pavement
surface. Profiling as well as visual/digital photographic inspections have been carried out on
both the ECS and normal fill embankments to understand the fill settlements and any heave
related movements.
Elevation surveys have been taken once every two weeks for the first few months after
the construction and thereafter monthly for the duration of the data collection. Surveys were
conducted using a Total Station device. After establishing control points, surface profiling is
carried out. This provides designers with information on movement of the pavement surface
from its initial recorded position. These movements can be either positive which indicates
heaving or it can be negative which indicates settlement.

Figure 5.6 Settlement readings from selected survey points

71

5.3 Hyperbolic Formulation

The hyperbolic formulation method has proved applicable for long term settlement
prediction in complex soil formations (Tan 1971, Kodamdaramaswamy and N. Rao 1980, N.
Rao and Somayajulu 1981). The two constant mathematical relation mentioned below provides
a practical solution for geotechnical problems like estimation of long term settlement. In this
case the relationship between settlement and time t is assumed to follow a hyperbolic curve :


   

(5.1)

Where A and B are the slope and the intercept of the straight line, respectively. If
settlement versus time t obeys the hyperbolic relationship, the transformed hyperbolic plot of
t/ against t should be a straight line and the ultimate total settlement is obtained from the
asymptotic line to the hyperbolic formulation. Taking the limits of the above Equation as t
approaches infinity, the total settlement is given by 1/A, which is the reciprocal of the slope of
the straight line (Kodamdaramaswamy and N. Rao 1980).
5.3.1 Long Term Settlements Using Field Elevation Data with Hyperbolic Formulation
The predicted settlement using hyperbolic formulation matches the measured
settlement values. The hyperbolic relationship was an approximate fit of the survey elevation
data over the entire range of settlement-time response. The straight line was prominent and
obvious irrespective of various shapes of settlement versus time curves.
The hyperbolic formulation analysis was performed at both control section and test
section where the following equations are determined using linear relationships:
Settlement at the control section
S = t / (0.0107 x t + 5.5982) mm
Settlement at the control section
S = t / (0.0295 x t + 4.8726) mm

72

Figure 5.7 Hyperbolic plot of t (days) versus t/s (day/mm) corresponding to the settlement
records.

5.4 Comparison of Field Results with Hyperbolic Formulation

Vertical displacement (or settlement) data from the elevation survey over the pavement
surface were collected from the last four years of both normal fill and ECS fill. This data was
then compared with hyperbolic formulation and the results matched closely with the measured
settlements. Figure 5.8 and 5.9 shows a good match of hyperbolic formulation values with the
measured elevation survey data.

73

Figure 5.8 Comparison of Elevation Survey data with Hyperbolic Formulation for ECS

Figure 5.9 Comparison of Elevation Survey data with Hyperbolic Formulation for Normal Fill

5.5 Comparison of Hyperbolic Formulation Results with Numerical Modeling Results

Terashi (2003) indicated that nearly 60 percent of on-land application in Japan and
perhaps roughly 85 percent of Nordic applications are for the settlement reduction and
improvement of stability of embankment. The settlement reduction creates a more economical
solution for embankment systems.
74

The behavior of treated sections has been comprehensively studied through field
observation of full scale physical model, laboratory testing and numerical simulation. However,
the cost of constructing and monitoring treated and untreated test embankments are quite high.
An alternative method such as numerical experiment or simulation by means of appropriate
methods such as finite element and finite difference techniques is essentially required. The
numerical simulation of these two embankments systems were realized by means of Finite
Difference and Finite Element methods using 2D analysis program. The aim of this study is to
investigate the influence of geometric configurations using 2D numerical simulations of the two
test embankments. Particular attention is given to the vertical displacements or settlement.
Subsequent comparisons are made to study the long term settlement behaviors between the
findings of 2D numerical simulations and those from the actual measured field data used by the
hyperbolic formulation method of the two full scale embankments (test section and control
section).
As a ground improvement technique, the ECS has been used as an alternative to
increase bearing capacity, reduce settlements, and increase deep-seated slope stability of
embankments over soft soil.
In addition to the treated embankment, other embankment section was also constructed
without treatment and this section was considered as control section. Therefore, this site has
one treated and one control sections for performance evaluation of ECS and normal fill
materials.
After four years, the measured maximum settlements were approximately between 50
to 200mm at control section as shown in Figure 5.10 and Figure 5.11.

It looks like the

settlements had become relatively stable after approximately few years since construction.

75

Figure 5.10 Vertical displacement at SH 360 -Control Section

Figure 5.11 Close view of vertical displacement at SH 360 - Control Section

Four cases were analyzed to evaluate the long term influence of the (1) Vertical
settlement on control section and test section using Mohr Coulomb Model for embankment fill
(PLAXIS, FEM) (2) Vertical settlement on control section and test section using Cam-Clay
Model for embankment fill (PLAXIS, FEM). (3) Vertical settlement on control section and test
section using Mohr Coulomb Model for embankment fill (FLAC, FDM). (4) Extension of vertical
settlement on control section and test section using hyperbolic formulation to predict long term
settlements. These results are presented in the next section.

76

5.5.1 Vertical Displacement at Pavement Surface

The vertical displacements (or settlement) for normal fill and ECS fill from the
numerical analysis were compared near the pavement surface because in the field, maximum
settlement at normal fill was recorded near the pavement surface and very less settlement was
seen at the pavement surface of ECS fill. The long term maximum settlement at pavement
surface was then determined using hyperbolic formulation and these settlements are
determined at 10, 20 and 30 years using the hyperbolic formulation results are used as a
benchmark to compare the numerical analysis modeling results.
The numerical analysis, FEM and FDM results from Chapter 4 are compared with
hyperbolic formulation results at 10, 20 and 30 years to check the long term settlements over
the pavement and these results are summarized in Figure 5.12 and Figure 5.13. These results
are also presented in Table 5.1, Table 5.2 and Table 5.3.

Figure 5.12 Comparison of predicted and measured long term settlement results using different
methods at ECS Section

77

As expected, the first case with ECS fill modeled in PLAXIS (FEM) shows that the
results are reasonably close to the long term interpretations based on hyperbolic formulation of
the initial measured data. The second case with ECS using Cam-Clay Model yielded a good
match and the same observation is noted for the third case with FDM modeling.

Figure 5.13 Comparison of predicted and measured long term settlement results using different
methods at Control Section

The overall settlement observed in the control section is much higher than the ECS
which is an expected outcome. Figure 5.13 compares different modeling test results on the
control section with the hyperbolic formulation results. The first case with normal fill predicts
moderately close results of the maximum settlement over the hyperbolic formulation method
using Mohr-Coulomb Model. The second case with normal fill predicts reasonably close
maximum settlement to the hyperbolic formulation using Cam-Clay Model. The third case was
developed in FDM where normal fill over-predicts the maximum settlement at surface than as
per case forth, hyperbolic formulation.
78

Table 5.1 Hyperbolic Formulation Results

Time (years)
10

20

30

Control section

81.9

87.3

89.2

Test section

32.43

33.15

33.40

Settlement (mm)

Table 5.2 PLAXIS, FEM Results using MCM / SSM for Embankment
Time (years)
10

20

30

Control Section

75 / 64

113 / 95

140 / 115

Test Section

23 / 26

30 / 34

34 / 39

Settlement (mm)

Table 5.3 FLAC, FDM Results using MCM for Embankment

Time (years)
10

20

30

Control Section

161

184

200

Test Section

31

33

34

Settlement (mm)

5.6 Summary
This chapter discusses instrumentation data and comparison of numerical model results
with hyperbolic formulation.. Instrumentation data from vertical inclinometer installed at ECS
location shows one of the locations has horizontal movement out of permissible limit. Possible
reason of movement can be soil erosion, inadequate subsurface drainage systems, vibration of
79

moving traffic and poor construction practices. All FEM and FDM methods provided closer
predictions with the extended interpretations of measured settlements. The only exception is the
FDM models predictions for Control section and these variations are attributed to interface
behavioral variations in the FDM. Overall, FEM modeling provided very good match of long term
settlements of both Control and test sections.

80

CHAPTER 6
SUMMARY, CONCLUSIONS AND FUTURE RESEARCH RECOMMENDATIONS
6.1 Introduction
The main objective of the present research was to study the use of different constitutive
models used in numerical analysis to model light weight embankment and local embankment
fills over a soft subsoil and then recommend the use of the models for future designs of light
weight embankment sections. The objectives of this research were accomplished and the
results of this research are presented in Chapter 4 and Chapter 5. Some of the salient research
findings of this research are summarized in the following section.
6.2 Summary and Conclusions
In this research project mainly two types of materials were studied for embankment to
better check the settlement near pavement surface. Various test were performed on both test
and control section of an embankment material to know the engineering properties which can
be used in numerical analysis. A series of laboratory tests were performed which include tests
like Atterberg limits, compaction, sieve analysis, hydrometer, direct shear test and free vertical
swell. Numerical modeling using FDM and FEM models are used to determine the long term
settlement behaviors of both control section and test sections. Mohr coulomb model were used
for embankment material modeling in the FDM analysis while both Mohr-coulomb and CamClay Models were adopted for FEM analysis. The hyperbolic formulation was studied and used
on elevation survey data and then the method was used to predict the long term settlements.
Based on the present experimental program, numerical analyses and hyperbolic formulation
prediction results, the following conclusions are drawn:
1. Finite difference analysis need appropriate constitutive model for a particular soil type and
takes considerable time for complete settlement analysis.

81

2. Finite element analysis is more powerful and gives better results that match with
interpretations of long term settlements using the measured data.
3. Light weight material fill shows less settlement when compared to normal material fill
settlements because the unit weight of LWA being one third to that of normal fill.
4. Vertical stresses are less in LWA embankment when compared to normal fill embankment
and this implies stress induced settlements will be lesser in LWA embankment.
5. Swelling characteristics test data showed that untreated section swell was about 2.5% where
as for the treated sample it was almost negligible.
Based on predicting settlements of embankment, there are a few uncertainties in the
modeling which are mentioned below:
1.

Uncertain time history of loading.

2.

Economic problems that restrict testing and analysis.

3.

Uncertain stratigraphy and water level fluctuations.

4.

Movement due to shearing stresses.

6.3 Future Research Recommendations

Below are few recommendations for future research in the use of low weight materials
as an embankment fill:
1. Developing a method to obtain reliable soil profiles of the study location.
2. Other low weight materials with different properties can be analyzed before using it as a fill
material in field.
3. Studying other constitutive models on these materials
4. Tension in the soil need to be studies in FDM analysis to get more accurate prediction.
5. Documentation of any study location with all data accessible.
6. Doing the same work in 3D modeling and comparing the results with 2D.

82

REFERENCES
Sireesh Saride, Sunil K. Sirigiripet, Anand J. Puppala, and Richard Williammee (2009),
Performance of Expanded Clay Shale (ECS) as an Embankment Backfill,.
Sireesh Saride, Sunil K. Sirigiripet, Anand J. Puppala, Richard Williammee (2008),
Performance of Expanded Clay Shale (ECS) as an Embankment Backfill
Tor Erik Frydenlund and Roald Aaboe. (2001), Long Term Performance and Durability
of EPS as a Lightweight. EPS Geofoam, 3rd International Conference, Salt Lake City,
December 2001
Zhang, Min-Hong and Odd E. Gjorv. (1991), Title No. 88-M19: Characteristics of
Lightweight Aggregates for High-Strength Concrete. ACI Materials Journal March-April 1991:
150-158
Gustavsson, Henry, Olli Ravaska, and ke Hermansson (2002), The Effect of
Insulation on Road Surface Icing. Cold Regions Engineering Cold Regions Impacts on
Transportation and Infrastructure. Ed. Kelly S. Merrill. Reston, VA: American Society of Civil
Engineers, 278-289
Sungmin Yoon, Umashankar Balunaini; Irem Z. Yildirim; Monica Prezzi and Nayyar Z.
Siddiki (2009), Construction of an Embankment with a Fly and Bottom Ash Mixture: Field
Performance Study
Patelunas, G. M. (1986), High volume fly ash utilization projects in the United States
and Canada. Rep. No. CS-4446, Electric Power Research Institute, Palo Alto, California.
Srivastava, L. and Collins, R. J. (1989), Ash utilization in highways: Delaware
demonstration project. GS-6481, Research Project No.2422-3, Interim Rep., Electric Power
Research Institute, Palo Alto, California.

83

Usmen, M. A., Head, W. J., and Moulton, L. K. (1983), Use of coal associated wastes
in low-volume roads. Transportation Research Record. 898, Transportation Research Board,
National Research Council, Washington, D.C., 268277.
Rehage, J. A. and Schrab, G. E. (1995), Environmental performance assessment of
coal combustion byproduct use sites: Road construction applications. TR-105127, Research
Project No. 2796-01, Final Rep., Electric Power Research Institute, Palo Alto, Calif.
An Deng and Jin-Rong Feng (2009), Granular Lightweight Fill Composed of Sand and
Tire Scrap ASCE
Peter J. Bosscher, Thncer B. Edil and Senro Kuraoka (1997), Design of highway
embankments using tire chips
Humphrey, D.N., Whetten, N., Weaver, J., and Recker, K. (2000), Tire Shreds as
Lightweight Fill for Construction on Weak Marine Clay, Proceedings of the International
Symposium on Coastal Geotechnical Engineering in Practice, Balkema, Rotterdam, pp. 611616.
Rossignolo, J. A., Agnesini, M. V. C. and Morais, J. A.(2003), Properties of HighPerformance LWAC for Precast Structures with Brazilian Lightweight Aggregates, Cement and
Concrete Composites, 77-82
Pioro, L. S. and Pioro, I. L. (2004), Production of Expanded-Clay Aggregate for
Lightweight Concrete from Non-Self bloating Clays, Cement and Concrete Composites, 639643
Vasina, M., Hughes, D. C., Horoshenkov, K. V. and Lapcik, L. (2005), The Acoustical
Properties of Consolidated Expanded Clay Granulates, Applied Acoustics, Article in press.
Th. Zimmermann, A Truty and J.L Sarf (2006), "Numerical simulation of underground
works and application to cut and cover construction" Taylor and Francis Group, London, ISBN 0
415 39748 0

84

Mladenovic, A., Suput, J. S., Ducman, V. and Skapin, A. S. (2004), Alkali-Silica

Reactivity of Some Frequently Used Lightweight Aggregates, Cement and Concrete Research,
Vol. [34], 1809-1816.
O. Arioz, K.Kilinc, B. Karasu, G. Kaya, G. Arslan, M. Tuncan, A. Tuncan, M. Korkut and
S. Kivrak (2008), A Preliminary Research On The Properties of Lightweight Expanded Clay
Aggregate.
Kodandaramaswamy K., Narasimah Rao, S. (1980),The prediction of settlements and
heave in clays, Canadian Geotechnical Journal 17, 623 631.
M.A. Al-Shamrani / Engineering Geology 77 (2005) 1734, Applying the hyperbolic
method and Ca/Cc concept for settlement prediction of complex organic-rich soil formations.
Paulo J. Venda Oliveira, Luis J. L. Lemos, and Paulo A. L. F. Coelho, ASCE (2010),
Behavior of an A typical Embankment on Soft Soil: Field Observations and Numerical
Simulation
DeMerchant, R. M and Valsangkar, A. J. (2002), Plate load tests on geo-grid reinforced
expanded shale lightweight aggregate, Journal on Geomembrane and Geotextiles, Vol 20,
Issue 3, pp172-190.
Holm, T. A., and Ooi, O. S. (2003), Moisture dynamics in lightweight aggregate and
concrete.6th International Conference on the Durability of Concrete, Thessaloniki, Greece.
Expanded Shale, Clay and Slate (ESCS) (2004), Light weight aggregate for
geotechnical applications Information Sheet 6001 www.escsi.org
Stoll, R. D., and T. A. Holm, T. A. (1985), Expanded shale lightweight fill: Geotechnical
properties, Journal of Geotechnical Engineering, ASCE, Vol. 111, No.8.
Holm, T. A., and Valsangkar, A. J. (1993), Lightweight artificial and waste materials for
embankments over soft soils Lightweight aggregate soils mechanics: Properties and
Applications, Transportation Research Record, Issue 1422, Washington, D.C., pp 7-13.

85

Karim, A. M. K. (1997), Building highway embankments of fly/bottom ash mixtures.

Joint Highway Research Project, Final Rep. No. FHWA/IN/JTRP-97/1, Purdue Univ., W.
Lafayette, Ind.
Kim, B. (2003), Properties of coal ash mixtures and their use in highway
embankments. Ph.D. dissertation, School of Civil Engineering, Purdue Univ., W. Lafayette, Ind.
www.escsi.org, (1960) Story of the Selma, ESCSI
Bernadette Dupony and David L.Allen (2002), "Movements and Settlement of Highway
Bridge Approachs", Report No. KTC-02-18/SPR-220-00-1F, Kentucky Transportation Center,
Lexington, Kentucky
Lenke, L. R. (2006), Settlement Issues-Bridge Approach Slabs. Rep.No. NM04MNT02, New Mexico Department of Transportation.
Schaefer, V.R. and Koch, J.C. (1992), Void Development under Bridge Approaches,
Rep. No. SD90-03, South Dakota Department of Transportation.
Robert D. Stoll and Thomas A. Holm, Expanded Shale Lightweight Fill: Geotechnical
Properties
Childs, K., Porter, D. L., and Holm, T. A. (1983), "Lightweight Fill Helps Albany Port
Expand," Civil Engineering, ASCE, Apr., pp. 54-57.
John P. Ries, Thomas A. Holm (ECSCI), Lightweight Concrete- Two Millennia of
Proven Performance.
Hopkins, T.C. (1969), Settlement of highway bridge approaches and embankment
foundations. Rep.No.KYHPR-64-17, HPR-1(4), Kentucky Transportation Center, Lexington, Ky.
Kramer, S. L.,and Sajer, P. (1991), Bridge approach slab effectiveness. Final Rep.No.
WA-RD 227.1, Washington State, DOT, Olympia, Wash.
Seo, J. (2003), The bump at the end of the bridge: An investigation. Dissertation,
Texas A&M Univ., College Station, Tex.

86

Stoll, R. D.,and Holm, T. A. (1985), Expanded shale lightweight fill: Geotechnical

properties. J. Geotech. Engrg., 111(8), 10231027.
Valsangkar, A. J.,and Holm, T. A. (1990), Geotechnical properties of expanded shale
lightweight aggregate. Geotech. Test. J., 13(1), 1015.
Valsangkar, A. J.,and Holm, T. A. (1999), Mechanical durability of expanded shale
lightweight aggregate. Geotech. Test. J., 22(4), 329333.
Wahls, H. E. (1990), NCHRP synthesis of highway practice no. 159: Design and
construction of bridge approaches, Transportation Research Board, Washington, D.C.
Hao Wang and Xiaoming Huang (2006), "Study of Ground and Embankment Treatment
Techniques in Road Widening", Pavement Mechanics and Performance (GSP 154) Part of
GeoShanghai 2006, Proceedings of Sessions of GeoShanghai 2006
John

Dunnicliff,

with

assistance

to

Gorden

E.Green,

(1993),

"Geotechnical

instrumentation for monitoring field performance".

Coulomb, C. A. (1776), Essai sur une application des regles des maximis et minimis a
quelquels problemesde statique relatifs, a la architecture. Mem. Acad. Roy. Div. Sav., vol. 7,
pp. 343387
Miki, H. (1996), An overview of lightweight banking technology in Japan, pp. 10-30. In
Proceeding of International Symposium on EPS Construction Method. Tokyo. Japan.
Holm, T. A.,and Ooi, O. S. (2003), Moisture dynamics in lightweight aggregate and
concrete. Proc., 6th Int. Conf. on the Durability of Concrete, ACI International, Farmington Hills,
Mich.
Holm, T. A. and Valsangkar, A. J. (1993), Lightweight artificial and waste materials for
embankments over soft soils. Lightweight aggregate soils mechanics: Properties and
applications, Transportation Research Board, Washington, D.C., 713
Stewart, C. F. (1985), Highway structure approaches. Rep.No. FHWA/CA/SD-85-05,
California DOT, Sacramento, Calif.

87

Briaud, J. L., James, R. W.,and Hoffman, S. B. (1997), NCHRP synthesis 234:

Settlement of bridge approaches (the bump at the end of the bridge), Transportation Research
Board, Washington, D.C.
Zhang, Min-Hong and Odd E. Gjorv., (1991), Title No. 88-M19: Characteristics of
Lightweight Aggregates for High-Strength Concrete. ACI Materials Journal March-April: 150158.
Moore, William M., (1970), Fired-Clay Aggregates for Use in Flexible Bases. Highway
Research Record Number 307: Synthetic Aggregates and Granular Materials: 33-43.
Mehdiratta, G.Rai and Michael O.Noggle. (1986), Geotechnical Properties and
Engineering Applications of Lightweight Aggregate. Texas Civil Engineer
Ledbetter, W. B., W. M. Moore, and B. M. Gallaway, (1971), A Synthetic Coarse
Aggregate Classification System Final Report (Research Report 81-15F). College Station,
Texas: Texas Transportation Institute, Texas A&M University.
Gustavsson, Henry, Olli Ravaska and ke Hermansson (2002), The Effect of Insulation
on Road Surface Icing. Cold Regions Engineering Cold Regions Impacts on Transportation
and Infrastructure. Ed. Kelly S. Merrill. Reston, VA: American Society of Civil Engineers, 278289.
Brettmann, Tracy and Cibor, Joseph M. (1991), Use of Lightweight Aggregate Fill for a
Highway Embankment on Compressible Soil (Baytown, Texas). Report No. 0401-2436.
Houston, Texas: Fugro-McClelland (Southwest), Inc.
Stark et al. (1995), Differential Movement of the Embankment/Structure InterfaceMitigation and Rehabilitation.
George Machan and Victoria G. Bennett (2008), "Use of Inclinometers for Geotechnical
Instrumentation on Transportation Projects". Transportation Research Circular E-C129
Edii, T. B. and Bosscher. P.I, (l994) Engineering Properties of Tire Chips and Soil
Mixtures," Geotechnical Testing Journal, GTJODJ. Vol. 17. No. 4, pp. 453-464

88

Joaquim Peiro and Spencer Sherwin (2005), "Finite difference, finite element and finite
volume methods for partial differential equations", Department of Aeronautics, Imperial College,
London, UK, Handbook of Materials Modeling. Volume I: Methods and Models, 132 c Springer.
Yeh, S.T. and Su, C. K. (1995), "EPS flow fill and structure fill for bridge abutment
backfill", Report No. CDOT-R-SM-95-15, Colorado Department of Transportation, Denver, Col.,
U.S.A., August 1995, 17 pp.
Schaefer, Vernon R and Koch, Jay C. (1992), Void Development under Bridge
Approaches. Final Report No SD90-03, South Dakota Department of Transportation.
Kramer, Steven L and Sajer, Peter (1991). Bridge Approach Slab Effectiveness,
Final Report, Washington State Department of Transportation, December.
Laguros, J. G., Zaman, M. M., and Mahmood, I. U. (1990), Evaluation of Causes of
Excessive Settlements of Pavements Behind Bridge Abutments and their Remedies; Phase II.
(Executive Summary), Oklahoma Department of Transportation, Report No. FHWA/OK 89 (07).
Wahls, H.E. (1990), Design and Construction of Bridge Approaches, NCHRP synthesis
159. Washington,
D.C: Transportation Research Board, National Research Council. Wolde-Tinsae and
Aggour (1990), Structural and Soil Provisions for Approaches to Bridges
Hoppe, E.J (1999), Guidelines for the Use, Design, and Construction of Bridge
Approach Slabs. Virginia Transportation Research Council, VTRC00-R4.
Abu-Hejleh, N., Hanneman, D., White, D. J., and Ksouri, I. (2006), Flowfill and MSE
Bridge Approaches: Performance, Cost and Recommendations for Improvements. Report No.
CDOT-DTD-R-2006-2. Colorado Department of Transportation, Denver.
Greimann, L. F., Yang, P-S., and Wolde-Tinsae, A. M. (1986), Nonlinear analysis of
integral abutment bridges. Journal of Structural Engineering, ASCE, 112(10), 2263-2280.
Jayawickrama, P., Nash, P., Leaverton, M. and Mishra, D. (2005), Water Intrusion in

89

Base/Subgrade Materials at Bridge Ends. TxDOT Report, FHWA/TX-06/0-5096-1,

Texas Tech University, Lubbock, Texas.
Bergado, D.T., and Patawaran, M.A.B. (2000), Recent developments of ground
improvement with pvd on soft Bangkok clay. Proc. Intl. seminar on Geotechnics in Kochi 2000,
Kochi, Japan, October, 2000.
Barksdale, R.D., and Bachus, R.C. (1983). Design and Construction of Stone
Columns, Vol.1 Rep. No. FHWA/RD-83/026, Federal Highway Administration, Washington,
D.C.
Porbaha, A. (2000), State of the Art in Deep Mixing Technology: Design
Considerations. Ground Improvement, 4(3), 111-125.
Tawatchai Tanchaisawat, D.T. Bergado and Taweesak

Kanjananak, (2007),"

Lightweight Geomaterials for Bridge Approach Utilization on Soft Ground Area",Kasetsart J.

(Nat. Sci.) 41 : 357 - 363 (2007)
D. J. Thompsett, A. Walker, R. J. Radley and B. M. Grieveson (1995)," Design and
construction of expanded polystyrene embankments", Construction and Building Materials, Vol.
9, No. 6, pp. 40341 I.
Patelunas, G. M. (1986)," High-volume fly ash utilization projects in the United States
and Canada". Final report.
Srivastava, L. and Collins, R.J. (1989), Ash Utilization in Highways: Delaware
Demonstration Project, Interim Report, GS-6481, Research Project 2422-3, Electric Power
Research Institute, Palo Alto, CA, USA.
Alleman, E. J., Fox, P. J., and Battista, D. D. (1996). Performance evaluation of
highway embankment constructed using coal ash. Joint Highway Research Project, Final Rep.
No. FHWA/IN/JHRP-96/5, School of Civil Engineering, Purdue Univ., W. Lafayette, Ind.
Edil, T. and Bosscher, P. (1994), "Engineering Properties of Waste Tire Chips and Soil
Mixtures," Geotech. Testing J., ASTM, Vol. 17, No. 4, pp. 453-464.

90

Olufikayo Aderinlewo and Nii-Attoh Okine (July 2008), Sensitivity analysis of a scrap tire
embankment using Bayesian influence diagrams. Science Direct.
Rossignolo, J. A., Agnesini, M. V. C. and Morais, J. A., (2003) Properties of HighPerformance LWAC for Precast Structures with Brazilian Lightweight Aggregates, Cement and
Concrete Composites, 77-82.
Su-Chen Huanga, Fang-Chih Changb, Shang-Lien Lob, Ming-Yu Lee, Chu-Fang Wang,
Jyh-Dong Lin, (2007), " Production of lightweight aggregates from mining residues, heavy metal
sludge, and incinerator fly ash", Journal of Hazardous Materials 144, 5258.
Murray, R.T., (1971), Embankments constructed on soft foundations: settlement study
at Avonmouth, Report LR, vol. 419. Road Research Laboratory, Crowthorne, p. 26.
Biot, M.A., (1941), General theory of three-dimensional consolidation, Journal of
Applied Physics 12, 155164.
Borges, J.L., (2004), Three-dimensional analysis of embankments on soft soils
incorporating vertical drains by finite element method. Computers and Geotechnics 31 (8), 665
676.
Leroueil, S., Tavenas, F., Trak, B., La Rochelle, P., and Roy, M. (1978), Construction
pore pressures in clay foundation under embankment Part I: The Saint-Alban test fill. Part II:
Generalized behavior. Can. Geotech. J., 151, 5482.
Tavenas, F. (1981), Some aspects of clay behaviour and their consequences on
modeling techniques. American Society for Testing and Materials, Special Technical
Publication No. 740, pp. 667-677.
Qian, J.H. and Yin, Z.Z., (1996), Geotechnical Principles and Calculation, Chinese
Water Conservancy Hydroelectric Press, Beijing. 720 pp.
O.C. Zienkiewicz and Y.K. Cheung, (1965), Finite Elements in the solution of Field
Problems, The Engineer, September 24th.

91

Mayne, P. W., Christopher, B.R., and DeJong, J., (2002), Subsurface Investigations
Geotechnical Site Characterization, Publication No. FHWA NHI-01-031, National Highway
Institute, Federal Highway Administration, Washington, DC, 300 pp.
Laurene V. Fausett, "Numerical Methods Algorithms and Applications
Bowles, J.E., (1979), Physical and Geotechnical Properties of Soils. McGraw-Hill,
New York. 410 pp.
Bowles, J. E. (2000), Foundation Analysis and Design, McGraw-Hill, New York.
GEO5,

Theoretical

Manual,

(2010),

http://www.finesoftware.eu/geotechnical-

software/help/fem/materials-models/
Geotechnical Design Manual, M 46-03.03, November 2010, Washington State
Department of Transportation
US Army Corps of Engineers, Engineering Manual (1995), EM 1110-2-1908,
Instrumentation of Embankment Dams and Levees
ASTM International (2003), ASTM Book of Standards, Vol. 4.13, ASTM, Philadelphia,
Pennsylvania.
PLAXIS 8.2, Material Model Manual
FLAC 4.0, Constitutive Model Manual

92

BIOGRAPHICAL INFORMATION
Seema Kalla was born in Kashmir, Jammu & Kashmir, India. She received her
bachelors degree in Civil Engineering from the Shivaji University, Kolhapur, India. The author
had six years of work experience in Offshore Geotechnical field. The author joined the
University of Texas at Arlington in August, 2009 as a MS candidate in Geotechnical
Engineering. During the course of her study the author worked as a graduate laboratory
assistant under Dr. Anand J. Puppala and had a chance to work in various research projects
involving laboratory tests and monitoring. The authors research interests include on Numerical
modeling and experimental studies of light weight materials in geotechnical applications.

93