Western University

Scholarship@Western
Electronic Thesis and Dissertation Repository

March 2014

Static and Seismic Performance of Geosynthetics-

Strengthened Pile Foundations
Ahmed M. Taha
The University of Western Ontario

Supervisor
Dr. M. Hesham El Naggar
The University of Western Ontario

Graduate Program in Civil and Environmental Engineering

A thesis submitted in partial fulfillment of the requirements for the degree in Doctor of Philosophy

© Ahmed M. Taha 2014

Follow this and additional works at: http://ir.lib.uwo.ca/etd

Part of the Geotechnical Engineering Commons

Recommended Citation
Taha, Ahmed M., "Static and Seismic Performance of Geosynthetics-Strengthened Pile Foundations" (2014). Electronic Thesis and
Dissertation Repository. Paper 1901.

This Dissertation/Thesis is brought to you for free and open access by Scholarship@Western. It has been accepted for inclusion in Electronic Thesis
and Dissertation Repository by an authorized administrator of Scholarship@Western. For more information, please contact [email protected].
STATIC AND SEISMIC PERFORMANCE OF GEOSYNTHETICS-
STRENGTHENED PILE FOUNDATIONS

by

Ahmed Taha

Graduate Program in Civil and Environmental Engineering

A thesis submitted in partial fulfillment

of the requirements for the degree of
Doctor of Philosophy

The School of Graduate and Postdoctoral Studies

The University of Western Ontario
London, Ontario, Canada

© Ahmed Taha 2014

i
 Abstract
Geosynthetic reinforcement in earth structures has been used extensively over the last
two decades. Extensive research has been carried out to investigate solutions to enhance
the lateral stability of pile foundations. This research is motivated by the need to install
piles in sites characterized by soft subsurface soil conditions, and often times, in seismic
active areas. This research work explores an innovative use of geosynthetics to enhance
the lateral performance of pile foundations. The static and seismic soil-structure-
interaction behaviors of geosynthetics-reinforced pile foundation systems were evaluated
using a series of reduced scale physical model tests performed on a shaking table in a 1G
environment. A laminar shear box containing a pile foundation model supporting a single
degree of freedom structure installed in different soil bed models was used in the
experiments. The soil models included: a layer of synthetic clay (Modified Glyben)
underlain by a sand layer (simulating a base case of soft soil); a layer of synthetic clay
sandwiched between a sand layer and an aggregate layer (simulating the case of
conventional ground replacement for the top soft soil); and a layer of synthetic clay
sandwiched between a sand layer and a geosynthetic-reinforced aggregate layer
(simulating the case of ground replacement of the top soft soil combined with
geosynthetic reinforcement using a microgrid mesh). A series of sine-sweep, harmonic
and scaled earthquake tests have been performed to identify the amplification and
resonance conditions of the foundation system and to identify various aspects of seismic-
soil-pile-geosynthetic reinforcement interaction effects. Lateral static load tests of this
system were performed using a one directional load system that was fixed on the laminar
shear box. The dynamic and static tests were simulated employing numerical models
developed using the finite element program Plaxis 3D. The results of both static and
dynamic tests showed that the microgrid reinforcement improved the lateral performance
of the pile foundation and reduced the vibration amplitudes of the supported structure.
The numerical analysis results were in close agreement with the dynamic and static
experimental results. The results of a parametric study for the investigated foundation
configuration and seismic loading demands showed that the requirements for engineered

ii
backfill can be reduced by more than 50% and the lateral seismic response can be
reduced by 50% by using geosynthetic reinforcement.
Key words: Seismic, static, piles, polymer strips reinforcement, microgrid mesh,
geogrid, soft clay, Shaking table tests, Scaled model tests, Layered soil, Plaxis 3D.

iii
 Co-Authorship Statement
This thesis is prepared in accordance with the regulation for Integrated-Article format
thesis stipulated by the school of graduate and post graduate studies at Western
University, London, Ontario, Canada. All the field testing, numerical modeling,
interpretation of results and writing of the draft and the final thesis were carried out by
the candidate himself, under the supervision of Dr. M. Hesham El Naggar. The
supervisor's contribution consisted of providing advice throughout the research program,
and reviewing the draft and the final thesis and publications results from this research.
The results of the field tests, and numerical modeling presented will be used in journals
and conferences publications, which will be co-authored with Dr. El Naggar.
Chapters 2, 3 and 4 of this thesis are the current versions of manuscripts that will be
submitted for publication as papers. Chapters 2 to 4 inclusive are coauthored by A. M.
Taha, M.H. El Naggar and A. Turan as described below:
Chapter 2: Experimental and Numerical Study on Lateral Behavior of Model

Geosynthetic-Reinforced Pile-Foundation Systems Subjected to Static Horizontal

Loading.

A version of chapter 2 will be submitted to the Geotextiles and Geomembranes Journal

Chapter 3: Experimental Study on the Dynamic Lateral Behavior of Geogrid

Reinforced Pile Foundation System

A version of chapter 3 will be submitted to the Geotextiles and Geomembranes Journal

Chapter 4: Numerical Study on the Dynamic Lateral Behavior of Geosynthetics-

Reinforced Pile Foundation System

A version of chapter 4 will be submitted to the Journal of Soil Dynamics and Earthquake

Engineering.
iv
 Acknowledgments
In the name of God, who is most compassionate and most merciful. All my praise is to
Allah, the Lord of the worlds. He is the Master of the Day of Judgment and keeps us on
the right path. We all depend on Him.
I would like to express my heartfelt appreciation and thanks to my supervisors Dr. M. H.
El Naggar and Dr. Alper Turan. The outcomes of this research could not have been
achieved without their partnership and support. I would like to acknowledge their
continuous support during my research. They were always friendly available with their
wise guidance and expertise.

I would like to thank Dr. Ashraf El-Damaty and Dr. Tim Newson and other members of
UWO Civil Engineering faculty, from which a gained lots of knowledge related to my
research work. I would also like to express my thanks for the kind assistance given by the
staff of the Boundary Layer Wind Tunnel (BLWTL) and University Machine Services
(UMS). I would like to specially thank Mr. Anthony Burggraaf and Gerry Dafoe for
training me on the shaking table and measuring instruments. I would like to thank M.
Richardson, W. Logan and Lama El-Naggar for their support in geotechnical laboratory
tests. I would like to thank Cody Ruthman and other UMS staff, who helped with the
assembling and mounting of the laminar soil container, this work could not have been
established without them.

I would like to express my gratitude to my father Mohammed Taha and my mother

Khadiga Saleh for their support and worm feelings during my study period. I would like
to express special thanks to my wife Hana and my brother Abdullah and all of my family
members for their extraordinary support and sacrifices.

v
I would like to express my deep thanks to Taibah University and the Saudi Cultural
Bureau of the Saudi Arabian Embassy for funding my scholarship and a great part of this
research at the University of Western Ontario.

vi
 Table of Contents

Abstract ........................................................................................................................................... ii

Co-Authorship Statement............................................................................................................... iv

Acknowledgments ........................................................................................................................... v

List of Tables ................................................................................................................................. xi

List of Figures .............................................................................................................................. xiii

NOMENCLATURE .................................................................................................................. xviii

Chapter 1 ......................................................................................................................................... 1

1.1 Introduction and problem overview .................................................................................... 1

1.2 Problem Definition and Methodology ................................................................................ 2

1.3 Research Objectives and plan ............................................................................................. 3

1.4 Literature Review on Related Studies................................................................................. 3

1.4.1 Dynamic performance of geogrid reinforced retaining walls ................................. 4

1.4.2 Numerical modeling of seismic behavior of geogrid reinforced

earth retaining structures ......................................................................................... 5

1.4.3 Nonlinear seismic performance of piles.................................................................. 6

1.4.4 Geogrid reinforced soils and piled supported systems. .......................................... 7

1.4.5 Background on polymer strips ................................................................................ 9

1.5 Contributions..................................................................................................................... 10

1.6 Scope of work ................................................................................................................... 10

1.7 Thesis Layout .................................................................................................................... 12

2. EXPERIMENTAL AND NUMERICAL STUDY ON LATERAL BEHAVIOR

OF MODEL GEOSYNTHETIC-REINFORCED PILE-FOUNDATION
SYSTEMS SUBJECTED TO STATIC HORIZONTAL LOADING ..................................... 17

2.1 Introduction ....................................................................................................................... 18

2.2 Background ....................................................................................................................... 20

vii
 2.3 Methodology ..................................................................................................................... 21

2.3.1 Model soil ............................................................................................................. 21

2.3.2 Soil parameters ...................................................................................................... 22

2.3.3 The model foundation system ............................................................................... 25

2.3.4 The Model geogrid ................................................................................................ 27

2.3.5 Model Stratigraphy ............................................................................................... 28

2.3.6 Placement of geogrid-reinforced engineered granular fill .................................... 29

2.3.7 Loading System and Instrumentation ................................................................... 30

2.3.8 Numerical model ................................................................................................... 33

2.3.9 Experimental and numerical test cases ................................................................. 36

2.4 Results and discussion ...................................................................................................... 37

2.4.1 Pile cap translation ................................................................................................ 37

2.4.2 Extension of microgrid (in Deep backfill case) .................................................... 42

2.4.3 Results of Numerical Analysis .............................................................................. 45

2.5 Summary and Conclusions ............................................................................................... 49

3. EXPIRMENTAL STUDY ON THE DYNAMIC LATERAL BEHAVIOR OF

GEOGRID REINFORCED PILE FOUNDATION SYSTEM ................................................ 54

3.1 Introduction ....................................................................................................................... 55

3.2 Background ....................................................................................................................... 56

3.3 Methodology ..................................................................................................................... 57

3.3.1 Reduced scale physical model .............................................................................. 57

3.3.2 The shaking table and soil container ..................................................................... 59

3.3.3 The model soil column.......................................................................................... 60

3.3.4 Model Foundation System .................................................................................... 62

3.3.5 Model Geosynthetics ............................................................................................ 64

viii
 3.3.6 Model Superstructure ............................................................................................ 65

3.3.7 Instrumentation ..................................................................................................... 68

3.3.8 Shaking Table Control System ............................................................................. 69

3.3.9 Placement of Model Soil Deposit ......................................................................... 70

3.3.10 Experimental Scenarios ........................................................................................ 72

3.4 Results and discussion ...................................................................................................... 76

3.4.1 Free Field Response .............................................................................................. 77

3.4.2 Pile cap response ................................................................................................... 78

3.4.3 Low Frequency Structure Response ..................................................................... 81

3.4.4 High Frequency Structure Response ..................................................................... 83

3.4.5 Pile cap rocking response ...................................................................................... 85

3.5 Summary and conclusion .................................................................................................. 88

4. NUMERICAL STUDY ON THE DYNAMIC LATERAL BEHAVIOR OF

GEOSYNTHETICS-REINFORCED PILE FOUNDATION SYSTEM ................................. 92

4.1 Introduction and problem overview .................................................................................. 92

4.2 Shaking table test description ........................................................................................... 94

4.3 The numerical model ........................................................................................................ 97

4.3.1 Problem dimensions .............................................................................................. 97

4.3.2 Interface conditions ............................................................................................... 99

4.3.3 Material models and parameters: ........................................................................ 100

4.3.4 The numerical model elements and mesh configuration .................................... 111

4.3.5 The numerical calculation process ...................................................................... 111

4.4 Free Field Ground Response Analysis Using Deep Soil ................................................ 112

4.5 Numerical Model Calibration Results............................................................................. 116

4.6 Dynamic Model and Its Verification .............................................................................. 116

ix
 4.6.1 Soil Model ........................................................................................................... 117

4.6.2 Piles-cap-geogrid-foundation model ................................................................... 120

4.6.3 Interface model ................................................................................................... 121

4.6.4 Ground response analysis ................................................................................... 123

4.7 Parametric Study Considering Dynamic Analysis .......................................................... 125

4.7.1 Effect of geogrid reinforcement on response to shaking with

varying intensity.................................................................................................. 126

4.7.2 Effect of geogrid reinforcement on response of foundation

subjected to shaking with different frequency .................................................... 128

4.7.3 Effect of engineered backfill thickness ............................................................... 130

4.7.4 Effect of stiffness of geosynthetics reinforcement on the system

performance ........................................................................................................ 131

4.8 Parametric Study Considering Pseudo-Static Analysis .................................................. 132

4.8.1 Building, pile cap model and geogrid ................................................................. 132

4.8.2 Pseudo-static seismic lateral loading calculation ................................................ 136

4.8.3 Summary of numerical pseudo-static analyses ................................................... 138

4.9 Results and Discussion ................................................................................................... 139

4.9.1 Effect of geosynthetic material stiffness and depth ............................................ 140

4.10 Summary and Conclusions ............................................................................................. 142

5. Summary and Recommendations for Further Research ......................................................... 149

5.1 Summary of thesis findings ............................................................................................ 149

5.2 Recommendations for Future Research .......................................................................... 153

Appendix A ................................................................................................................................. 154

Curriculum Vitae ........................................................................................................................ 155

x
 List of Tables

Table 2-1: Scaling relationships for primary system variables (Meymand, 1998). ...................... 21

Table 2-2 : Soil parameters ........................................................................................................... 24

Table 2-3: Model Piles calculations .............................................................................................. 26

Table 2-4: Engineering properties of Microgrid geogrid .............................................................. 27

Table 2-5: Summary of experimental and numerical cases .......................................................... 36

Table 3-1: Scaling relationships for primary system variables (Meymand, 1998). ...................... 59

Table 3-2: Model Piles calculations .............................................................................................. 64

Table 3-3: Engineering properties of geogrid ............................................................................... 65

Table 3-4: SDOF structures model and prototype frequencies ..................................................... 66

Table 3-5: Summary of tests ......................................................................................................... 73

Table 3-6: Summary of harmonic tests ......................................................................................... 76

Table 4-1: Granular Soil parameters ........................................................................................... 101

Table 4-2: The HSSMALL clay and aggregate parameters:....................................................... 109

Table 4-3: DeepSoil input parameters; model soil column. ....................................................... 115

Table 4-4: The HSSMAL prototype parameters of each soil layer ............................................ 119

Table 4-5: DeepSoil input parameters; prototype soil column. .................................................. 124

xi
Table 4-6: Summary of numerical model verification analyses ................................................. 125

Table 4-7: HSSMALL Soil parameters of each soil layer adopted for the equivalent static
study ............................................................................................................................................ 135

Table 4-8 : Uniaxial Geogrid GG200PE specifications .............................................................. 136

Table 4-9: Equivalent static forces and overturning moment calculation .................................. 138

Table 4-10: Equivalent static analysis cases ............................................................................... 139

Table 4-11: Cases 1-8 results comparison. ................................................................................. 141

xii
 List of Figures
Figure 1-1: Aggregate interlocking with geogrid (www.windfarmbop.com) ..................... 1

Figure 1-2: Component parts of Reinforced Earth Wall (Holtz, 1997) .............................. 4

Figure 1-3 Reinforcement strips extending from a mechanically stabilized wall during
construction (www.recocanada.com). ............................................................................... 10

Figure 2-1: Construction sand gradation ........................................................................... 22

Figure 2-2: The scaled piled cap geogrid model ............................................................... 26

Figure 2-3: Compaction of modified glyben .................................................................... 29

Figure 2-4: Geogrid extending between the ring and the metal disc ................................ 30

Figure 2-5: a. Schematic diagram of the testing setup, b. top view of the testing setup. .. 32

Figure 2-6: The Wire Line Position Sensors mounted on a frame and tied to the wires .. 33

Figure 2-7: a) Plaxis 3D numerical model of Soil and pile-cap-geogrid system b) Finite
element mesh. ................................................................................................................... 35

Figure 2-8: Lateral load-displacement response of model foundation and results of

numerical simulations (Cases #1 and 4). .......................................................................... 38

Figure 2-9: Lateral load-displacement response of model foundation and results of

numerical simulations (Cases #2 and 5). .......................................................................... 39

Figure 2-10: Lateral load-displacement response of model foundation and results of

numerical simulations (Cases # 3 and 6) .......................................................................... 40

Figure 2-11: experimental results comparison (cases 1-3) ............................................... 41

Figure 2-12: Numerical results comparisons (cases 5-8) .................................................. 42

xiii
Figure 2-13: Microgrid strain vs. pile cap pull load measured near to the piles face. ...... 43

Figure 2-14: Microgrid strain vs. pile cap pull load measured far from the piles face ..... 44

Figure 2-15: Global Microgrid deformation measured near and far from the pile face ... 45

Figure 2-16: Bending moment at pile head vs. pile cap lateral deflection........................ 46

Figure 2-17: Shear force at pile head vs. pile cap lateral deflection ................................. 46

Figure 2-18: Pile cap deflections for microgrid lengths of 90 cm and 20 cm................... 47

Figure 2-19: Maximum bending moments for microgrid lengths of 90 cm and 20 cm. ... 48

Figure 2-20: Maximum shear forces for microgrid lengths of 90 cm and 20 cm ............. 49

Figure 3-1: Laminar soil container covered with latex membrane and aligned with corner
profiles .............................................................................................................................. 60

Figure 3-2: Construction sand gradation ........................................................................... 61

Figure 3-3: The scaled piled cap geogrid model ............................................................... 63

Figure 3-4: a) The SDOF High system b) SDOF Medium system, c) ADXL 203
Accelerometer ................................................................................................................... 67

Figure 3-5: Instruments distribution diagram ................................................................... 69

Figure 3-6: The shaking table and control system ............................................................ 70

Figure 3-7: Compaction of Modified Glyben ................................................................... 72

Figure 3-8: The earthquake acceleration time history ...................................................... 74

Figure 3-9: Fourier spectrum of model scale Upland EQ (a) input signal (b) signal at
table top ............................................................................................................................. 75

xiv
Figure 3-10: Free field response comparison (SAMP-1 to SAMP-4). ............................. 78

Figure 3-11: The dynamic response comparison of the piled cap foundation (SLNF-1 to
SLNF-3). ........................................................................................................................... 79

Figure 3-12: Dynamic response of pile foundation with and without reinforcement
(UPEQ-L). ......................................................................................................................... 80

Figure 3-13: The dynamic response comparison of the piled cap foundation (SHNF 1-3)
........................................................................................................................................... 81

Figure 3-14: The dynamic response comparison of the low frequency SDOF
superstructure (SLNF-1 to SLNF-3). ................................................................................ 82

Figure 3-15: The dynamic response comparison of the low frequency SDOF
superstructure (UPEQ-L). ................................................................................................. 83

Figure 3-16: The dynamic response comparison of the high frequency SDOF (SHNF-1 to
SHNF-3). ........................................................................................................................... 84

Figure 3-17: The dynamic response comparison of the high frequency SDOF structure
(UPEQ-H). ........................................................................................................................ 85

Figure 3-18: The dynamic vertical response comparison measured at the pile tip (SLNF-1
to SLNF-3). ....................................................................................................................... 86

Figure 3-19: Vertical response at the pile tip comparison (SAMP-4) .............................. 87

Figure 4-1: The schematic of the shaking table test setup ................................................ 94

Figure 4-2: Input motion acceleration time history used in the shaking table test ........... 95

Figure 4-3: Fourier spectrum of the input motion used in the shaking table test ............. 96

Figure 4-4: Pile cap acceleration response ........................................................................ 96

xv
Figure 4-5 : Model pile foundation and microgrid connectivity....................................... 98

Figure 4-6: 3D FE model of geosynthetic reinforced pile-cap system ............................. 99

Figure 4-7: Typical sand shear stress-strain loop (calculated from the shaking table tests).
......................................................................................................................................... 102

Figure 4-8: Typical aggregate shear stress-strain loop (calculated from the shaking table
tests). ............................................................................................................................... 103

Figure 4-9: Glyben shear stress-strain loop (calculated from the the shaking table tests).
......................................................................................................................................... 104

Figure 4-10 : Hyperbolic stress-strain relationship in primary loading for a slandered

drained triaxial test (Schanz et al., 1999) ........................................................................ 106

Figure 4-11: Stiffness degradation for modified glyben (Turan et al., 2009a) ............... 108

Figure 4-12 : Mean curves defining G/Gmax versus γ relationships for gravelly soils at
various confining pressures along with standard deviation boundaries for reduced data set
(Rollins et al. ,1998)........................................................................................................ 113

Figure 4-13 : Mean Curves Defining D versus γ relationships for gravelly soils at various
confining pressures along with standard deviation boundaries for entire data set (Rollins
et al. ,1998). .................................................................................................................... 114

Figure 4-14: Measured vs. calculated acceleration using DeepSoil. .............................. 115

Figure 4-15: Calculated vs. measured responses of pile cap reinforced with geogrid.... 116

Figure 4-16 : Soil model general view ............................................................................ 117

Figure 4-17 : a) Soil column stratigraphy without geogrid, b) Soil column stratigraphy
after adding geogrid ........................................................................................................ 118

Figure 4-18: Pile-Cap–geogrid finite element mesh. ...................................................... 120

xvi
Figure 4-19: Pile cap translation, Static pull test vs. 35% improved interface modulus 123

Figure 4-20: Ground response analysis, 3D finite element model vs. DeepSoil. ........... 124

Figure 4-21: Cap motion comparison, Cases of 2 m backfill with and without geogrid
(0.1 g, 3 Hz). ................................................................................................................... 126

Figure 4-22: Cap motion comparison, Cases of 2 m backfill with and without geogrid
(0.2 g, 3 Hz). ................................................................................................................... 127

Figure 4-23: Cap motion comparison, Cases of 2 m backfill with and without geogrid
(0.3 g, 3 Hz). ................................................................................................................... 128

Figure 4-24: Cap motion comparison, Cases of 2 m backfill with and without geogrid
(0.2 g, 2 Hz). ................................................................................................................... 129

Figure 4-25: Cap motion comparison, Cases of 2 m backfill with and without geogrid
(0.2 g, 5 Hz). ................................................................................................................... 129

Figure 4-26: Pile cap acceleration time history comparison , ( cases of 1.6 m with
geogrid, 1.6m without geogrid and 6 m backfill without geogrid). ................................ 130

Figure 4-27: Effectiveness of conventional geogrid vs. polymer strips in reducing pile
cap response .................................................................................................................... 131

Figure 4-28 : Pile, cap, geogrid and superstructure model ............................................. 133

Figure 4-29 : Plaxis model 3D view ............................................................................... 134

Figure 4-30: Deformed model shape .............................................................................. 140

xvii
 NOMENCLATURE

c cohesive strength

cu undrained shear strength

Dr relative density

γ soil unit weight

G/Gmax shear modulus reduction factor

G dynamic shear modulus

G max maximum shear modulus

Gs secant shear modulus

gw/c ratio of fluids to solids by mass

M mass matrix

K stiffness matrix

C damping matrix

force matrix

Δt time interval

T fundamental period of superstructure

higher mode factor

Rd , Ro Ductility factors

xviii
V seismic base shear force

S(T) design spectral acceleration

seismic importance factor

W seismic dead load

(Vs)m shear wave velocities for the prototype

(Vs)p shear wave velocities for the model

w water content

w/gw ratio of water to water plus glycerin by mass

ε1 axial strain

φ friction angle

ψ dilatancy angle

γ shear strain

γr threshold shear strain

shear strain at Gs = 0.70 G0

λ geometric scaling factor

ν Poisson’s ratio

νur Unloading reloading poisson’s ratio

xix
ωn fundamental angular frequency

ρ mass density

σv overburden stress

σx confining stress

σ’3 minor principle stress

Pref reference pressure

E Young’s modulus

Ei initial stiffness

Ep Young’s modulus of pile material

E50 secant stiffness modulus at 50% strength

reference secant stiffness modulus at 50 % strength

oedometer loading modulus

reference oedometer loading modulus

elastic unloading reloading modulus

reference elastic unloading reloading modulus

m amount of stress dependency power

Jp prototype stiffness

Jm model stiffness

xx
q the deviatoric current shear stress

qa the deviatoric asymptotic value of the shear stress

qf the ultimate diviatoric stress

Rinterface structure/ soil interface strength ratio

τ shear stress

τmax maximum shear stress

D damping ratio

K2max shear modulus number for the soil = 50 for loose sand and 75 for dense
sand

σ’m mean effective confining stress of the soil in psf

xxi
 Chapter 1

1.1 Introduction and problem overview

Geogrids are polymeric material consisting of tensile ribs with openings of sufficient size
to allow interlock with the surrounding soil. This geogrid-soil interlock mechanism
allows the geogrid mesh to work as a reinforcement element, which enhances the soil
shear strength. Therefore, geogrids have been widely used in modern construction
technology. Several application examples include: geogrid reinforced earth retaining
walls (GRS); highway construction and expansion over soft soils; geogrid-reinforced
pile-supported highway embankments; and geogrid reinforced slopes. The use of
geogrids in flexible pavement construction is widely popular and demonstrates its
advantageous interaction with the aggregate used. The geogrid mesh is laid within the
aggregate base course and provides increased modulus and lateral confinement for the
crushed stones intruding the apertures of the geogrid (figure 1-1).

Figure 1-1: Aggregate interlocking with geogrid (www.windfarmbop.com)

This confinement system prevents the base course stones from dispersing apart under the
cyclic traffic loads (Koerner, 2005).

In other applications, geogrids have been implemented to reduce lateral wall deflections
arising from dynamic loads and uneven settlement of the supporting sub grades and
embankments. Due to the rising concern in the construction industry with regards to
mitigating the destructive effects of cyclic earthquake loads, this thesis explores the

1
development of a novel pile foundation system that incorporates the benefits of
geosynthetics reinforcement to enhance the lateral performance of pile foundation
systems.

1.2 Problem Definition and Methodology

Meeting serviceability limits in strong earthquake events has become the driving force in
the design for seismic safe structures. Therefore, researchers and engineers developed
several means to enhance the dynamic lateral stability of civil engineering structures. For
example, Geosynthetic reinforced earth retaining structures were constructed to sustain
high earthquake shaking. Therefore, this thesis investigates the effect of geosynthetics
reinforcement on the lateral response of low frequency and relatively higher frequency
superstructures supported by pile foundations. The investigation is carried out on two
model superstructures mounted on a model pile foundation and subjected to 1-G shaking
on a shaking table. The model geogrid and piles are embedded in a layered soil system
that includes a synthetic soft clay layer sandwiched between a granular backfill from top
and well graded sand from bottom. This composite system is subjected to three kinds of
1-D base excitation including, i) sine sweep, ii) harmonic iii) and scaled earthquake
motions. The research also investigates the static lateral response of the geogrid pile
composite system through subjecting the foundation cap to a static lateral pull
mechanism. The results derived from the composite system are compared with the base
case where the model pile foundation is not reinforced and another case with thicker
granular backfill layer simulating conventional ground replacement solution to improve
lateral response of piles. In addition, static and dynamic numerical models of the pile-
geogrid reinforcement composite system were developed and were verified using the
experimental results. The numerical models were then used to conduct a parametric study
to further examine the effects of different parameters on the lateral performance of the
pile-geogrid reinforcement composite system.

2
1.3 Research Objectives and plan
To the best of the author’s knowledge, the use of geosynthetics to improve the seismic
performance of pile foundations and to reduce intolerable pile lateral deflections have not
been investigated. The primary focus of this thesis is to prove that geosynthetic
reinforcement can enhance the lateral stability of pile foundation. Therefore, the
following plan was carried out:
 To carry out static lateral pull tests to investigate the static behavior of the
geogrid reinforced piled cap system.

 To carry out 1-G shaking table tests to investigate the dynamic behavior of the
geogrid reinforced piled cap system.

 To develop numerical static and dynamic models for the composite pile-geogrid
reinforcement system and calibrate/verify this model with the experimental
results.

 To carry out a parametric study to investigate the favorable effect of

geosynthetics enhanced pile foundation over a range of base excitation
amplitudes and frequencies. Also, to study the effect of changing the geogrid
stiffness on the lateral performance of the geosynthetics-reinforced pile
foundation system.

1.4 Literature Review on Related Studies

This thesis covers several research fields related to the individual components of the
geogrid reinforced pile foundation system investigated. Extensive research has been
carried out to investigate the soil-pile-superstructure interaction problem using 1-G
shaking. Also, the nonlinear soil-pile interaction problem has been widely studied both
experimentally and numerically. On the other hand, several researches have investigated
the performance of geogrid reinforced earth retaining structures and embankments
subjected to dynamic loading. This research was conducted employing finite element

3
numerical models that were calibrated against experimental results. The following
sections present some of the studies related to the studied problems.

1.4.1 Dynamic performance of geogrid reinforced retaining walls

The first Geosynthetic reinforced retaining wall was built in France in 1971.
Geosynthetic reinforced walls include reinforcing geogrids or geotextile as reinforcing
elements embedded in the engineered fill to counteract against the lateral earth pressure.
Compared to the conventional earth retaining walls, they are more flexible and can
support higher earth fills. Therefore, they suit seismic active sites (Holtz, 1997) (figure 1-
2).

Figure 1-2: Component parts of Reinforced Earth Wall (Holtz, 1997)

Geogrid-reinforced retaining walls performed very well in major seismic events in the
US and Taiwan. Sandri (1997) surveyed three geosynthetic-reinforced walls and four
geosynthetic-reinforced slopes that showed no visual evidence of distress. Also, 11
geogird-reinforced segmental retaining walls and slopes exceeding 4.6 m in height
survived the North Ridge earthquake in 1994, which had a magnitude of 6.7 (Sandri,
1997). Ling et al. (2001) evaluated several geosynthetic-reinforced soil retaining walls
and slopes during a major earthquake event in Taiwan. Their investigation revealed that
the geosynthetic-reinforced structures performed better than the unreinforced structures.

4
Moreover, his investigation emphasized the importance of proper seismic design of
geosynthetic-reinforced soil retaining structures.

Several other researchers investigated the performance of geosynthetic-reinforced walls

and arrived at the same conclusion in which several GRS walls were observed to survive
major earthquake events (e.g. White and Holtz (1996), Eliahu and Watt 1991, Collin et
al. (1992) and Nishimura et, al. (1996)). Additionally, several researches have been
undertaken to analyze and design geosynthetic-reinforced walls for satisfactory seismic
performance. A broad review of the techniques used for seismic analysis of geosynthetic-
reinforced walls, slopes and embankments was carried out by Bathurst and Alfaro
(1996). Bathurst and Alfaro (1996) argued that these design and analysis techniques are
based on approximated analytical solutions. They recommended using the enhanced
abilities of existing geotechnical finite element models, calibrated using experimental
simulation procedures, for tackling the complex geosynthetic-reinforced soil structures
problems with greater accuracy.

1.4.2 Numerical modeling of seismic behavior of geogrid reinforced

earth retaining structures
Several researchers investigated the seismic performance of geogrid-reinforced walls
using finite element and finite difference models. Bathrust and Hatami (1998) studied the
influence of reinforcement length, stiffness and layers spacing on the response of
reinforced walls to seismic loads using numerical modeling. Their study indicated that
the static and dynamic wall lateral displacement can be altered depending on the
reinforcement length, stiffness and distribution. Helwany et al. (2001) executed a finite-
element model to study the seismic behavior of a 0.9 m model segmental wall subjected
to EQ loading generated by a shaking table using the finite element program DYNA3D.
In their finite element simulation, they used a soil model that accounted for the nonlinear
hysteretic behavior of the backfill soil. They concluded that the results obtained from the
finite element simulation were consistent with the experimental results. Burke et al.
(2004) analyzed the seismic response of a full-scaled reinforced retaining wall using a 2-
D finite element model. Ling et al. (2004) carried out a dynamic finite element procedure
5
to analyze the behavior of geosynthetic-reinforced soil retaining wall. El-Emam and
Bathrust (2007) formulated a dynamic finite difference model to study the influence of
reinforcement parameters on the seismic response of a reduced-scale reinforced soil
retaining wall. Fakharian and Attar (2007) also used a dynamic finite difference model to
investigate the seismic behavior of a block-faced geogrid reinforced soil bridge
abutment. Liu (2009) investigated the reinforcement loads of a GRS wall subjected to
seismic loading during the service life using a finite element procedure. Anastasopulos et
al. (2010) used a numerical model to analyze the seismic performance of a bar–mat
retaining wall at model scale employing the finite element program ABAQUS. The
aforementioned research proved that both finite difference and finite element numerical
models are powerful tools for prediction of the performance and nonlinear behavior of
soils and earth retaining structures.

1.4.3 Nonlinear seismic performance of piles

Several researches have been undertaken to model the nonlinear seismic response of pile
foundations (e.g. El Naggar and Novak, 1996; El Naggar and Bentley, 2000; El Naggar
et al., 2005).

El Naggar and Novak 1996 developed an analytical model for the lateral response of
single piles and pile groups. This model include: 1) soil reaction elements that account
for the state of stress and gapping and slippage at both sides of the pile; 2) inner field
zone elements to model the nonlinear soil behavior by nonlinear spring that represent the
stiffness and a dashpot to simulate the soil hysteretic damping; 3) far field zone elements
that simulate the linear stiffness and damping and account for the propagation of waves
and energy dissipation; 4) pile modeled as conventional two node beam elements .

El Naggar and Bentley (2000) simulated the soil seismic nonlinearity and energy
dissipation using a spring with dynamic p-y curve model and a dashpot. The soil stiffness
in this simulation is derived from the dynamic p-y curve and is equivalent to the
combined inner and outer zones stifnesses. The dashpot is connected in parallel to the
unified spring and has a constant equivalent to the far field dashpot constant. The

6
computed lateral pile response from the dynamic p-y curve was in good agreement with
the measured results of a Statnamic load test (El Naggar, 1998). Moreover, El Naggar
and Bentley (2000) established an equation from regression analysis that relates the
dynamic soil resistance with the loading frequency and the static soil reaction. This
equation is used to generate dynamic p-y curves from the static curves in order to get
better estimation of the response of the structures-soil system to dynamic loading.
Dynamic p-y curves were implemented in the finite element program ANSYS. The
ANSYS simulation showed good agreement between the analytical and dynamic p-y
lateral models results.

El Naggar et al. (2005) developed a simplified Beam on Nonlinear Winkler Foundation

(BNWF) model for nonlinear seismic response analysis of offshore piles. The (BNWF) is
a simplified approach that accounts for nonlinear seismic-soil-pile-structure and is
commonly used in engineering practice. The (BNWF) uses spring based on the p-y curve
approach to represent soil resistance and dashpots to account for energy dissipation and
the hysteretic behavior of the soil. The pile was modeled as discrete beam column
elements. The damping of the pile segments was determined using the Rayleigh
approach. The discontinuity and gapping conditions at pile-soil interface were modeled
using special interface elements that account for relative movements of the pile against
soil. The displacement-time or acceleration time histories at different soil layers were
computed using Iwan-Mroz model for stress-strain and tangential shear modulus relation.
The result of this free field analysis was then used as the input motion at BNWF support
nodes. The comparison of El Naggar et al. (2005) finite element model results and the
experimental centrifuge test carried out by Wilson et al (1997) indicated a good
agreement when the pile damping ratio is between 3% and 5%.

1.4.4 Geogrid reinforced soils and piled supported systems.

Little research has been carried out to study the combined effect of piled geogrid
systems. Liu et al. (2007) presented a case history of a geogrid-reinforced and pile-
supported (GRPS) highway embankment. In this case history, ABAQUS finite-element
software was used to back analyze a (3D) fully coupled model. The horizontal length of
7
the finite-element mesh was three times the width of half the embankment base in order
to minimize the boundary effects. The piles were modeled by a 20-node quadratic brick
without pore pressure degrees of freedom. The interface soils and other soils were
modeled by 20-node quadratic displacement, with eight more excess pore pressure
degrees of freedom. The geogrid were modeled by eight-node quadrilateral that have a
capacity to resist only tensile force. The model of the pile was chosen as an isotropic
linear elastic material. The embankment fill, gravel and the surface coarse-grained fill,
was modeled using as linear elastic-perfectly plastic with Mohr–Coulomb failure
criterion. The four foundation soils were modeled as modified cam clay materials. The
comparison of the numerical model and field observation results showed that for
embankment higher than 2.5 m, the computed stress reduction ratio was consistent with
the measured values.

Another system that combines the favorable effect of the geosynthetic-reinforcement and
piles is geosynthetic-reinforced and pile-supported earth platforms. This system is
constructed over soft soils to build superstructures, such as tanks and embankments in a
single stage without long-standing waiting times. Also, building this foundation system
can significantly reduce total and differential settlements. Furthermore, geosynthetic-
reinforced and pile-supported earth platforms can reduce earth pressures and avoid
expensive excavation and refill engaged in typical situations. Compared with
unreinforced pile supported earth platforms, the geosynthetic-reinforcement can reduce
the settlement between pile caps. Han and Gabr (2002) carried out a numerical study on
the behaviour of geosynthetic-reinforced and pile-supported earth platforms over soft
foundations using the finite difference program FLAC 2D. They investigated the
influencing factors related to the height of embankments, tensile stiffness of
geosynthetic, and elastic modulus of pile material. The results were presented in terms of
settlement, stress concentration, and soil arching. The material constitutive model
adopted in this research was the nonlinear hyperbolic elastic model developed by Duncan
and Chang (1970). The results of this research verified that inclusion of geosynthetic in
earth platforms can reduce the total and differential settlements above the pile heads and
at the ground surfaces. The research also found that increasing the stiffness of the
8
geosynthetic reinforcement had a favourable effect in reducing soil arching. Han and
Gabr (2002) indicated that increasing the stiffness of the geosynthetic reinforcement
increased the stress concentration ratio and improved the stress transfer from the soft soil
to the pile. The study results also indicated that the maximum tension in geosynthetic
increases with increasing the stiffness of geosynthetic, increasing the height of
embankment fill and increasing the elastic modulus of the pile material.

1.4.5 Background on polymer strips

French architect/engineer Henri Vidal patented Reinforced Earth technology in 1963.
Simply, it involves combining layer of earthworks and tensile reinforcements to develop
a new and strong composite earth retaining structure. This technology is now commonly
used globally. Polymer strips are geosynthetic material developed to provide earth
reinforcement for earth retaining structures. Polymer strips to enhance the lateral stability
if an earth retaining structures known as mechanically stabilized wall. The mechanically
stabilized wall are built from modular blocks of various shapes and dimensions (figure 1-
2), and stacked together in the vertical direction. The polymer strips are embedded
between the modular blocks and extended into the soil. High frictional resistance develop
between the strips and soil and provide lateral stiffness to these kind of walls (figure 1-3)
(www.recocanada.com).

9
Figure 1-3 Reinforcement strips extending from a mechanically stabilized wall
during construction (www.recocanada.com).

1.5 Contributions
This thesis explored a novel means of enhancing the lateral performance of pile
foundations using geogrid reinforcement embedded in granular backfill. Before this
novel system, the use of geogrid reinforcement was limited to earth structures, pavements
and embankments applications. The static and dynamic results derived from this thesis
may encourage researchers to carry out further field investigations to examine the
quantitative performance of this novel system. Moreover, the dynamic and static tests
revealed the importance of using a well-designed backfill thickness in reducing the
lateral response of superstructures supported on piled foundations. The findings of this
thesis together with full scale tests may present an economical alternative to using thick
pile caps or large ground replacement efforts. The numerical models used in this thesis
can be used as the basis for studying this foundation system at the field scale.

1.6 Scope of work

This thesis encompasses static and dynamic experiments to investigate the lateral
performance of the geogrid-pile foundation system. The experimental testing results are
10
used to calibrate and verify static and dynamic finite element models. These numerical
models are then used to perform a comprehensive parametric study to evaluate the
important design parameters of the developed foundation system. To carry out these
tasks, the following steps are required:

1- Preparing synthetic soft clay (modified glyben) through mixing bentonite, water
and glycerin.

2- Construction of layered soil beds in the lamina container. These soil beds
included: a soil bed representing the base case (soft clay overlying sand layer); a
soil model representing the conventional ground replacement solution (the glyben
layer sandwiched between an aggregate layer from top and a sand layer from
bottom) and a soil model representing the proposed novel geosynthetic-reinforced
backfill (the glyben layer sandwiched between a geosynthetic-reinforced
aggregate layer from top and a sand layer from bottom).

3- Installing a geogrid mesh within the top granular backfill layer for the case of
geosynthetic-reinforced backfill.

4- Installing a 2X2 model pile group in the layered soil.

5- Fabricating then fixing the model low frequency and relatively high frequency
single degree of freedom structures on top of the pile cap foundation.

6- Performing lateral static pull tests on the pile foundation system with and without
geogrid reinforcement and comparing its response with both the base case and the
pile foundation system embedded in the backfill layer.

7- Subjecting the geogrid-reinforced pile foundation system to harmonic, sine sweep

and scaled earthquake 1-D base motions using a shaking table.

8- Subjecting the same pile foundation system without the geogrid mesh to the same
harmonic, sine sweep and scaled earthquake base motions using the shaking table.

11
 9- Subjecting the same pile foundation system without reinforcement within the
backfill layer to the same base excitation.

10- Employing the finite element program Plaxis 3D (Brinkgreve et. al, 2012) to
develop the static and dynamic numerical models and validating these models
with the experimental results.

11- Carrying out static numerical parametric study to investigate the effect of the
length and stiffness of the geogrid reinforcement on the lateral performance of the
geogrid pile foundation system.
12- Carrying out dynamic numerical parametric study to investigate the effect of the
input ground motion amplitudes and frequency on the geogrid pile foundation
system.

13- Carrying out a parametric study to investigate the effect of the depth and stiffness
of geogrid on the lateral performance of the geogrid pile foundation system
through the analysis of the response of a building subjected to pseudo-static
seismic lateral loading.

1.7 Thesis Layout

This thesis is organized in the following chapters:

Chapter 1 presents the objectives, the literature review, the original contributions of this
thesis, the scope of work and the organization of this thesis.

Chapter 2 presents the experimental work carried out to investigate the static lateral
behavior of the geogrid-reinforced pile foundation system. It describes the methodology
undertaken to apply a static lateral load on the pile cap and the extension behavior of the
geogrid mesh. Chapter 2 also describes the finite element model developed to simulate
the static lateral pull test. The results collected from the static test were used to verify this
numerical model. Also, a parametric study is carried out to investigate the effect of

12
length and stiffness of the geogrid reinforcement on the static lateral performance of the
geogrid pile foundation system.

Chapter 3 presents the experimental work carried out to investigate the dynamic lateral
behavior of the geogrid-reinforced pile foundation system. Chapter 3 explains the
methodology employed to construct the elements of the experimental work and the
details of the shaking table tests. Then, results of the shaking table tests are presented and
discussed, and conclusions are presented.

Chapter 4 presents a finite element model that is established to simulate the dynamic
behavior of the geogrid-reinforced pile foundation system. Chapter 4 also include an
investigation carried out to study the dynamic lateral behavior of the new foundation
system on a prototype scale. Chapter 4 describes the elements of the numerical model
and the soil constitutive models used to build the dynamic finite element model. The
results derived from Chapter 3 were used to verify this numerical model. In addition, a
parametric study is carried out to investigate the effect of the frequency of the base
motion and amplitude, and the soil reinforcement stiffness on the dynamic lateral
performance of the geogrid pile foundation system. Finally, a parallel seismic parametric
study was carried out to investigate the effect of geogrid stiffness and depth on the lateral
performance of the geogrid pile foundation system.

Chapter 5 presents the summary of the thesis and provides recommendations for future
research that can be carried out benefiting from the methodology and results of this
thesis.

References:

[1] Anastasopoulos, I., Georgarakos, T., Drosos, V., and Kourkoulis R. 2010. Seismic
performance of bar-mat reinforced-soil retaining wall: Shaking table testing versus
numerical analysis with modified kinematic hardening constitutive model. Soil
Dynamics and Earthquake Engineering, 13 ,1089-1105.

13
[2] Bathurst, R.J.and Alfaro, M.C.1996. Review of seismic design, analysis and
performance of geosynthetic rein- forced walls, slopes and embankments.
International Symposium on Earth Reinforcement, keynote lecture, Kyushu, 23-52.

[3] Bathurst, R.J.and Hatami, K., 1998. Seismic response analysis of a geosynthetic
reinforced soil retaining wall. Geosynthetics International, 5 (1–2), 127–166.

[4] Brinkgreve, R., Egin, E. and Swolfs, W. User’s Manual for Plaxis 3D 2012 , Delft
University of Technology & PLAXIS bv, The Netherlands.

[5] Collin, J. G., Chouery-Curtis, V. E., and Berg, R. R. 1992. Field observations of
reinforced soil structures under seismic loading. Proceedings of the International.
Symposium on Earth Reinforcement practice, H. Ochiai, N. Yasufuku, and K.
Omine, eds., Vol. 1, Balkema, Rotterdam, The Netherlands, 223–228.

[6] Cai, Z. and Bathurst, R.J. 1995. Seismic response analysis of geosynthetic
reinforced soil segmental retaining walls by finite element method. Computers and
Geotechnics, 17 (4), 523-546.

[7] El-Emam, M.M. and Bathurst, R.J. 2007. Influence of reinforcement parameters on
the seismic response of reduced-scale reinforced soil retaining walls. Geotextiles
and Geomembranes, 25 (1), 33-49.

[8] Eliahu, U. and Watt, S. 1991. Geogrid-reinforced wall withstands earthquake.

Geotechnical Fabrics Report, 9 (2), 8–13.

[9] El Naggar, M.H., 1998. Interpretation of Lateral Statnamic Load Test Results.
Geotechnical Testing Journal, ASTM, Vol. 21, No. 3, pp. 169-179.

[10] El Naggar, M.H. and Novak M. 1996. Influence of foundation nonlinearity on

offshore towers response. Journal of Geotechnical Engineering, 122 (9), 717-721.

[11] El Naggar, M.H. and Bentley, K.J. 2000. Dynamic analysis for laterally loaded
piles and dynamic p-y curves. Canadian Geotechnical Journal, 37 (6), 1166-1183.

14
[12] El Naggar, M.H., Shayanfar, A.S., Kimiaei, M. and Aghakouchak, A.A. 2005.
Simplified BNWF model for nonlinear seismic response analysis of offshore piles
with nonlinear input ground motion analysis. Canadian Geotechnical Journal. 42
(2), 365-380.

[13] Fakharian, K. and Attar, I.H. 2007. Static and seismic numerical modeling of
geosynthetic-reinforced soil segmental bridge abutments. Geosynthetics
International, 14 (4), 228-243.

[14] Han, J. and Gabr M. A. 2002. Numerical Analysis of Geosynthetic-Reinforced and

Pile-Supported Earth Platforms over Soft Soil. Journal of Geotechnical and
Geoenvironmental Engineering, 128 (1), 44-53.

[15] Helwany, S.M.B., Budhu, M., McCallen, D. 2001. Seismic analysis of segmental
retaining walls. I: Model verification. Journal of Geotechnical and
Geoenvironmental Engineering, 127 (9), 741-749.

[16] Holtz, R. D., Christorpher, B, R., and Berg, R. R., Geosynthetic Engineering 1 st ed,
BiTech Publishers Ltd, BC, Canada.

[17] Koerner, R. M., 2005, Designing With Geosynthetics 4th ed, Pearson Education
Ltd., New Jersey, USA.

[18] Ling, H.I., Liu, H.B., Mohri, Y., and Kawabata, T. 2001. A Bounding Surface
Model for Goegrid. Journal of Engineering Mechanics, ASCE, 127 (9), 963-967.

[19] Liu, H. 2009 .Analyzing the reinforcement loads of geosynthetic-reinforced soil

walls subject to seismic loading during the service life. Journal of Performance of
Constructed Facilities, 23 (5), pp. 292-302.

[20] Liu, H.L., Ng, C.W.W., and Fei, K. 2007. Performance of a geogrid-reinforced and
pile-supported highway embankment over soft clay: Case study, Journal of
Geotechnical and Geoenvironmental Engineering, 133 (12), pp. 1483-1493.

15
[21] Nishimura, J., Hirai, T., Iwasaki, K., Saito, Y., and Morikshima, M. 1996.
‘‘Earthquake resistance of geogrid reinforced soil walls based on a study conducted
following the southern Hyogo earthquake.’’ Proceedings., International.
Symposium. on Earth Reinforcement Practice ., IS-Kyushu ’96, Earth
Reinforcement Practice., H. Ochiai, N. Yasufuku, and K. Omine, eds., Balkema,
Rotterdam, The Netherlands.

[22] Sandri, D. 1997. A performance summary of reinforced soil structures in the

greater Los Angeles area after the Northridge earthquake, Geotextiles and
Geomembranes, 15 (4-6), pp. 235-253

[23] White, D. M., and Holtz, R. D. (1996). ‘‘Performance of geosynthetic reinforced

slopes and walls during the Northridge, California earthquake of January 17,
1994.’’ Proceedings. International. Symposium. on Earth Reinforcement Practice.,
IS-Kyushu ’96, Earth Reinforcement Practice., H. Ochiai, N. Yasufuku, and K.
Omine, eds., Balkema, Rotterdam, The Netherlands,Vol. 2, 965–972.

16
 Chapter 2

2. EXPERIMENTAL AND NUMERICAL STUDY ON

LATERAL BEHAVIOR OF MODEL GEOSYNTHETIC-
REINFORCED PILE-FOUNDATION SYSTEMS
SUBJECTED TO STATIC HORIZONTAL LOADING

This chapter introduces an innovative use of polymer strips as part of a novel foundation
concept, polymer strips reinforced pile foundation system, where polymer strips are used
to enhance the lateral static resistance of the pile foundation system. The lateral static
behavior of this foundation system was evaluated using a series of reduced scale physical
model test performed on the model foundation installed in different soil beds contained in
a small size soil container. These soil beds included: a soil bed representing the base
case (soft clay overlying sand layer); a soil model representing the conventional ground
replacement solution (the glyben layer sandwiched between an aggregate layer from top
and a sand layer from bottom) and a soil model representing the proposed novel
geosynthetic-reinforced backfill (the glyben layer sandwiched between a geosynthetic-
reinforced aggregate layer from top and a sand layer from bottom).

A one directional loading system was attached to the soil container containing a three
layer model soil deposit, which included a layer of synthetic clay (Modified Glyben)
sandwiched between a lower and a surficial granular layers, was used. The model pile-
cap system was installed through the surficial granular layer, which was reinforced using
a microgrid mesh and underlying clay and sand layers. A series of lateral static pull tests
have been conducted to investigate the influence of the polymer strips reinforcement on
the foundation system behavior. This study also presents the results of a 3D finite
element model calibrated against the physical test results. The experimental results and
numerical analyses showed that the lateral resistance of a pile cap system was enhanced
by the microgrid mesh embedded in the surficial granular layer. The results also showed

17
that the lateral resistance of the pile cap system was significantly influenced by the
thickness of engineered backfill.

2.1 Introduction

Piles are used to transfer vertical and lateral loads to competent soil layers along the pile
shaft and at its toe. Pile foundations are designed to provide a certain level of lateral
resistance while maintaining the lateral deformation within limits required by the
structural engineer. Examples of structures that are required to resist significant lateral
loads include (i) high rise buildings subjected to wind and/or earthquake loads, (ii)
marine structures subjected to horizontal forces due to the impact of vessels during
mooring and wave action, (iii) offshore structures subjected to wind and wave loads. Pile
foundations designed based on the serviceability limit offers a more cogent approach
because it accounts for the displacement limit of the structure (Zhang, 2009).

A significant number of full scale lateral pile load tests have been performed in
accordance with well-established standard procedures (e.g. ASTM D3966, ASTM
D7383). However, only a limited number of experimental studies using reduced scale
models were conducted to examine the static lateral behavior of piles. For example,
Maymand (1998) applied incremental lateral static loads on single model piles and 3X3
model pile group and recorded the pile cap deflection and pile bending moments.
Matsumoto (2004) executed a series of static horizontal load tests on a model piled raft in
1-g field. His research considered two variables: the height of the pulling force; and the
rigidity of the pile head connection.

Numerous analytical studies were also performed. Kimura et al. (1995) performed 3D
elasto-plastic finite element analysis to study the ultimate lateral behavior of pile groups
in layered soil. The soil’s constitutive behaviour was modeled using a Druker-Prager
Model and the piles were modeled using nonlinear beam elements. Wakai et al. (1999)
also used 3D elasto-plastic finite element analysis to simulate a number of model tests of
free and rigid head pile groups subjected to lateral loading in homogeneous soil profiles.
Zhang et al. (1999) developed a finite element code that implemented p-y curves (Reese
18
et al., 1974), Brown’s p-multiplier approach (Brown et al., 1998) and the p-multiplier
factors proposed by McVay et al. (1998) to predict the response of laterally loaded pile
groups. Yang and Jeremic (2002) investigated the pile behavior under lateral loads in
layered elasto-plastic soils and demonstrated good agreement between the numerical and
experimental results.

Several researchers investigated the use of geosynthetics with geotechnical structures.

Piles and geogrids were used together as part of a hybrid system, denoted as the geogrid
reinforced and pile-supported (GRPS), to limit total and differential settlements observed
in embankments (Liu et al., 2007). Several researchers investigated the (GRPS) system
(e.g. Maddison et al., 1996; Han and Akins, 2002; Han and Gabr, 2002; and Liu et al.,
2007) and demonstrated its advantages with regard to improving stability of
embankments and reducing their settlements. On the other hand, Bathurst and Alfaro
(1996), Zhenqi and Bathurst (1995), Sandri (1997), EL-Emam and Bathurst (2004)
conducted experimental and numerical studies and showed that the geogrid
reinforcement improved the seismic behaviour of earth retaining structures. This is
attributed to the increased shear resistance due to the friction and interlocking
mechanisms at the soil-reinforcement interfaces. These studies collectively demonstrate
the utility of geosynthetics to improving the static and seismic lateral performance of
geotechnical structures.

In this study, experimental and numerical investigations were conducted to evaluate the
enhancement of the lateral performance of pile foundations reinforced using polymer
strips. The physical model of the polymer strips reinforced pile foundation system
comprised a model pile group that included 4 piles rigidly connected to a rigid pile cap.
The model pile group was installed in a multi-layer soil deposit that included a cohesive
layer underlain by a sand layer. The model soil bed included a granular top layer
representing engineered granular material typically used to replace weak surficial
cohesive soils. The cohesive soil layer was modeled using synthetic clay, Modified
Glyben, which consists of bentonite mixed with water and glycerin. A microgrid mesh
was embedded within the top granular fill layer and connected to the piles to evaluate the

19
static performance of a polymer strips-reinforced pile foundation system. The pile cap
was loaded laterally using a pulling mechanism and fixed on the soil container. The
results of the static tests were used to calibrate/verify finite element model constructed
using commercial finite element software Plaxis 3D. The calibrated/verified model was
used for further parametric analyses in order to investigate various aspects of interaction
between the foundation, soil and polymer strips.

2.2 Background

Reduced scale model tests offer the advantage of studying the response of pile-soil-
superstructure systems in a controlled environment. If they carried out properly, scaled
model tests are also advantageous in seismic studies because they are able to give
economic and realistic information about ground amplification, pore water pressure
variations, non-linear soil behavior and soil structure interaction. Also, reduced scale
model tests offer an economical alternative for full scale tests to run parametric studies
(Turan et al., 2008). Rocha (1957) pioneered developing scale model similitude
relationships for soil mechanics problems. The stress-strain behavior of soils was scaled
assuming that they are linearly proportional in models and prototypes. Rocha’s linear
scaling concept was set to account for the changes in the stress system present in a 1-g
environment. Iai (1989) continued Rocha’s research and derived similitude relationships
assuming that the constitutive stress-strain relationship was independent of the confining
stress if a proper scaling factor is used. Iai (1989) validated this assumption using plane
strain compression tests under varying confining stresses. He derived his similitude
relations based on the geometric and density scaling factors, (λ) and (λP), respectively.
Meymand (1998) successfully used this modeling approach to study soil, pile and
superstructure interaction in a 1-g environment. The present study followed the same
modeling approach. Table 2-1 shows the scaling relations for the variables contributing
to the primary modes of system response.

20
Table 2-1: Scaling relationships for primary system variables (Meymand, 1998).

Parameter Scaling Parameter Scaling Parameter Scaling

factor factor factor

Length λ Acceleration 1 Mass 1

density

Force Shear wave Stress λ

velocity

Stiffness Time Strain 1

Modulus λ Frequency EI

* λ: geometric scaling factor

2.3 Methodology

2.3.1 Model soil

The model soil stratigraphy comprised three layers of soils: i) a layer of silty sand
overlain by, ii) a layer of synthetic clay (modified glyben) to represent the cohesive soil
layer and iii) a surfacial layer of engineered granular backfill. The bottom granular layer
was modeled using the fine granular material with a grain size distribution depicted in
Figure 2-1. The overlying cohesive layer was modeled using Modified Glyben. Modified
Glyben provides favorable characteristics for scaled physical model tests. The primary
advantages of modified glyben for reduced scale model tests are: i) it consolidates at a
very slow rate after application of confining pressure, and thus it can be used in 1-g and
N-g tests without observing a consolidation stage; ii) it resists desiccation due to drying,
and iii) its mechanical properties do not significantly change with time, which facilitates
the multiple use in physical tests (Turan et al. 2009a). The laboratory results showed that
21
modified glyben have a normalized shear modulus versus shear strain amplitude response
that is independent on the confining stress, and glycerin to water ratio (Turan et al. 2008).
The modified glyben was prepared by mixing bentonite with glycerin and water. The
fluid to dry bentonite ratio was 85% and the water to glycerine ratio was 50%. These
ratios were chosen to simulate prototype soft clay. Hand vane shear tests (Pilcon) were
used to measure the undrained shear strength of modified glyben mixture after
preparation and placement in soil container. After a light compaction, one drop of a
sliding hummer, the undrained shear strength of the model clay was about 19 kPa at the
bottom of the clay layer and 8 kPa near the top with a density of 1406 kg/m 3. The
surficial granular layer used to model engineered granular fill had an average particle size
of 5 mm.

120

100
Percent Passing

80

60

40

20

0
0.01 0.1 1 10

Grain Size (mm)

Figure 2-1: Construction sand gradation

2.3.2 Soil parameters

The geotechnical parameters of the three soil layers are summarized in this section. The
strength properties of soils were obtained from direct shear tests and hand vane
apparatus. In the absence of any stiffness measurements, the static soil stiffness was
22
estimated using empirical correlations for granular layers. The static soil stiffness for
granular layers were estimated from the following relations derived by Vermeer and
Schanz (1998) that estimated the Young’s soil modulus based on the confining stress,
i.e.:

Where σx is the confining stress, Pref =100 kPa, E E50.; and shear modulus G =
E/(2(1+) and  is Poisson’s ratio.

The elastic modulus values for the aggregate (surficial layer) and sand (bottom layer)
used in the tests were derived from Equations 1 and 2, i.e.:

Eaggregate=

[3]

Esand= [4]

The elastic modulus values of aggregate and sand were evaluated from Equations 3 and 4
considering the low confinement stresses at the mid depth of each granular layer. For the
cohesive layer comprising of modified glyben, the stiffness was estimated based on the
correlation shown in equation [5] and presented in Turan et al. (2009a). The shear wave
velocity (Vs) of the glyben layer was estimated based on the undrained shear strength
determined using the hand (lab) vane apparatus and (Vs) value was estimated using the
undrained shear strength (cu) of glyben using the empirical equation given below.

Vs = 26.52 Ln (cu) -33 [5]

23
The dynamic shear modulus of the glyben is calculated as:

Gmax =  Vs2 [6]

Where  is soil density = unit weight/gravity acceleration.

The calibrated finite element numerical model showed that some of the clay elements
was deformed by the applied lateral load up to a shear strain level amounted to 0.1 %.
This level of shear strain corresponds to 60 % G/Gmax taken from the glyben degradation
curve (Turan et al. 2009a). Therefore, the static shear modulus of glyben, Gstatic, was
assumed to be 60 % of the dynamic shear modulus obtained from an average clay Vs of
36 m/s. Also, the measured undrained shear strength (using lab shear vane) of glyben
near the backfill interface was less than that at the bottom by around 56 % due to the
overburden stress buildup. Therefore, a layer of weak clay was induced near the backfill
surface, which has been replaced with strong granular layer when the depth of backfill
was increased. The structure-glyben interface strength was assumed in the analysis to be
10 % of the glyben strength due to using a smooth cap and acrylic piles. Table 2.2 shows
the soil parameters used in the numerical analysis.

Table 2-2 : Soil parameters

Aggregate Glyben Sand

Unit weight, (kN/m3) 16.1 13.8 17.35

C (kPa) 0 10 0

peakdegrees) 50 0 40

24
(Appendix A)

Gstatic (kPa) 1194 1093 3222

Poisson’s ratio,  0.15 0.2 0.3

The model foundation system

The reduced scale model foundation comprised four 600 mm long and 19 mm diameter
acrylic tubes. These model piles were rigidly connected to an aluminum cap with 200
mm X 200 mm plan dimensions and 6 mm thickness which correspond to a concrete cap
of 0.7 m thickness. Four plastic rings were fabricated and clamped on the piles to
facilitate the load transfer between the microgrid and the piles. The microgrid mesh was
extended and pressed inside two halves of rings using four steel bolts. The rings were
fixed in their vertical position using a plastic bolt. Figure 2-2 shows the details of the
foundation model. Table 2-3 provides a summary of the scaled model corresponding
prototype dimensions evaluated using the scaling relationships presented in Table 2-1.

25
Figure 2-2: The scaled piled cap geogrid model

Table 2-3: Model Piles calculations, λ = 20.

Parameter Model pile (Acrylic) Calculated Prototype pile

(Steel)

Outer diameter (mm) 19 380

Length (mm) 600 12000

Young's modulus (kPa) 3.2 X 106 200 X 106

Flexural rigidity, EpI (kN.m2) 1.82 X 10-2 5.81 X 104

Axial rigidity, EpA (kN) 6.03 X 105 3.5 X 106

26
2.3.3 The Model geogrid

The microgrid mesh was introduced within the aggregate layer at a depth of 3.5 times the
pile diameter (i.e 66 mm). The geosynthetic mesh used in the model tests was a bi-axial
knitted Microgrid manufactured by STRATAGRID. STRATAGRID Microgrid is
manufactured utilizing a complex knitting process and polymer coatings to provide
desired engineering properties and constructed of high molecular weight and high
tenacity polyester yarns. The yarns provide significant tensile capacity through precision
knitting into a dimensionally firm, uniform network of apertures. The model stiffness
value of the microgrid at 2% was Jm = 110 kN/m. With a geometrical scale factor λ = 20,
the prototype stiffness Jp= Jm λ2 = 110 x 202 = 44000 kN/m (El Emam and Bathurst,
2004). Table 2-4 shows the main engineering properties of the STRATAGRID
microgrid. STRATAGRID microgrid was considered as model size polymer strips in
this study.

Table 2-4: Engineering properties of Microgrid geogrid

Index Properties Test Method Value

Ultimate Strength ASTM D-6637 Method A 29.2 kN/m

Creep Limited Strength ASTM D-5262/D-6992 16.8 kN/m

Strength @ 2% Strain ASTM D4595 2.2 kN/m

Strength @ 5% Strain ASTM D4595 (MD) 7.3 kN/m

Strength @ 5% Strain ASTM ASTM D4595 (CMD) 4.4 kN/m

Aperture Size Measured 2.54 x 6.35 mm

27
2.3.4 Model Stratigraphy

The testing program was commenced by assembling the soil container (Turan et al.
2009b). A wood board covered with fine steel grits was then bolted to the bottom of soil
container. The steel grits was glued to the wood board using epoxy and installed on the
bottom of container. Prior to soil placement, the sides of the soil container were covered
with latex membrane to prevent soil particles from migrating out of the soil container
through the openings between the lamina. The initial granular layer of the model soil was
placed inside the container in lifts of 20 mm. Each lift of sand was compacted by tamping
until an 80 % relative density had been achieved. Subsequently, the pile-cap system was
installed. To eliminate excessive settlement during the testing program, the lumped mass
representing the superstructure was placed and maintained on the cap for 24 hours to
expedite the piles immediate settlement. Subsequently, sand backfilling proceeded in lifts
until a total layer thickness of 36 cm was reached.

Blocks of soft modified glyben were prepared by pressing weighed quantities into a small
box with dimensions of 70x70x50 mm. These blocks were then placed over the sand
layer, positioned side by side and compacted using a drop hammer until the target total
layer thickness was reached. Each layer of clay had 40 mm thickness. Figure 3 depicts
the compaction process of modified glyben. Subsequently, a granular layer was placed on
top of the modified glyben layer to simulate a situation where the soft cohesive layer is
over-excavated and replaced with engineered granular fill. The details of placement of
this layer are presented in the following subsection. In field applications, when soft
cohesive soils exist at or near the ground surface, the soft clay is typically excavated and
replaced with engineered backfill.

28
 Figure 2-3: Compaction of modified glyben

2.3.5 Placement of geogrid-reinforced engineered granular fill

The granular layer was placed on top of cohesive soil layer to represent typical
excavation and replacement with engineered granular fill. This is a common practice
where shallow layers of compressible soils are encountered. Once the granular soil was
placed in lifts, the microgrid mesh was laid within the granular layer at the specified
depth. The mesh was connected to four model piles using the clamps (rings). To measure
the global microgrid extension during the static pull test, two thin wires were hocked on
two 50 mm diameter smooth metal discs extending out of the soil container towards two
wire position sensors. The discs were glued at the surface of the microgrid and aligned at
the two centerlines of the 2x2 pile group. The two discs were attached at 55 and 300 mm

29
from the face of the microgrid ring to examine the strain behavior of the microgrid mesh
near and far from the pile face (see Figure 2-4).

Figure 2-4: Geogrid extending between the ring and the metal disc

2.3.6 Loading System and Instrumentation

The horizontal static load was applied to the pile cap directly through a tension wire
connected to a pulling mechanism which was fixed at the side of soil container. The
pulling mechanism comprised of a hand-operated winch and rollers that allow the cable
to run smoothly and horizontally to the pile cap top (Figure 2-5 a and b).

30
(a)

31
 (b)

Figure 2-5: a. Schematic diagram of the testing setup, b. top view of the testing
setup.

The pull load applied to the pile cap was measured using an OMEGA 500 N load cell
connected to the pulling wire. The pile cap horizontal translation was measured using a
25 mm linear variable displacement transducer (LVDT) made by Hoskin Scientific Ltd.
The LVDT sensor has a measuring range of 25 mm with a sensitivity of 7.3/7.5 mv/v.
Figures 5a and b show the LVDT and the load cell, which were used to record
displacements and loads while the pile cap was being pulled slowly. The global strain of
the microgrid mesh was measured by two 100 mm range Penny & Giles Draw Wire
Position Sensors (Figure 2-6). The Wire Position sensors has a resolution of ± 0.07 mm.
The clamps of the wire sensors were tied to two thin wires extending out of the soil
container towards the two 50 mm diameter smooth metal discs. The position sensors
were mounted on the frame supporting the soil container. The output signals from LVDT,
load cell and position sensors were recorded at a computer station that housed the data
acquisition system panels and data logging software.
32
 Figure 2-6: The Wire Line Position Sensors mounted on a frame and tied to the
wires

2.3.7 Numerical model

The numerical model was established using the commercial finite element software
Plaxis 3D. The numerical model comprised ten-node tetrahedron elements to simulate the
soil volume, six-node plate elements to simulate the pile cap, shell elements to simulate
the microgrid mesh and three-node beam elements to simulate piles. The rings that
connected the piles to the microgrid mesh were simulated using larger pile tube elements.
Interface elements were used at pile-soil interfaces. Figure 2-7 depicts the three
dimensional view of the numerical model.

33
The pile-soil interfaces were modeled as embedded interface elements having their
strength equivalent to a percentage of the adjacent soil shear strength, denoted as
Rinterface. The pile-soil interface elements are three nodes elements connecting both
sides of pile nodes and soil nodes within the soil’s tetrahedron elements. These interface
surfaces allow relative movement between piles and soil elements (Brinkgreve et al.,
2012). Within the backfill layer, the Pile-Soil Rinterface strength is assumed to be 90 % of
the adjacent soil strength for the total depth of 8 cm to account for minor slippage
associated with the relatively small static loading. . In order to simulate the
Microgrid/soil interlock at the geogrid mesh level, the geogrid-soil element deformation
is assumed continuous along the common edge between geogrid and soil elements.

The constitutive behavior of the soil and interface was modeled using elastic-perfectly
plastic stress-strain relationship and Mohr-Coulomb failure criterion. Around the soil
volume, general boundary conditions were automatically imposed around the soil volume
according to the following rules (Plaxis manual, 2011).

 Vertical model boundaries with their normal in x-direction (i.e. parallel to the
yz-plane) are fixed in x-direction (Ux =0) and free in y- and z-direction.
 Vertical model boundaries with their normal in y-direction (i.e parallel to the
xz-plane) are fixed in y-direction (Uy =0) and free in x- and z-direction.
 Vertical model boundaries with their normal neither in x- nor in y-direction are
fixed in x- and y- direction (Uy =Ux=0) and free in z-direction.
 The model bottom boundary is translational fixed in all directions
(Ux=Uy=Uz=0).
 The ground surface is translational free in all directions.

Plaxis allows users to execute the finite element model in phases. The numerical
calculations were executed in 18 phases to simulate the geostatic equilibrium,
construction sequence as well as incremental application of lateral loads. The pull load
was applied on a node at the center of the model pile cap in increments starting from 100
N and increased by 100 N increments at each loading stage until a maximum load of 800
34
N was reached. Other numerical models that included a microgrid-reinforced granular
layer with different microgrid embedment depths were developed and analyzed. The
results of these analyses were compared with the results of the base case where no
microgrid reinforcement was present.

(a)

(b)

Figure 2-7: a) Plaxis 3D numerical model of Soil and pile-cap-geogrid system

b) Finite element mesh.

35
2.3.8 Experimental and numerical test cases
Table 2-5 shows the summary of the experimental and numerical test cases. As shown in
Table 2-5, three physical modeling cases and their corresponding numerical models were
considered and analyzed: (i) foundation with thin and thick engineered backfill with
microgrid; (ii) foundation with thin and thick engineered backfill without microgrid; and
(iii) the base case of foundation in soft clay without backfill and without microgrid. Table
2-6 provides the thickness of each of the three layers considered in the different test
cases.

Table 2-5: Summary of experimental and numerical cases

Test No. Test case description

1 Static pull test , foundation in deep backfill with Microgrid

2 Static pull test , foundation in deep backfill without Microgrid

3 Static pull test , foundation in shallow backfill without Microgrid

4 Plaxis model of static pull, foundation in deep backfill with Microgrid.

5 Plaxis model of static pull, foundation in deep backfill without Microgrid.

6 Plaxis model of static pull, foundation in shallow backfill without Microgrid.

7 Plaxis model of static pull, foundation in shallow backfill with Microgrid.

8 Plaxis model of static pull, foundation in soft clay without backfill (base case).

36
Table 2-6 : Thickness of different soil layers considered in different test cases

Test Case No Surficial backfill layer Soft clay layer Bottom sand layer

Thickness (m) Thickness (m) Thickness (m)

1,2,4,5 0.08 0.23 0.36

3,6,7 0.03 0.28 0.36

8 0 0.31 0.36

2.4 Results and discussion

The results are presented in terms of the load-displacement response of the model
foundation, shear forces and bending moments of the piles as well as strain in the
microgrid material. The results from the different cases are compared to evaluate the
effects of using the microgrid material on the foundation response and the possibility for
reducing the requirements for engineered backfill.

2.4.1 Pile cap translation

The measured load-displacement response curves of the model foundations were
compared with the responses simulated using the numerical models and results were
depicted in Figures 2.8 and 2.9. Figure 2.8 shows the experimental static pull test results
and the numerical simulation of the static test for the deep backfill case with microgrid
(Cases 1 and 4). Figure 2.9 shows the experimental and numerical static pull test results
for the deep backfill case without the embedment of the microgrid mesh (Cases 2 and 5).
The experimental and numerical results shown in Figures 2.8 and 2.9 demonstrate good
agreement between the two sets. Comparing the results in Figures 2.8 and 2.9 indicates
that the addition of the microgrid mesh increased the lateral stiffness of the foundation
and reduced its lateral displacement. For example, a 3 mm displacement was caused by
700 N loads in Case 2, where no microgrid reinforcement was used, while the same
37
amount of displacement was caused by 800 N when the microgrid reinforcement was
applied (Case 1). This indicates about 15 % improvement in lateral resistance.

900
Static test _Deep Backfill with Microgrid
800 Plaxis Result_Deep Backfill with
Microgrid
700 Static Result trend line

600
Pull load (N)

500

400

300

200

100

0
0 0.5 1 1.5 2 2.5 3 3.5
Horizontal displacement (mm)

Figure 2-8: Lateral load-displacement response of model foundation and results of

numerical simulations (Cases #1 and 4).

38
 800
Static test_ Deep Backfill Without Microgrid

700 Plaxis Result_ Deep Backfill Without

Microgrid

600
Pull load (N)

500

400

300

200

100

0
0 0.5 1 1.5 2 2.5 3 3.5 4
Horizontal displacment (mm)

Figure 2-9: Lateral load-displacement response of model foundation and results of

numerical simulations (Cases #2 and 5).

39
Figure 2-10 compares the experimental and numerical results for Cases 3 and 6, where no
microgrid was used and a shallow layer of aggregate backfill was considered. Figure 2-
10 indicates a good agreement between the experimental and numerical result. Also, the
comparison between Case 3 (thin aggregate backfill layer) and Cases 2 (thick aggregate
backfill layer) revealed that reducing the backfill layer reduced the lateral stiffness of the
pile group and resulted in larger lateral displacement of the pile foundation.

700
Static test
600

500

400
Pull load (N)

300

200

100

0
0 0.5 1 1.5 2 2.5 3 3.5
Horizontal displacment (mm)

Figure 2-10: Lateral load-displacement response of model foundation and results of

numerical simulations (Cases # 3 and 6)

Figure 2-11 compares the experimental Cases 1 to 3. The results indicate that the pile cap
foundation resistance to lateral deformation improved by 15 % at 3 mm displacement due
to the addition of microgrid. Also, the data trend shows that the resistance of the pile cap
foundation system to lateral loading was degraded by 20 % due to reducing the thickness
of granular backfill to half.
40
 900
Static test _Deep Backfill with Microgrid

Static test_ Deep Backfill Without Microgrid

800 Static test_Shallow backfill without Microgrid

Static test_Deep Backfill With Microgrid Trend Line

700 Static test_Beep Backfill Without Microgrid Trend Line

600

500
Pull load (N)

400

300

200

100

0
0 0.5 1 1.5 2 2.5 3 3.5

Horizontal displacement (mm)

Figure 2-11: experimental results comparison (cases 1-3)

Figure 2-12 compares the results of the numerical analysis for cases 5 to 8. It can be
noted from Figures 2-12 that the pile foundation performance has been improved due to
increasing the granular backfill. Also, it shows the superiority of the polymer strips
solution as the shallow backfill with microgrid resulted in the least lateral displacement.
The results show that using the microgrid reinforcement can allow smaller backfill and
yet improved response. This can represent an effective solution and significant saving in

41
cost and construction time in sites where deep backfill be required following
conventional ground replacement design.

900

800

700

600

500
Pull load (N)

400
Plaxis Result_ Deep Backfill Without
300 Microgrid
Plaxis Result_ No Backfill Without Microgrid
200 Plaxis Result_ Shallow Backfill Without
Microgrid
100 Plaxis Result_ Shallow Backfill With
Microgrid
0
0 0.5 1 1.5 2 2.5 3 3.5 4 4.5
Horizontal displacment (mm)

Figure 2-12: Numerical results comparisons (cases 5-8)

2.4.2 Extension of microgrid (in Deep backfill case)

Figure 2-13 shows the measured microgrid extension near the piles during the static pull
test. The microgrid lateral deflection, at a distance of 56 mm from the pile ring was
subtracted from the pile deflection at the microgrid level to obtain the magnitude of
microgrid extension. This value was divided by the microgrid length of interest to obtain
the microgrid strain. Figure 2-13 shows the stiffness behavior of the microgrid and
indicates an increasing stiffness as a result of increase in deflection (i.e. hyper-elastic
stiffness). Figure 2-14 shows the variation of foundation lateral resistance with the
microgrid strain far from the piles. It can be noticed from Figure 2-14 that the microgrid
strain increases as the pile cap pull load increases. This behavior indicates stiffening in
the lateral behavior of the foundation system due to the hyper-elastic behavior of the
42
microgrid mesh. Thus, it can be expected that with increasing lateral loads, the efficiency
of the polymer strips reinforcement will increase. Comparing the results in Figure 2-13
shows that the microgrid strain near the pile (only the first 5 mm) was in the range of 5 to
25%. However, Figure 2-14 shows that the rest of the microgrid mesh experienced strain
in the range of 0.5 to 2%. It may be inferred that the average strain for the microgrid
mesh up to the maximum load applied was about 2 to 2.5%. Therefore, the strength and
stiffness parameters of the microgrid used in the numerical models would be evaluated at
2% strain.

900
800
700
Pile cap pull load (N)

600
500
400
300
200
100
0
0 5 10 15 20 25 30
Microgrid Strain (%)

Figure 2-13: Microgrid strain vs. pile cap pull load measured near to the piles face.

43
 900

800

700
Pile cap pull load (N)

600

500

400

300

200

100

0
0 0.5 1 1.5 2 2.5
Microgrid strain (%)

Figure 2-14: Microgrid strain vs. pile cap pull load measured far from the piles face

Figure 2-15 displays the microgrid extension near and far from the piles measured at
WIRE1 and WIRE2, respectively. The results show that in the initial loading phase, the
microgrid reinforcement was not engaged. However, as the lateral load increased, the
microgrid reinforcement was engaged due to the interlocking of the granular material
within the microgrid, which increased the shearing resistance of the aggregate in the
vicinity of the microgrid. It can also be noticed that the microgrid (and the interlocked
aggregates) stiffness increased as the lateral load increased due to further mobilization of
the interlocking mechanism. In addition, Figure 15 indicates that the extension that the
polymer strips experiences under a given lateral load dissipated along the microgrid mesh
away from the foundation.

44
 1000
900
Pile Cap Pull Load (N) 800
700 WIRE1
600 WIRE2
500
400
300
200
100
0
0 0.5 1 1.5 2 2.5
Microgrid Deformation (mm)

Figure 2-15: Global Microgrid deformation measured near and far from the pile
face

2.4.3 Results of Numerical Analysis

The results of the numerical parametric study that was conducted to investigate various
aspects of the mechanical interaction between the foundation, soil and microgrid
reinforcement are presented in this section.

Bending Moment and Shear Forces and in Piles

The piles section-forces and moments calculated employing the numerical models that
were calibrated using the results of the physical tests are presented herein. Figure 2-16
demonstrates the effect of microgrid reinforcement on the maximum bending moment of
the piles as the static lateral load increased. It can be noted that the bending moment
decreased due to the addition of the microgrid mesh by approximately 8%. However, the
reduction in shear force was marginal (only 2%) compared to the reduction in the
bending moment as noted from Figure 2-17.

45
 12.00
B-M Without Microgrid
10.00
Max Bending Moment (Nm)

B-M With Microgrid

8.00

6.00

4.00

2.00

0.00
0 200 400 600 800 1000

Pile Cap Pull Load (N)

Figure 2-16: Bending moment at pile head vs. pile cap lateral deflection

250.00
Shear force - Without
geogrid
Shear force - With geogrid
200.00
Max Shear Force (N)

150.00

100.00

50.00

0.00
0 100 200 300 400 500 600 700 800 900
Pile Cap Pull Load (N)

Figure 2-17: Shear force at pile head vs. pile cap lateral deflection
46
Effect of Microgrid Length

Figure 2-18 compares the pile cap deflections calculated for two microgrid lengths. The
microgrid lengths considered in the parametric analyses were, 90 cm and 20 m. The
results shows that increasing the microgrid mesh length reduced the lateral pile cap
deflection by around 12%. Moreover, Figures 2-19 and 2-20 demonstrate the effect of
microgrid length on the maximum bending moment and shear forces on the piles. It can
be noted from Figures 2-19 and 2-20 that as the microgrid mesh extent from the pile
foundation increased, the bending moment decreased by around 5%. Also, it can be
noticed that this reduction in shear force was marginal compared to the reduction in the
bending moment.

900
90 cm Microgrid
800
20 cm Microgrid
700
Pile Cap Pull Load (N)

600

500

400

300

200

100

0
0 0.5 1 1.5 2 2.5 3 3.5 4

Pile cap lateral defliction (mm)

Figure 2-18: Pile cap deflections for microgrid lengths of 90 cm and 20 cm.

47
 12
20 cm Microgrid
Max Bending Moment (N) 10 90 cm Microgrid

0
0 200 400 600 800 1000

Pile Cap Pull Load (N)

Figure 2-19: Maximum bending moments for microgrid lengths of 90 cm and 20 cm.

48
 250
20 cm Microgrid

200 90 cm Microgrid
Max Shear Force (N)

150

100

50

0
0 200 400 600 800 1000

Pile Cap Pull load (mm)

Figure 2-20: Maximum shear forces for microgrid lengths of 90 cm and 20 cm

2.5 Summary and Conclusions

This chapter presented static model tests and numerical simulations to investigate the
lateral behavior of a model polymer strip reinforced pile foundation system. The
following conclusions were derived from the results of reduced scale physical tests and
corresponding numerical analyses:

 The lateral resistance of the foundation system was increased by 15 % due to the
addition of the model microgrid mesh and extending it near the boundary of the
soil container.
 The peak lateral resistance of the foundation system decreased by 20 % due to
reducing the thickness of the surficial granular backfill layer.

49
 The numerical results show that the addition of the polymer strips would allow
reducing the required soft ground replacement by 50%, while providing improved
lateral performance.
 The numerical results showed that the addition of the geosynthetic reinforcement
reduced the bending moment by 8% and reduced the shear force by 2%.
 The numerical results also showed that extending the geosynthetic reinforcement
increased the lateral stiffness of the foundation system and reduced the lateral
displacement by 12 %.
 The numerical results indicated that extending the microgrid mesh farther from
the pile foundation bending decreased the bending moment by around 5%.
 The parametric study indicated that the extent of the geosynthetic reinforcement
influences the level of enhancement it provides. Thus, the length of the
reinforcement should be optimized for the specific case considered.
 Overall, the model scale experimental and numerical results showed that the
beneficial effects of the geosynthetic reinforcement increased as the applied load
increased. Thus, further improvement of the lateral performance of the
geosynthetic-reinforced foundation is expected if the foundation is allowed to
experience larger displacement.

50
References:
[1] Bathurst, R.J. and Alfaro, M.C. 1996. Review of seismic design, analysis and
performance of geosynthetic rein- forced walls, slopes and embankments.
International Symposium on Earth Reinforcement, keynote lecture, Kyushu, 23-52.

[2] Brown, D. A., Morrison, C., and Reese, L. C. 1 988. Lateral load behavior of pile
group in sand. Journal of Geotechnical. Engineering. ASCE, 114 (11), 1261–1276.

[3] Cai, Z., and Bathurst, R.J. 1995. Seismic response analysis of geosynthetic
reinforced soil segmental retaining walls by finite element method. Computers and
Geotechnics, 17 (4), 523-546.

[4] El-Emam, M. M., and Bathurst, R.J. 2007 .Influence of reinforcement parameters
on the seismic response of reduced-scale reinforced soil retaining walls.
Geotextiles and Geomembranes, 25 (1), 33-49.

[5] El-Emam, M. M., and Bathurst, R.J. 2004. Experimental design, instrumentation
and interpretation of reinforced soil wall response using a shaking table.
International Journal of Physical Modelling in Geotechnics, 4, 13-32.

[6] Han, J., and Akins, K. 2002. Use of geogrid-reinforced and pile supported earth
structures. Proceedings. Deep Foundations 2002: An International Perspective on
Theory, Design, Construction, and Performance, ASCE, Reston, Va.,pp. 668–679.

[7] Han, J., and Gab, M. A. 2002. Numerical analysis of geosynthetic reinforced and
pile-supported earth platforms over soft soil. Journal of Geotechnical and
Geoenvironmental. Engineering, 128 (1) , 44–53.

[8] Meymand, P. J. 1998. Shaking table scale model tests of nonlinear soil-pile-
superstructure interaction in soft clay. Ph.D. Dissertation, University. of California,
Berkeley, CA.

[9] Iai, S. 1989. Similitude for shaking table tests on soil-structure-fluid models in 1g
gravitational field. Soils and Foundations, 29 (1), 105–118.
51
[10] Kimura, M., Adachi, T., Kamei, H., and Zhang, F. 1995. 3-D Finite element
analyses of the ultimate behavior of laterally loaded cast-in-place concrete piles. In
G. N. Pande and S. Pietruszczak, editors, Proceedings of the Fifth International
Symposium on Numerical Models in Geomechanics, NUMOG V,. A. A. Balkema,
September 1995. pp. 589-594.

[11] Maddison, J. D., Jones, D. B., Bell, A. L., and Jenner, C. G. 1996. Design and
performance of an embankment supported using low strength geogrids and vibro
concrete columns. Geosynthetics: Applications, design and construction, De Groot,
Den Hoedt, and Termaat, eds., Balkema, Rotterdam, The Netherlands, 325–332.

[12] Matsumoto, T., Fukumura, K., Pastsakorn, K., and Horikoshi, K. 2004.
Experimental and analytical study on behavior of model piled rafts in sand
subjected to horizontal and moment loading. International Journal of Physical
Modeling in Geotechnics, 4 (3), 2042-6550.

[13] McVay, M. C., Zhang, L. M., Molnit, T., and Lai, P. 1998. Centrifuge testing of
large laterally loaded pile groups in sands. Journal of Geotechnical and
Geoenvironmental Engineering, ASCE, 124 (10), 1016–1026.

[14] Reese, L. C., Cox, W. R., and Koop, F. D. 1974. Analysis of laterally loaded piles
in sand. Proceedings. 6th Annual. Offshore Technology. Conference. Offshore
Technol. Conference., Richardson, Tex., 473–483.

[15] Rocha, M. 1957. The possibility of solving soil mechanics problems by the use of
models, Proceedingd.4th Int. Conference. Soil Mechanical. Foundation.
Engineering, 1, 183-188.

[16] Schanz, T. and Vermeer, P. A. 1998. On the stiffness of Sands, Getechnique, 48,
383-387.

52
[17] Sandri, D. 1997. A performance summary of reinforced soil structures in the
greater Los Angeles area after the Northridge earthquake, Geotextiles and
Geomembranes, 15 (4-6), pp. 235-253.

[18] Tatsuoka, F., Sakamoto, M., Kawamura, T., and Fukushima, S. 1986. Strength and
deformation characteristics of sand in plane strain compression at extremely low
pressures, soils and foundations, 26 (1), 65-84.

[19] Turan, A., Hinchberger, S.D., and El Naggar, M.H. 2009a .Mechanical
characterization of an artificial clay. Journal of Geotechnical and
Geoenvironmental Engineering, 135 (2), 280-290.

[20] Turan, A., Hinchberger, S.D., and El Naggar, M.H. 2009b. Design and
commissioning of a laminar soil container for use on small shaking tables. Soil
Dynamics and Earthquake Engineering, 29 (2), 404-414.

[21] Turan, A., Hinchberger, S.D., and El Naggar, M.H. 2008 Investigation of dynamic
performance of bentonite-glycerin-water based artifitial caly. GRC Research
Report No: GEOT-08-01, The University of Western Ontario, London, ON,
Canada.

[22] Wakai, A., Gose, S., and Ugai, K. 1999. 3-D Elasto-Plastic finite element analysis
of pile foundations subjected to lateral loading. Soil and Foundations, 39 (1), 97-
111.

[23] Yang, Z and Jeremic, B. 2005. Study of soil layering effects on lateral loading
behavior of piles. Journal of Geotechnical and Geoenvironmental Engineering,
ASCE, 131(6), 1090-0241.

[24] Zhang, M. M. Limin and Lai, P. 1999. Numerical analysis of laterally loaded 3x3
to 7x3 pile groups in sands. Journal of Geotechnical and Geoenvironmental
Engineering, 125 (11), 936-946.

53
 Chapter 3
3. EXPIRMENTAL STUDY ON THE DYNAMIC LATERAL
BEHAVIOR OF GEOGRID REINFORCED PILE
FOUNDATION SYSTEM

The use of geosynthetic reinforcement in earth structures have increased significantly

over the last two decades. The advantages associated with the use of geosynthetics have
been well documented. Some of the typical applications include slopes reinforcement,
improvement of embankments foundations, mechanically stabilized retaining structures
and subgrade improvement for roads.

Certain subsurface conditions may dictate special foundation solutions such as piled
foundation systems. Such foundations are widely used in seismically active areas, where
they are expected to resist significant lateral loads. However, the weak subsurface
conditions that dictate the use of pile foundation systems result in low lateral foundation
resistances. This chapter introduces an innovative use of geosynthetics as part of a novel
foundation concept called geosynthetics-reinforced pile foundation, where polymer strips
are used to enhance the lateral resistance of the pile foundation system.

The seismic pile-soil-geosynthetics interaction of this system was evaluated using a

series of reduced scale physical model tests performed on a shaking table in 1G
environment. A uni-directional laminar shear box containing a three layer soil
stratigraphy, which included a layer of synthetic clay known as modified glyben (Turan
et al. 2009a) sandwiched between lower and upper layers of granular materials, was used
in the physical model tests. The model pile-cap system that supported a single degree of
freedom structure was installed and a series of tests were performed using dynamic
loadings in the form of sine sweep, harmonic and scaled earthquake signals in order to
identify the amplification and resonance conditions of the foundation system and to
evaluate various aspects of the pile-soil-geosynthetic interaction.

54
3.1 Introduction

Poor ground conditions near the surface pose some foundation challenges for design
engineers. In many cases, deep foundation systems (piles) are sought as the practical
foundation option to address the challenges of weak foundation soils. Pile foundations
are widely used in seismically active areas, where they are expected to resist significant
lateral loads. However, the weak subsurface conditions near the ground surface often
dictate expensive and time consuming ground replacement, pile inclination or soil
stabilization practices in order to enhance the lateral resistance of the pile foundations.

This chapter introduces the results of reduced scale physical model tests of a novel
geosynthetics-reinforced pile foundation system. The physical model of the pile
foundation system was composed of 4 model tube piles connected with a rigid steel cap.
The steel cap is connected to a model single degree of freedom structure (SDOF). The
SDOF was designed and fabricated to examine the effect of model reinforcing polymer
strips (microgrid mesh) on the seismic response of structures with different natural
periods. The foundation system was installed in a multi-layer soil deposit that comprised
a soft cohesive layer underlain by a sand layer. The model soil stratigraphy in the laminar
container included a granular top layer representing the engineered granular material,
which is often used to replace the weak surficial cohesive soil. The cohesive soil layer
was modeled using the Modified Glyben (Turan et al. 2009a). A microgrid mesh was
embedded within the top granular fill layer and was connected to the piles. The dynamic
performance of the pile-cap-microgrid system has been studied using shaking table tests
in 1g environment. Various forms of dynamic loads such as harmonic loads, sine sweep
and scaled earthquake signals have been used to assess the performance of the microgrid
reinforcement to improve lateral foundation response.

55
3.2 Background

Geosynthetics are polymeric materials that consist of tensile ribs with openings of
sufficient size to allow interlocking with the surrounding soils. Examples of practical use
of geosynthetics include geogrids, geonets, cellular confinement and polymer strips. The
geosynthetics-soil interaction mechanism allows the geosynthetics to work as a
reinforcing element, which provides the soil with tensile strength and enhances its shear
resistance. Therefore, geosynthetics have been widely used in modern construction
practices. Several researchers have investigated the use of geosynthetic-reinforced soils
to enhance the resistance of the geotechnical structures. Studies such as Bathurst and
Alfaro (1996), Zhenqi and Bathurst (1995), Helwany et al. (2001), El-Emam (2007) and
Liu (2009) investigated the use of geogrid reinforcement to enhance the stability of earth
retaining walls. Others investigated the performance of geosynthetics as part of
geosynthetics reinforced embankment support systems (e.g. Liu et al., 2007).

The use of polymer strips reinforcement to improve the lateral resistance of pile
foundation systems under dynamic loading conditions was studied in this chapter using a
reduced scale physical modeling approach in a 1g environment. Despite the limitations of
the 1g shaking table tests with regard to achieving the suitable stress fields for testing
cohesionless soils, these tests have been used extensively to examine the dynamic
behaviour of polymer strips reinforced earth retaining structures and the seismic soil-pile-
superstructure interaction problem. Sakagutchi et al. (1992) conducted a set of reduced
scale shaking table tests on a modular block soil wall. Matsumo et al. (1998) investigated
the influence of increasing the length of soil reinforcement through a set of shaking table
tests on six model walls. Koseki et al. (1998) investigated the dynamic lateral behaviour
of a conventional and reinforced soil model walls through a series of shaking table tests.
El Emam and Bathrust (2007) carried out comprehensive shaking table tests to examine
the influence of geogrid reinforcement parameters such as stiffness, length and vertical
spacing on the seismic response of reduced scale soil retaining wall.

56
Extensive research was conducted to investigate the seismic soil-pile-superstructure
interaction (SSPSI) problem using 1g shaking table tests. Kubo (1969) was the first to
perform shaking table model pile tests. Gohl (1991), Liu and Chen (1991), Yan et al.
(1991), Sreerama (1993), Markis et al. (1997), Tao et al. (1998) are some examples of
such studies. Most of these studies have investigated the seismic response of single piles
and small pile groups in cohesionless soil. Although many tests have shown deviations
from a specific model similitude rule, few scale model tests have been calibrated with
prototype tests and been proven successful.

The use of geosynthetics to improve the seismic performance of piled foundations and to
reduce intolerable pile lateral deflections is a new concept. Thus, this research is carried
out to investigate the dynamic lateral performance of geosynthetics-reinforced pile
foundation system using reduced scale shaking table tests.

3.3 Methodology

This section describes the testing facility, provides a brief description of the model
testing laminar container, and summarizes the similarity rules used to develop the scaled
models. This section also gives a detailed description of the model soils, model
foundation system and the single degree of freedom system that represents the
superstructures. The instrumentation and data acquisition systems as well as the testing
plan and procedures followed in this study are explained.

3.3.1 Reduced scale physical model

Reduced scale model tests offer the advantage of studying the response of pile-soil-
superstructure systems in a controlled environment. If they carried out properly, reduced
scale model tests are also advantageous in seismic studies because they are able to give
economic and realistic information about ground amplification, pore water pressure
variations, non-linear soil behavior and soil structure interaction. Also, reduced scale
model tests offer an economical alternative for full scale tests to run parametric studies
(Turan et al., 2008). In this research, several reduced scaled model tests were carried out
57
on a shaking table. The main feature of shaking table tests is that they are executed in in a
1-g environment which does not induce the elevated confining stress field required for
proper modeling of cohesionless soils ,as in centrifuge tests (Meymand, 1998).

Rocha (1957) started developing scale model similitude relationships for soil mechanics
problems. Rocha scaled the stress-strain behavior of soils assuming that they are linearly
proportional in models and prototypes. Rocha’s linear scaling concept was set to account
for the changes in the stress system present in a 1g environment. Iai (1989) continued
Rocha’s research and derived similitude relationships assuming that the constitutive
stress-strain relationship is independent on the confining stress if a proper scaling factor
is used. He validated this assumption using plane strain compression tests under varying
confining stresses (Tatsuoka et al., 1986). The tests results showed that this assumption is
valid for axial strains lower than the strains at failure peak stress. Iai (1989) derived his
similitude relations based on the geometric and density scaling factors, (λ) and (λP),
respectively. Meymand (1998) developed a modeling approach suitable for soil, pile and
superstructure conditions. In his method, the primary modes of system response are
defined first. Then, prototype values for the variables contributing to these modes are
selected. Similarity rules are then used to verify that the calculated implied prototype
response is reasonable. Table 3-1 shows the scaling relationships for the variables
contributing to the primary modes of system response.

58
Table 3-1: Scaling relationships for primary system variables (Meymand, 1998).

Parameter Scaling Parameter Scaling Parameter Scaling

factor factor factor

Length λ Acceleration 1 Mass 1

density

Force Shear wave Stress λ

velocity

Stiffness Time Strain 1

Modulus λ Frequency EI

* λ: geometric scaling factor

3.3.2 The shaking table and soil container

The model soil bed was formed in a laminar container placed on a 1.22 m x 1.22 m 1-D
shaking table. The laminar container comprises of 24 rectangular shaped laminae stacked
to create an 807 mm high, 900 mm long and 450 mm wide box with 2 mm clearance
between laminae. The laminae are machined using solid high strength aluminum alloy
section. Figure 1 shows the details of laminar container used in this study (Turan et al.,
2009b). The shaking table was equipped with pneumatic and electrical actuators
controlled by a digital control module that allows the generation of various types of
signals. The shaking table used in this study is a one-dimensional table that can generate
various types of dynamic loads, such as harmonic, sine sweep and white noise signals.

59
Figure 3-1: Laminar soil container covered with latex membrane and aligned with
corner profiles

3.3.3 The model soil column

The model soil stratigraphy comprised three layers of soils starting from bottom to top:
i) a layer of silty sand; ii) a layer of synthetic clay (modified glyben) to represent the
cohesive soil layer; and iii) a surfacial layer of engineered granular backfill. The bottom
granular layer was modeled using the fine granular material with grain size distribution
depicted in Figure 3-2. The peak angle of internal friction of the sand was measured as
40 from direct shear tests.

60
 120

Percent Passing 100

80

60

40

20

0
0.01 0.1 1 10
Grain Size (mm)

Figure 3-2: Construction sand gradation

The overlying cohesive layer was modeled using modified glyben, which provides
favourable characteristics for scaled physical model tests. The primary advantages of
modified glyben for reduced scale model tests are: i) it consolidates at a very slow rate
after application of confining pressure, thus it can be used in 1-g and N-g tests without a
observing a consolidation stage; ii) it resists desiccation due to drying; and iii) its
mechanical properties do not significantly change with time, which facilitates its
multiple use as part of the physical testing program. Also, laboratory results showed that
modified glyben have a normalized shear modulus versus shear strain amplitude
response that is independent on the confining stress, and glycerin to water ratio (Turan et
al. 2009a). Therefore, it is suitable for 1G environment tests.

The modified glyben was prepared by mixing bentonite with glycerine and water. The
fluid to dry bentonite ratio was 85% and the water to glycerine ratio was 50%. These
ratios were chosen to simulate prototype soft clay. Hand-held vane shear test (Pilcon)
was used to measure the undrained shear strength of the modified glyben mixture after
preparation and placement in the soil container. After light compaction, the undrained
shear strength of the model clay Cu was about 19 kPa at the bottom of the clay layer and
8 kPa near the top with a density of 1406 kg/m3.

61
The corresponding average shear wave velocity for an average Cu of 15 kN/m2 was
estimated to be 40 m/s using the empirical relation reported in Turan et al. (2009a), i.e.

Vs = 26.52 Ln (cu) – 33 [1]

Using similitude relationships shown in Table 1, with a geometric scaling factor λ=20,
The model clay shear wave velocity of 40 m/s correspond to a prototype clay shear wave
velocity of approximately 113 m/sec. This shear wave velocity corresponds to a soft to
medium consistency for prototype clay.

The main source of the shear resistance of the microgrid mesh is derived from the
interlocking mechanism with the surrounding backfill. Therefore, the top engineered
backfill layer was simulated using aggregates with an average particle size of 5 mm. The
peak angle of internal friction of the aggregates was measured as 0 from direct shear
tests.

3.3.4 Model Foundation System

The reduced scale foundation model was made of four acrylic tubes with a length of 60
mm and a diameter of 19 mm. These piles were rigidly connected to a steel cap of 20 X
20 mm and 6 mm thickness which correspond to a concrete cap of 0.7 m thickness. Four
plastic rings were fabricated and installed to facilitate the load transfer between a
microgrid mesh and the piles. The microgrid mesh was extended and pressed inside the
rings using four steel bolts. Rings were fixed in their vertical position using a plastic bolt.
Figure 3-3 shows the details of the model foundation. Table 3.2 provides a summary of
the model scaling calculations.

62
Figure 3-3: The scaled piled cap geogrid model

63
Table 3-2: Model Piles calculations, λ = 20

Parameter Model pile (Acrylic) Calculated Prototype

pile (Steel)

Outer diameter (mm) 19 380

Length (mm) 600 12000

Young's modulus (kN/m2) 3.2 X 106 200 X 106

Flexural rigidity, EpI (kN.m2) 1.82 X 10-2 5.81 X 104

Axial rigidity, EpA (kN) 6.03 X 105 3.5 X 106

3.3.5 Model Geosynthetics

The microgrid mesh was introduced within the aggregate layer at a depth of 3.5 times the
pile diameter (i.e 66 mm). The geosynthetics were simulated in the physical model tests
by using a bi-axial knitted Microgrid manufactured by STRATAGRID. STRATAGRID
Microgrid was manufactured utilizing a complex knitting process and polymer coatings
to provide various engineering properties and constructed of high molecular weight and
high tenacity polyester yarns. The yarns provide significant tensile capacity through
precision knitting into a dimensionally firm, uniform network of apertures. The model
stuffiness value of the Microgrid at 2% Jm = 110 kN/m. With a geometrical scale factor λ
= 20, the prototype stiffness Jp= Jm λ2 = 110 x 202 = 44000 kN/m (EL-Emam and
Bathurst 2004). Table 3.3 shows the main engineering properties of the STRATAGRID
Microgrid.

64
Table 3-3: Engineering properties of geogrid

Index Properties Test Method Value

Ultimate Strength ASTM D-6637 Method A 29.2 kN/m

Creep Limited Strength ASTM D-5262/D-6992 16.8 kN/m

Strength @ 2% Strain ASTM D4595 2.2 kN/m

Strength @ 5% Strain ASTM D4595 (MD) 7.3 kN/m

Strength @ 5% Strain ASTM ASTM D4595 (CMD) 4.4 kN/m

Aperture Size Measured 2.54 x 6.35 mm

3.3.6 Model Superstructure

The superstructure was modeled using a single degree of freedom system. The model
single degree of freedom (SDOF) structure was machined using a 30 mm long steel
column with a square cross-section of 25x25 mm, supporting a 9.5 kg mass. This mass
was connected to the column with a 10 mm diameters bolt (see Figure 3-4). The natural
frequency of the SDOF structures was determined via a sine sweep test conducted on the
shaking table while the SDOF structure was rigidly clamped on top of the table. The
natural periods of the SDOF structures correspond to two structures with two different
natural frequencies. Table 3.4 shows the model and prototype natural frequencies of the
two structures. The SDOF with lower natural frequency is denoted low frequency (LF)
SDOF and the SDOF with the higher natural frequency is denoted high frequency (HF)
SDOF.

65
Table 3-4: SDOF structures model and prototype frequencies

Simulated BLDG BLDG BLDG SDOF

BLDG Type Height Period T Frequency – Frequency –
(m) (sec)* prototype (Hz) model (Hz)

Steel moment 18.5 0.45 2.24 10.03

frame

Steel moment 26.5 0.59 1.71 7.64

frame

*Ta = 0.05 h0.75 (NBCC CI.4.1.8.11.3(c))

(a) SDOF Low Frequency

66
 (b) SDOF High Frequency

(c) Accelerometer incased

Figure 3-4: a) The SDOF High system b) SDOF Medium system, c) ADXL 203
Accelerometer

67
3.3.7 Instrumentation

Ten accelerometers (AC1–AC10) were used to monitor the acceleration response of the
shaking table, the model soil deposit, the piled cap and the SDOF model structure. The
accelerometers (type ADXL203) were small size high precision devices with dual-axis.
The ADXL203 accelerometers were capable of measuring accelerations for a range of
±5g with a sensitivity of ±0.3 % mv/g. In addition, a laser displacement transducer
(DISP1) (Matsushita-KDCL) was fixed on the laminar container frame to measure the
shaking table motion, which could be used to verify the base acceleration signal
measured by AC1 on top of the table. The output signals were recorded as analog
voltages that were proportional to acceleration and displacement. Figure 3-4 (c) shows an
accelerometer containing capacitors, resistors and power supply elements and encased in
a plastic box for protection against humidity and contamination due to soil contact. The
signal output and power cables of each accelerometer were connected to an amplifier
box, which was connected to the data acquisition system.

68
Figure 3-5 shows the positions of all accelerometers in/on the soil container and on the
model piled-cap-structure system. AC1 was used to measure the response on the shaking
table. The soil response was monitored by accelerometers AC2 to AC6 embedded in the
soil column at a distance 20 cm from the edge of the pile cap as depicted in Figure 5.
AC8 was fixed on top of the pile cap to monitor the pile cap lateral response. AC7 and
AC9 were fixed at the edges of the pile cap with 90 degrees angle with horizon to
measure the vertical response at both edges of pile cap. The response of SDOF structure
was monitored using AC10, which was mounted at the top of the lumped mass model.

Figure 3-5: Instruments distribution diagram

3.3.8 Shaking Table Control System

The shaking table system was managed by an in-house control software (FRE) developed
at the Western University BLWT (Boundary Layer Wind Tunnel). The FRE software is
capable of running the table with various vibration signals such as harmonic, sine weep
69
and white noise. The vibration system responds to a low voltage signal by producing a
corresponding displacement through the actuator of the shaking table, as shown in the
schematic diagram in Figure 3-6. The FRE is also capable of simultaneously generating
the control signal and recording output from displacement and acceleration sensors.

Figure 3-6: The shaking table and control system

3.3.9 Placement of Model Soil Deposit

The testing program was commenced by assembling the laminar container on top of the
shaking table. A wood board covered with steel grits was then bolted to the shaking table
at the bottom of the lamina container. The steel grits were glued to the wood board using
epoxy and installed on the table to ensure transfer of the table motion to the soil column
through friction forces at the steel grit-soil interface. Prior to soil placement, the sides of
the laminar box were covered with latex membrane to prevent soil particles from
migrating out of the box through the openings between the laminae. The initial bedding
70
layer of the model soil column was backfilled inside the box using layers of sand. Each
layer of sand was compacted after applying several cycles of harmonic shaking signals
until an 80 % relative density has been achieved (which was determined based on
monitoring the height of the sand in the box).

Sand backfilling proceeded in layers where each layer was compacted by shaking each
20 kg of sand into a 2 cm layer until a total layer depth of 36 cm is reached. Blocks of
soft modified glyben were then placed over the last sand layer. The blocks of soft
modified glyben were prepared by pressing weighed quantities of clay in a small cubical
acrylic box with dimensions of 70 x 70 x 50 mm. The clay blocks were positioned side
by side and compacted using a drop hammer. Each layer of clay had 40 mm thickness
with 1406 Kg/m3 bulk density. Figure 3-7 depicts the compaction process of Modified
Glyben.

The clay layer thickness was varied in accordance with the following scenarios studied as
part of this research: i) shallow backfill case with the clay removal/aggregate backfill
thickness of 30 mm, and ii) deep backfill case with the clay removal/aggregate backfill
thickness of 80 mm. The aggregate backfill was compacted using the same drop hammer.
Each 30 mm layer of aggregate weighed 20 kg and compacted into a density of 1650
kg/m3. The aggregate backfill was continued in layers up to the bottom of the model cap
foundation. After the compaction process, the clay undrained shear strength was
measured using the hand vane shear apparatus. The average undrained shear strength cu
of the top 100 mm of clay was determined as 9 kPa. Beneath that, the clay layer had an
average cu of 19 kPa. This increase in clay strength was due to the compaction effort.

The model pile-cap system was placed into the soil bed after all layers were prepared as
described. In order to eliminate excessive settlement during the testing program, the
lumped mass was placed and maintained on top of the cap for 24 hours to have the piles
complete the immediate settlement.

71
 Figure 3-7: Compaction of Modified Glyben

3.3.10 Experimental Scenarios

Several shaking table tests were carried out with the objectives of i) studying the
influence of the geosynthetics reinforcement on the dynamic performance of the pile
foundation, ii) studying the influence of the geosynthetics reinforcement on the dynamic
response of the single degree of freedom structures, and iii) exploring the potential for
using geosynthetics reinforcement to reduce the thickness of the aggregate backfill. The
physical model of the geosynthetics-reinforced pile-structure system was excited using
three main types of signals: sine sweep signals, harmonic signals and a scaled earthquake
acceleration time history. Table 3-5 presents a summary of the shaking table tests.
72
Table 3-5: Summary of tests

Predominant Peak Ground

Type of input
Test ID frequency Purpose
motion
(Hz) Accel. (PGA), g

Response of LF SDOF system:

SLNF-1
1) deep backfill without
SLNF-2 Sine Sweep 5 - 19 0.25 geogrid case, 2) deep backfill
with geogrid, and 3) shallow
SLNF-3
backfill without geogrid

Response of HF SDOF
SHNF-1 system: 1) deep backfill
without geogrid case, 2)
SHNF-2 Sine Sweep 5 - 19 0.25
deep backfill with geogrid,
SHNF-3 and 3) shallow backfill
without geogrid

UPEQ-L Random Response of LF and HF SDOF

earthquake 16 0.3 systems to scaled Upland
UPEQ-H time history Earthquake time histories.

SAMP-1
Ground motion
SAMP-2
Harmonic amplification: shallow &
16 0.06 to 0.1
signal deep backfill with or without
SAMP-3
geogrid (Table 3-6)
SAMP-4

Notes: LF is Low natural frequency; HF is High natural frequency

The sine weep excitation was used to examine the response of the LF and HF SDOF
structures and the pile cap model (cases SLNF and SHNF). The shaking table control
73
system enables users to generate various sine sweep signals with a range of frequencies
at a specific maximum amplitude. The soil-structure model has been excited with a 5
minute long sine sweep signal with frequencies ranging from 5 Hz to 19 Hz. The sine
sweep test was repeated after installing the microgrid mesh at a depth of 3.5 times the
pile diameter within the aggregate backfill. This sine sweep test was also performed to
investigate the effect of reducing the thickness of the backfill layer on the system
response.

The geosynthetics-soil-foundation-structure model was also subjected to scaled

earthquake time histories (cases UPEQ-L & UPEQ-H). The earthquake time history was
generated from the original 1990 Upland earthquake with a peak ground acceleration of
0.25g and a predominant frequency of 3.6 Hz (see Figure 3-8). The predominant
frequency of 3.6 Hz corresponded to a scaled model frequency of 16 Hz based on a
geometric scaling factor of 20. The shaking table was used to simulate the scaled
earthquake acceleration time history. As shown in Figures 3-9a and 3-9b, the comparison
of the Fourier spectrum of scaled input signal and response measured at table top show
close agreement in terms of amplitude and frequency content. The minor discrepancies
could be due to the limited abilities of the actuator and the control system due to lacking
a motion correction feedback system.

0.3
0.2
Acceleration g

0.1
0
-0.1
-0.2
-0.3
0 5 10 15 20 25
Time (sec)

Figure 3-8: The earthquake acceleration time history

74
 0.024
0.022
0.02
0.018
0.016
Fourier Amplitude

0.014
0.012
0.01
0.008
0.006
0.004
0.002
0
1 10 100
Frequency (Hz)

(a)

0.024
0.022
0.02
0.018
0.016
Fourier Amplitude

0.014
0.012
0.01
0.008
0.006
0.004
0.002
0
1 10 100
Frequency (Hz)

(b)

Figure 3-9: Fourier spectrum of model scale Upland EQ (a) input signal (b) signal at
table top

75
The model was also excited with four harmonic signals with 16 Hz frequency and
amplitudes ranging from 0.06g to 0.1g (cases SAMP1 to SAMP4). Table 3.6 presents a
summary of these cases and the amplification factor calculated from the acceleration
response of each case. The amplification factor was obtained from dividing the amplitude
of soil response by the base excitation amplitude. This group of tests was performed to
investigate the input motion amplification through the soil column. Similar tests were
also conducted to investigate the cases where the microgrid mesh was present in the
backfill. This harmonic signal was used to investigate the response of the system with
reduced-thickness backfill layer.

Table 3-6: Summary of harmonic tests

Case # Base Amplification
acceleration factor
(g) Shallow
Deep backfill Deep backfill backfill
with microgrid without microgrid without
microgrid
SAMP-1 0.061 3.705 3.951 4.754
SAMP-2 0.090 7.444 8.889 9.778
SAMP-3 0.100 7.400 8.500 9.400
SAMP-4 0.120 6.667 8.333 10.00

3.4 Results and discussion

The results of the shaking table tests performed in this study are summarized in this
section. The results include the response of free field ground response analyses, the
response of the model geosynthetics-reinforced pile foundation system and the response
of the SDOF superstructure models. The results are discussed comparatively considering
the cases of shallow backfill and deep backfill. The feasibility of using polymer strips to
enhance the dynamic induced lateral behaviour of the model pile-cap-superstructure
system was also investigated.

76
3.4.1 Free Field Response

The results of harmonic tests that were conducted in the absence of model
foundation/structure system are presented here (SAMP1 to SAMP4). The results
obtained from AC6 are summarized in Figure 3-10 and Table 3.6. Figure 3-10 shows the
amplification factors for lateral acceleration plotted against the input base acceleration
amplitudes. It can be seen that the amplification factor showed an increasing trend with
the increase of excitation amplitude due to the increased non-linearity experienced by the
dynamically loaded soil. The results also indicated that this case, where aggregate
backfill had the smaller thickness experienced the largest amplification. The case of
deeper backfill exhibited a lower amplification factor. The lowest amplification was
observed when the microgrid mesh was present within the deep backfill. Around 23 %
reduction in the amplification factor from almost 8.5 to 6.5 can be observed at a base
acceleration of 0.12 g. This behaviour can be attributed to the stiffening of soil column
due to the combined effect of increased backfill thickness and the addition of microgrid.
The large amplification factors can be attributed to the fact that the models were excited
with a harmonic load, that has a frequency of 16 Hz, very close to that of the system
valued at 15.4 Hz, and thus resonance conditions occurred.

77
 14.5 Shallow backfill withiout microgrid

Deep backfill without microgrid

12.5
Amplification factor

Deep backfill with microgrid

10.5

8.5

6.5

4.5

2.5
0.05 0.06 0.07 0.08 0.09 0.10 0.11 0.12 0.13
Base Acceleration (g)

Figure 3-10: Free field response comparison (SAMP-1 to SAMP-4).

3.4.2 Pile cap response

Figure 3-11 shows the results of the sine sweep test with respect to the lateral response
measured at the pile cap using AC8 (cases SLNF 1-3). This test was run while the LF
SDOF structure was affixed on top of the pile cap. The response comparison shows that
the presence of the microgrid mesh had a significant effect on the lateral response of the
pile cap level. This could be attributed to the lateral stiffening of the foundation due to
the addition of the microgrid reinforcement. When the microgrid reinforcement was
present, the Fourier amplitude was reduced around 15% from 2.6 to 2.2 at a frequency of
10 Hz’s.

78
 3
Shallow backfill without microgrid
Deep backfill without microgrid
2.5
Deep backfill with microgrid

2
Fourier Amplitude

1.5

0.5

0
0 5 10 15 20 25
Frequency (Hz)

Figure 3-11: The dynamic response comparison of the piled cap foundation (SLNF-
1 to SLNF-3).

Figure 3-12 depicts the Fourier amplitudes for three test cases excited using scaled
earthquake excitation (cases UPEQ-L). It can be seen that increasing the backfill depth
and the addition of microgrid reduced the lateral seismic response of the pile cap.

79
 0.12
ACC8_Shallow backfill without microgrid
0.1 ACC8_Deep backfill without microgrid
ACC8_Deep backfill with microgrid
Fourier Amplitude

0.08

0.06

0.04

0.02

0
5 15 25 35 45
Frequency (Hz)

Figure 3-12: Dynamic response of pile foundation with and without reinforcement
(UPEQ-L).

Figure 3-13 shows the Fourier amplitudes of the sine sweep tests for the HF SDOF
structure was fixed on top of the pile cap (cases SHNF 1-3). Similar to the case of LF
SDOF, the response comparison shows that the addition of the microgrid attenuated the
response at the pile cap level. Also, Figure 3-13 shows that the reduction of backfill
caused the lateral cap response to increase.

80
 5
Shallow backfill without microgrid
4.5
Deep backfill without microgrid
4 Deep backfill with microgrid
3.5
Fourier Amplitude

2.5

1.5

0.5

0
0 2 4 6 8 10 12 14 16 18 20 22 24
Frequency (Hz)

Figure 3-13: The dynamic response comparison of the piled cap foundation (SHNF
1-3)

3.4.3 Low Frequency Structure Response

The sine sweep test results of the LF SDOF structure were obtained from AC10 and are
illustrated in Figure 3-14 (cases SLNF 1-3). The comparison of the Fourier amplitudes
for the aforementioned test cases shows that the model with the smaller backfill thickness
displayed the highest response of the SDOF superstructure at all frequencies. Increasing
the backfill thickness significantly reduced the response. A further reduction in the
response of the SDOF superstructure was observed for the case where the microgrid was
present. This additional reduction in the structural response was small due to three
reasons: the thicker backfill has already stiffened the system sufficiently; the relatively
shallow embedment depth of the microgrid (i.e. low overburden), which did not allow the
full mobilization of the resisting forces from microgrid; and the relatively low amplitude
input motion, which resulted in small deformations (i.e. strains in the microgrid), which
resulted in lower stiffness of the microgrid (because its stiffness is hyper-elastic, with
higher stiffness for larger strains as shown in Chapter 2). Therefore, it can be expected

81
that the beneficial effect of the geosynthetics reinforcement would be more prominent for
larger input motions with larger amplitudes and in situations with optimized backfill
thickness.

8
Shallow backfill without microgrid
7 Deep backfill without microgrid
Deep backfill with microgrid
6
Fourier Amplitude

0
0 2 4 6 8 10 12 14 16 18 20 22 24

Frequency (Hz)

Figure 3-14: The dynamic response comparison of the low frequency SDOF
superstructure (SLNF-1 to SLNF-3).

Figure 3-15 shows the Fourier amplitudes of the three test cases excited using the scaled
earthquake excitation (see UPEQ-L). Despite that at low frequencies the three backfill
cases contributed to almost the same response, It is can be seen that at frequencies
beyond 16 Hz both increasing the backfill depth and addition of microgrid attenuated the
lateral seismic response of the low frequency SDOF superstructure.

82
 0.3
Shallow backfill without geogrid
0.25 Deep backfill without geogrid
Deep backfill with geogrid
Fourier Amplitude

0.2

0.15

0.1

0.05

0
0 5 10 15 20 25
Frequency (Hz)

Figure 3-15: The dynamic response comparison of the low frequency SDOF
superstructure (UPEQ-L).

3.4.4 High Frequency Structure Response

Figure 3-16 shows the sine sweep tests results that were carried for the HF SDOF
structure system (cases SHNF-1 to 3). It can be noted from Figure 3-16 that the
microgrid attenuated the response of the HF SDOF structure only slightly. The
attenuation remained insignificant compared to what was achieved for the LF SDOF
structure. In the sine sweep load, with constant acceleration amplitude, the lateral
displacement amplitudes of the cap associated with the High frequency SDOF would be
lower than the displacement amplitudes associated with the low frequency SDOF. Thus,
the strains in the microgrid are lower for the high frequency SDOF structure compared to
the low frequency SDOF structure. Therefore, the effect of microgrid was not noticeable
in the lateral vibration of the cap connected with the high frequency SDOF structure.

83
 3.5
Shallow backfill without microgrid
3 Deep backfill without microgrid
Deep backfill with microgrid
2.5
Fourier Amplitude

1.5

0.5

0
0 5 10 15 20
Frequency (Hz)

Figure 3-16: The dynamic response comparison of the high frequency SDOF
(SHNF-1 to SHNF-3).

Figure 3-17 compares the Fourier amplitudes of the three backfill test cases for the scaled
earthquake motion UPEQ-H. In contrast to the LF SDOF system test results, Figure 3-17
indicates that there was almost no change in the structure response due to the addition of
microgrid. Comparing the results in Figures 3-15 and 3-17 indicates that the structural
response of the HF SDOF structure was almost half the response of the LF SDOF
system. The reduction in the response amplitudes for the HF SDOF structure resulted in
reduced lateral foundation deformations, and hence reduced strains in the microgrid, i.e.,
it was not engaged and consequently was ineffective at this low level of lateral
displacement.

84
 0.14
Deep backfill with microgrid
0.12
Deep backfill witout microgrid

0.1
Foureir amplitude

0.08

0.06

0.04

0.02

0
0 5 10 15 20 25 30 35 40 45 50 55

Frequency (Hz)

Figure 3-17: The dynamic response comparison of the high frequency SDOF
structure (UPEQ-H).

3.4.5 Pile cap rocking response

Figure 3-18 shows the Fourier amplitude spectrum of the vertical acceleration response
to the sine sweep signal measured at the corner of the pile cap for three test cases (SLNF-
1 to 3) measured by AC9. It can be noted that the vertical acceleration amplitudes at the
pile cap was attenuated as a result of adding the microgrid mesh. Figure 3-18 also shows
that reducing the thickness of the backfill layer increased the response and caused
lengthening in the predominant period. The sine sweep tests revealed that increasing the
backfill thickness from 8 cm to 3 cm reduced the peak cap rocking motion by 15 % and
the addition of microgrid reinforcement further reduced the peak rocking motion by 9 %.
These results show the favourable effects of adding the microgrid mesh on the pile cap

85
rocking vibration mode. It also indicates that selecting a proper backfill depth is vital in
controlling the rocking mode of vibration of the pile cap.

1.4
1.3 Deep backfill with geogrid
1.2
1.1 Deep backfill without geogrid
1 Shallow backfill without geogrid
0.9
Fourier Amplitude

0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0
4 6 8 10 12 14 16 18
Frequency (Hz)

Figure 3-18: The dynamic vertical response comparison measured at the pile tip
(SLNF-1 to SLNF-3).

After removing the structure, harmonic shaking tests with frequency = 16 Hz and
amplitude = 0.12 g were performed (cases SAMP-1 to 4). Figure 3-19 shows the results
of the harmonic signal excitation measured at the corner of the pile cap by AC9 for the
cases of deep backfill with and without microgrid and the shallow backfill case. The
comparison between these three cases shows that, as expected, using a thicker granular
backfill attenuated the rocking vibration of the cap and hence, the vertical response. It
can also be noticed that the addition of the microgrid mesh further reduced the rocking
motion at the pile cap due to stiffness contribution by the microgrid mesh.

86
 0.35 Deep backfill with geogrid

0.3 Deep backfill without geogrid

0.25
Acceleration Amplitude (g)

Shallow backfill without

geogrid
0.2

0.15

0.1

0.05

0
0.5 0.6 0.7 0.8 0.9 1 1.1 1.2 1.3
-0.05

-0.1
Time (sec)

Figure 3-19: Vertical response at the pile tip comparison (SAMP-4)

87
3.5 Summary and conclusion

This chapter presented the results of a series of reduced scale shaking table tests
performed to study the influence of geosynthetics reinforcement on the dynamic response
of SDOF superstructures supported on pile foundations. The results presented here
provide physical evidence to the effectiveness of the proposed foundation concept and
can be used to calibrate numerical models. The following is a summary of observations
made during the tests and the conclusions derived from this experimental study;

 As expected, increasing the thickness of the backfill layer reduced the ground
motion amplification and the maximum dynamic response of the soil column to
harmonic loading. The addition of the microgrid mesh further reduced the lateral
response by 23%.
 The scaled earthquake and sine sweep tests indicated that the microgrid
reinforcement resulted in reducing lateral response of the pile foundation, even
for the case with thicker engineered backfill. The maximum lateral response of
the pile cap decreased farther for the case of the geosynthetics-reinforced backfill.
 The lateral response of the SDOF systems to the scaled earthquake and sine
sweep tests decreased for the case of microgrid-reinforced backfill. The sine
sweep tests results indicated that the thick backfill reduced the maximum lateral
response of the low frequency SDOF structure and the microgrid reinforcement
reduced it further. The effectiveness of the microgrid reinforcement in reducing
the dynamic response was more pronounced at higher inertial interaction
associated with the low frequency SDOF due to the larger dynamic loads
associated with larger deformations in the microgrid.
 The rocking vibrations of the pile cap due to the sine sweep tests and harmonic
loading were reduced due to the geosynthetics reinforcement. The sine sweep
tests revealed that increasing the backfill thickness from 3 cm to 8 cm reduced the
peak cap rocking motion by 15 % and the addition of microgrid reinforcement
further reduced the rocking motion by 9 %.
References
88
[1] Bathurst RJ, Alfaro MC. 1996. Review of seismic design, analysis and
performance of geosynthetic rein- forced walls, slopes and embankments.
International Symposium on Earth Reinforcement, keynote lecture, Kyushu, 1996.
p. 23-52.

[2] Cai, Z., and Bathurst, R.J. 1995. Seismic response analysis of geosynthetic
reinforced soil segmental retaining walls by finite element method. Computers and
Geotechnics, 17 (4), 523-546.

[3] El-Emam, M.M., Bathurst, R.J. 2007. Influence of reinforcement parameters on the
seismic response of reduced-scale reinforced soil retaining walls. Geotextiles and
Geomembranes, 25 (1), 33-49.

[4] El-Emam, M. M., and Bathurst, R.J. 2004. Experimental design, instrumentation
and interpretation of reinforced soil wall response using a shaking table.
International Journal of Physical Modelling in Geotechnics, 4, 13-32.

[5] Gohl, W. 1991. Response of pile foundations to simulated earthquake loading:

experimental and analytical results, Ph.D. Dissertation, Univ. of British Columbia.

[6] Helwany, S.M.B., Budhu, M., and McCallen, D. 2001. Seismic analysis of
segmental retaining walls. I: Model verification. Journal of Geotechnical and
Geoenvironmental Engineering, 127 (9), 741-749.

[7] Iai, S., 1989. Similitude for shaking table tests on soil-structure-fluid models in 1 g
gravitational field. Soils and Foundations, 29 (1), 105–118

[8] Koseki,J., Munaf, Y., Tatsuoka, F., Tateyama, M., Kojima, K. and Sato, T. 1998.
Shaking and tilt table tests of geosynthetic-reinforced soil and conventional-type
retaining walls. Geosynthetics International, 5 (1), 73-96.

[9] Kubo, K. 1965. Experimental study of the behaviour of laterally loaded piles.
Proceedings of the 6th International. Conference on Soil Mechanics and Foundation
Engineering, Montreal, Vol. 2, 275-279.
89
[10] Liu, H. and Chen, K. 1991. Test on behavior of pile foundation in liquefiable soils,
Proceedings of the 2nd International Conference on Recent Advances in
Geotechnical Engineering and Soil Dynamics, St. Louis, Vol.1, 233-235.

[11] Liu, H.L., Ng, C.W.W., and Fei, K. 2007. Performance of a geogrid-reinforced and
pile-supported highway embankment over soft clay: Case study. Journal of
Geotechnical and Geoenvironmental Engineering, 133 (12), 1483-1493.

[12] Liu, H. 2009. Analyzing the reinforcement loads of geosynthetic-reinforced soil

walls subject to seismic loading during the service life. Journal of Performance of
Constructed Facilities, 23 (5), 292-302.

[13] Makris, N., Tazoh, T., Yun, X., and Fill, A. 1997. Prediction of the measured
response of a scaled soil-pile-superstructure system. Soil Dynamics and Earthquake
Engineering, 16, 113-124.

[14] Matsumo, O., Tsutsumi, T., Yokoyama, K. and Saito, Y. 1998. Shaking table tests
and analysis of geosynthetic-reinforced soil retaining walls. Geosynthetic
International, 5 (1-2), 97-126.

[15] Meymand, P. 1998. Shaking table scale model tests of nonlinear soil-pile-
superstructure interaction in soft clay, Ph.D. Dissertation, University of California,
Berkeley.

[16] Michalowski R L. 1998. Soil reinforcement for seismic design of geotechnical

structures. Computers and Geotechnics, 23 (1-2), 1-17.

[17] Rocha, M. 1957. The possibility of solving soil mechanics problems by the use of
models. Proceedings of 4th International Conference of Soil Mechanics and
Foundations Engineering, Vol.1 pp, 183-188.

[18] Sakaguguchi, M., Muramatsu, M. and Nagura, K. 1992. A discussion on reinforced

embankment structures having high earthquake resistance”. Proceedings of the

90
 International Symposium on Earth Reinforcement Practice, IS-Kyushu’92, Ochiai,
Hayashi and Otani (Editors), Fukuoka, Japan, pp. 287-292.

[19] Sreerama, K. 1993. Dynamic pile-soil-pile interaction using model tests under
simulated earthquakes. Ph.D. Dissertation, University of Missouri-Rolla, MO.

[20] Tao, X., Kagawa, T., Minowa, C., and Abe, A. 1998. Verification of dynamic soil-
pile interaction”. Proceedings of the 3rd Conference on Geotechnical Earthquake
Engineering and Soil Dynamics, ASCE, Seattle, 1199-1210.

[21] Tatsuoka, F., Sakamoto, M., Kawamura, T., and Fukushima, S. 1986. Strength and
deformation characteristics of sand in plane strain compression at extremely low
Pressures. Soils and Foundations, 26 (1), 65-84.

[22] Turan, A., Hinchberger, S.D., El Naggar, M.H. (2009a). Mechanical

characterization of an artificial clay. Journal of Geotechnical and
Geoenvironmental Engineering, 135 (2), 280-290.

[23] Turan, A., Hinchberger, S.D., El Naggar, M.H. (2009b). Design and
commissioning of a laminar soil container for use on small shaking tables. Soil
Dynamics and Earthquake Engineering, 29 (2), 404-414

[24] Yan, L., Byrne, P., and Dou, H. 1991. Model studies of dynamic pile response
using hydraulic gradient shaking table tests. Proceedings of the 6th Canadian
Conference of Earthquake Engineering, Toronto, 335-342.

91
 Chapter 4
4. NUMERICAL STUDY ON THE DYNAMIC LATERAL
BEHAVIOR OF GEOSYNTHETICS-REINFORCED
PILE FOUNDATION SYSTEM

This chapter presents the finite-element (FE) analysis for simulating the dynamic
performance of geosynthetics-reinforced pile foundation system. The (FE) models were
established using the program Plaxis 3D (Brinkgreve et al., 2012). The numerical models
were verified against the shaking table test results of a model scale geosynthetics-
reinforced pile foundation system as described in Chapter 3. A parametric study was
carried out to investigate the effect of different design parameters on the effectiveness of
the proposed geosynthetics-reinforced pile foundation system. These parameters
included: the frequency and amplitude of ground motion; the stiffness and strength of the
geosynthetic reinforcement, the location of the reinforcement within the backfill material
and the thickness of the backfill material. The numerical results indicated that the
geosynthetic-reinforcement greatly reduced the maximum lateral response of the pile cap
connected with the low frequency single degree of freedom structural model.

4.1 Introduction and problem overview

Geosynthetics are polymeric material consisting of tensile ribs with openings of
sufficient size to allow interlocking with the surrounding soil. This geosynthetics-soil
interlocking mechanism allows the geosynthetics to work as a reinforcement element,
which enhances the soil shear strength. The geogrid mesh is laid within the aggregate
engineered fill provides increased modulus and lateral confinement for the crushed stones
intruding the apertures of the geogrid. Therefore, geosynthetics have been widely used in
modern construction technology. Several application examples include geogrid
reinforced earth retaining walls (GRS), highway construction and expansion over soft
soils, geogrid-reinforced pile-supported highway embankments and geogrid reinforced
slopes. In these applications, geogrids have been widely implemented to reduce lateral
92
wall deflections arising from dynamic loads and uneven settlement of the supporting sub
grades and embankments.

Due to the rising concerns in the construction industry with regards to mitigating the
destructive effects of cyclic earthquake loads, several researches investigated the use of
geogrid-reinforced soils to enhance the seismic resistance of geotechnical structures.
These studies include: Bathurst and Alfaro (1996); Cai and Bathurst (1995);
Michalowski (1998); Helwany et al. (2001); Ling et al. (2004); Christopher (2004); Liu
(2009); El-Emam and Bathurst (2007); and Fakharian and Attar (2007). These studies
collectively demonstrated the superior performance of geogrid-reinforced walls to resist
lateral dynamic loading and provided verified finite element numerical models to analyze
the dynamic behavior of these walls.

In parallel, several researches investigated the seismic pile-soil interaction problem

through experimental and numerical studies. Several investigators such as Yegian and
Wright (1973), Angelides and Roesset (1980), Randolph (1981), Faruque and Desai
(1982), Trochanis at al. (1988, 1991) Wu and Finn (1997), and Bentley and El Naggar
(2000) developed finite element models to analyze the dynamic response of piles. These
researches proved that the finite element method is a powerful tool for analyzing the soil-
pile-structure interaction (SSPSI) problem.

Pile foundations are typically used when the ground conditions near the surface cannot
support the structural loads. In these situations, piles are used to transfer the loads to
more competent soil layers at larger depths below the ground surface. In many cases,
these foundations are subjected to significant lateral loads due to seismic or wind
loading. However, the weak subsurface conditions that dictate the use of pile foundations
result in low lateral foundation resistance, which poses a challenge to design engineers.
Conventional solutions involve either ground improvement techniques (e.g. jet grouting,
vibro compaction, etc.) or ground replacement for a substantial depth of the weak
surficial soil. These solutions are typically expensive and cause significant construction
delays.

93
This chapter introduces the dynamic analysis of an innovative use of geosynthetics,
where geogrid is used to enhance the lateral resistance of pile foundations. The shaking
table tests carried out to investigate the behavior of a model geogrid-pile foundation
system were used to calibrate a finite element model developed using the program Plaxis-
3D, which was then used to conduct a parametric study. The results of the parametric
stud are presented and discussed herein.

4.2 Shaking table test description

The numerical model developed in this research was calibrated using the results of
shaking table tests. Figure 4-1 shows a schematic diagram of the shaking table test setup
that include: i) the soil model; ii) the laminar box (soil container); and iii) the shaking
table and, iv) the geogrid-pile foundation model.

Figure 4-1: The schematic of the shaking table test setup

Figure 4-1 shows three layers of soil from bottom to top: i) a silty sand to represent
underlying soil layer; ii) a synthetic soil layer (modified glyben) to simulate soft clay
94
soil; and iii) coarse aggregate with relatively small sizes to represent a backfill of course
grained soil. Modified glyben provide favourable characteristics for scaled physical
model tests (Turan et al., 2009). The initial bedding layer of the model soil column was
backfilled inside the box using layers of sand. Then, blocks of soft modified glyben were
placed over the last sand layer and were compacted using a drop hammer to form a
uniform soft clay layer. Finally, the aggregate backfill was compacted using the same
drop hammer.

The shaking table was excited using a harmonic base motion with frequency = 16 Hz and
amplitude = 0.06 g. The frequency of this harmonic signal is the scaled model frequency
of an excitation frequency of 3.6 Hz using a geometric scaling factor of 20. Figures 4-2
and 4-3 show the input motion time history and Fourier spectra. Figure 4-4 shows the
pile cap acceleration response for the case of geogrid-reinforcement embedded in the top
aggregate layer.

0.08

0.06

0.04
Acceleration (g)

0.02

0
12.8 13 13.2 13.4 13.6 13.8 14 14.2
-0.02

-0.04

-0.06
Time (sec)

Figure 4-2: Input motion acceleration time history used in the shaking table test

95
 1
0.9
0.8
Fourier Amplitude (g-s)

0.7
0.6
0.5
0.4
0.3
0.2
0.1
0
14 14.5 15 15.5 16 16.5 17 17.5 18
Frequency (Hz)

Figure 4-3: Fourier spectrum of the input motion used in the shaking table test

0.4
0.3
0.2
Acceleration (g)

0.1
0
-0.1
-0.2
-0.3
-0.4
12.8 13 13.2 13.4 13.6 13.8 14 14.2
Time (sec)

Figure 4-4: Pile cap acceleration response

96
4.3 The numerical model
4.3.1 Problem dimensions
Figure 4-5 shows the piles, the geogrid connectivity and pile cap dimensions. The model
pile cap was fabricated from aluminum block with dimensions 200mmx200mm x6mm,
representing a prototype concrete pile cap of dimensions 4.00x4.00x0.7m. The model
piles were fabricated of four acrylic tubes 600 mm long and 20 mm in diameter,
representing steel piles 12 m long and 380 mm in diameter. These piles were rigidly
connected to a pile cap. Four plastic rings were fabricated and installed to connect the
microgrid mesh to the piles. The microgrid mesh was extended and affixed to the rings
using four steel bolts. Rings were fixed in their vertical position using a plastic bolt. The
microgrid mesh had a stiffness of 110 kN/m at 2% strain, representing a prototype
geosynthetic material (e.g. polymer strips) with stiffness of 44,000 kN/m.

Figure 4-6 shows the 3D finite element model of the geosynthetic pile foundation
composite system. This model simulated the dimensions of the real soil column, geogrid,
pile and cap dimensions of the shaking table test setup. The soil column consisted of a
soft glyben layer sandwiched between an aggregate layer from top and a bedding sand
layer. The geogrid mesh was introduced within the aggregate layer at a depth of 3.5 times
the pile diameter (i.e. 66 mm).

97
Figure 4-5 : Model pile foundation and microgrid connectivity

98
 Figure 4-6: 3D FE model of geosynthetic reinforced pile-cap system

4.3.2 Interface conditions

During the shaking table test, it was noticed that relative movement of the pile cap with
respect to the top aggregate layer had occurred. This behaviour was modeled using
interface elements at the bottom of the pile cap with reduced pile cap-soil interface
strength that was assumed to be 1 % of the soil strength. In contrast, the interlock
behaviour, which is expected to occur at the geogrid-aggregate interface was modeled
with nodal deformation compatibility which constrains the relative translation between
the geogrid mesh and the surrounding soil (Brinkgreve et. al, 2012). To facilitate these
two modeling procedures, the aggregate layer was split into two layers 3 cm and 5 cm in
which the reduced interface strength was allocated to the top layer and the rigid interface
layer was assumed at the bottom.

99
4.3.3 Material models and parameters:
The following section presents the description of the material models used in the
dynamic numerical analysis and the methodology followed to obtain the engineering
parameters that were used in these models.

4.3.3.1 Linear elastic soil model and parameters

The stress strain behavior of the bottom sand layer was simulated as a liner elastic soil
model to reduce the computational effort and time. This is justified since non-linearity
was not expected to occur in this layer due to its relatively high strength and stiffness,
and the relatively low shaking excitation amplitude applied and the associated small
strain amplitude experienced. Figure 4.7 provides the stress-strain hysteretic loop
calculated at a soil element within the sand layer indicating a maximum strain of 0.006%
(6x10-5), which is considered to be small strain and the soil stiffness can be represented
by the low strain shear modulus (i.e. maximum shear modulus, Gmax). The angle of
internal friction and shear modulus parameters of the sand layer was obtained via
laboratory direct shear tests. The sand (and aggregate) shear modulus were evaluated
using the equation proposed by Seed and Idriss (1970), i.e.:

G0 = 1000 K2max (σ’m)0.5 [1]

where:

G0 is the maximum (low strain) shear modulus in (psf)

K2max is the shear modulus number for the soil = 50 for loose sand and 75 for dense sand

σ’m is the mean effective confining stress of the soil in psf

The shear wave velocity can then be calculated as,

Vs = [2]

100
Table 4-1: Model granular Soil parameters

Parameter Sand

Unit weight, (kN/m3) 17.35

Relative density Dr (%) 80

Vsm (m/s) 50

peak 40

Gmax (kPa) 4427

Poisson’s ratio () 0.3

Damping ratio % 2

101
 300

200

100
Shear Stress (Pa)

0
-0.008 -0.006 -0.004 -0.002 0 0.002 0.004 0.006 0.008 0.01
-100

-200

-300

-400
Shear Srain %

Figure 4-7: Typical sand shear stress-strain loop (calculated from the shaking table
tests).

4.3.3.2 Nonlinear soil model and parameters

In contrast to the sand layer, it was expected that the aggregate and glyben layers would
undergo nonlinear stress-strain behavior due to the significant amplification of the input
motion and the relatively lower strength and stiffness of the two layers due to reduced
overburden pressure. Figures 4.8 and 4.9 provide the stress-strain hysteretic loops
calculated at soil elements within the aggregate and glyben layers, respectively,
indicating maximum strains of 0.05% (5x10-4) and 0.025% (2.5x10-4), which are
considered to be relatively high strain and the soil stiffness should be represented by
shear modulus corresponding to the strain level. Therefore, the nonlinear stress-strain
behaviors of the aggregate and glyben layers were modeled using the Plaxis built-in
Hardening Soil (HS) model with small strain stiffness (HSSMALL model).

102
 400

300

200
Shear Stress t(Pa)

100

0
-0.06 -0.04 -0.02 0 0.02 0.04 0.06
-100

-200

-300

-400
Shear strain g (%)

Figure 4-8: Typical aggregate shear stress-strain loop (calculated from the shaking
table tests).

103
 80

60

40
Shear Stress t (Pa)

20

0
-0.04 -0.03 -0.02 -0.01 0 0.01 0.02 0.03
-20

-40

-60

-80
Shear strain g (%)

Figure 4-9: Glyben shear stress-strain loop (calculated from the the shaking table
tests).

The Plaxis (HSSMALL) model accounts for the stiffness degradation for different types
of soils when subjected to primary deviatoric loading (Brinkgreve et al., 2012).
Simultaneous irreversible plastic strains develop in soils during deviatoric loading
causing the yield surface to change in size. In contrast to the elastic perfectly-plastic soil
models, the yield surface of the HS model expands due to plastic straining and varies in
the principle stress space. Also, the HS model captures the variation of the plastic axial
strain with the deviatoric stress by using the theory of plasticity rather than the theory of
elasticity and introducing a yield cap. The granular soil dilatancy is also considered in the
HS model. The model is characterized by stress dependent stiffness according to power
law, i.e., the unloading-reloading Young’s modulus is given by:

[3]

104
where, is the reference soil elastic modulus , and the cohesion (c), angle of internal
friction () are the soil effective strength parameters; Pref is the reference confining stress
taken at 100 kPa; : is the confining stress calculated at the mid depth of the clay layer;
and m is an exponent taken as 1 for soft clay.

The plastic straining due to primary deviatoric loading is represented by the reference
secant stiffness in standard drained triaxial test, , and the plastic straining due to
primary compression is represented by the reference tangent stiffness for primary
odometer loading, . can be estimated as a fraction of the soil elastic modulus,
using the relation:

= /3 [4]

while, is estimated as a fraction of , i.e.:

/1.25 [5]

The elastic unloading / reloading is described by the reference soil elastic modulus ,
and Poisson’s ratio (νur). Finally, the shear strength is evaluated according to the Mohr-
Coulomb failure criterion in terms of the strength parameters cohesion (c), angle of
internal friction () and dilatancy angle (ψ).

The basic equation of the HS model is the hyperbolic relation between the vertical strain,
ε1, and the deviatoric stress, q, in primary triaxail loading is formulated as:

- ε1 = for q < qf [6]

The hyperbolic relation can be visualized in figure 4-10 as follows:

105
Figure 4-10 : Hyperbolic stress-strain relationship in primary loading for a
slandered drained triaxial test (Schanz et al., 1999)
Where

Ei: is initial stiffness

Ei = [7]

q: is the current shear stress

qa: is the asymptotic value of the shear stress

qf: is the ultimate diviatoric stress , where:

qf = (c - ) [8]

Rf = qf / qa

In the cyclic loading applications, the HS model is limited to the assumption that soil
undergoes elastic loading and reloading behavior with no hysteretic damping. Some of
the limitations of the HS model in dynamic applications can be resolved by using the HS-
Small Strain (HSSMALL) model. The HSSMALL model accounts for soil stiffness at
106
small strain and non-linear degradation at large strain levels (Brinkgreve et al., 2012).
The HSSMALL model uses the Hardin and Drenvich (1972) hyperbolic law to relate the
shear modulus at large strains to small-strain properties. The Hardin and Drenvich (1972)
hyperbolic law is formulated as:

[9]

where the threshold strain γr is given by:

γr = [10]

Hence, in order to generate stiffness modulus reduction curve that covers the range of
soil stiffness from the low strain value (i.e. dynamic stiffness, G0) to the observed
hysteretic behaviour, two more parameters are required:

 The initial or very small-strain shear modulus, G0.

 The shear strain parameter at which the secant shear modulus is reduced to
about 70% of G0.

Evaluation of the nonlinear soil model parameters for glyben layer:

Surarak et.al (2012) carried out comprehensive triaxial and odometer testing on Bangkok
soft clay to determine its Hardening Soil model parameters , and .The
obtained parameters were used to calibrate a Plaxis triaxial test model. The research
revealed that the experimental stress strain curves were comparable with the Plaxis
curves. Therefore, the HS model parameters of the soft Bangkok clay were adopted in
this numerical calibration with undrained (B) analysis (Brinkgreve et al, 2012). Glyben
undrained shear strength value was obtained from the average measured undrained shear
strength cu.

The shear modulus at a given shear strain level is obtained using Equation [9] and the
shear strain parameter g0.7 was estimated from the shear modulus degradation curve of
modified glyben at G/Gmax = 70 % ,as shown in Figure 4-11 (Turan et al., 2009b).
107
 Figure 4-11: Stiffness degradation for modified glyben (Turan et al., 2009a)

Evaluation of the nonlinear soil model parameters for aggregate layer:

In this research the HS small strain model parameters for the aggregate layer were
calculated as follows. Instead of running traxial tests, the low strain shear modulus, G0,
was calculated using Equation [1]. In order to reflect the difference in overburden
pressure, the aggregate layer was subdivided into two sub-layers: top layer (above
geogrid); and bottom layer (below geogrid). HSSMALL stiffness parameters were
estimated based on the static shear modulus Gstatic . The Gstatic was estimated as 10% of
Go and the soil elastic modulus is calculated as 2Gstatic(1+ν). Knowing
estimating the confining stress σ3 from:

σ3= Ko σv [11]

Where
108
Ko = 1-sinφ [12]

σv = γsoil h [13]

The reference unloading reloading Young’s modulus was calculated using Equation
[3]. The angle of friction of aggregate was measured as 50º (Appendix A). Then,

and were calculated from Equations [4] and [5] respectively. was estimated
from the following equation:

= [14]

The shear modulus at any strain level is obtained from Equation [9].The shear strain
parameter g0.7 was estimated from the shear modulus degradation curve of aggregate at
G/Gmax = 70 % ,as shown in Figure 4-12 (Rollins et al., 1998). Table 4.2 summarizes the
clay and aggregate HSSMALL model parameters,

Table 4-2: The model HSSMALL clay and aggregate parameters:

Parameter Symbol Top Aggregate Bottom Aggregate Glyben

Soil unit weight (kN/m3) g 16 16 13.8

Secant stiffness in standard 2783 4680 800

2
drained triaxial test (kN/m )

109
Tangent stiffness for primary 2227 3750 850
2
oedometer loading (kN/m )

Unloading/reloading stiffness 8350 14.4X103 8000

2
(kN/m )

Power for stress-level M 1 1 1

dependency of stiffness

Cohesion 1 1 1

Friction angle φ 50 50 NA

Dilatancy angle Ψ 20 20 NA

Undrained Shear Strength cu NA NA 15

Reference shear modulus at small 36.3 X 103 61.1X103 377X103

strains (kN/m2)

Threshold strain, g 0. 007 0.007 0.06

Poisson’s ratio 0.2 0.2 0.2

4.3.3.3 Soil damping

The damping ratio of soils is an important parameter in numerical modelling of the
dynamic behaviour of soil-supported structures (Ju and Ni, 2007). The damping ratios of
the aggregate, glyben and sand layers were calculated from the stress strain loops
obtained from the acceleration data collected during the shaking table tests. Figures 4-7
to 4-9 show typical shear stress-strain hysteretic loops of the sand, aggregate and glyben
layers, respectively. The damping ratio was simulated in the numerical model via the two
Rayleigh damping coefficients αR and βR (Ju and Ni, 2007). These coefficients were used
to determine the damping matrix C, which is formulated as a function of the mass and
stiffness matrices as follows:
110
C = αR M + βR K [15]

Rayleigh damping coefficients αR and βR were determined from the damping ratio and
two different circular natural frequencies (ω1 and ω2) through the relationship :

αR= 2 ω1ω2 (D1 ω2 – D2 ω2) / ( - ) [16a]

βR = 2 (D2 ω2 – D1 ω2) / [16b]

4.3.4 The numerical model elements and mesh configuration

The soil was modeled using 10 node tetrahedral elements. The piles were modeled as
three node beam column elements using the built-in embedded piles elements. The pile
cap was modeled using plate elements. The Interface elements were introduced at the
bottom of the pile cap in order to simulate the separation that was noticed during the
shaking tests. The interface elements were modeled as 6 nodes triangular elements that
allow separation and relative deformation between the cap and soil as the deformation
exceeds the stiffness of the interface layer.

In the calibration phase, the numerical model simulated the configuration, dimensions
and boundary conditions of the physical model test setup. The vertical boundaries of the
model were surrounded with prescribed displacement surfaces that allow free translation
in the 1-D horizontal X and Z directions while restraining the translation at the transverse
Y direction. This scheme simulates the free horizontal translation of the lamina section
during 1-D shaking. At the bottom of the model, a prescribed surface was attached
through which a harmonic base excitation signal of 16 Hz and 0.06g amplitude was
introduced.

4.3.5 The numerical calculation process

The dynamic calculation process was divided into four main stages, namely: the initial
condition stage (to evaluate initial geostatic stresses); the construction of the geogrid-
pile-cap system stage; and the forced vibration analysis stage. In the fourth stage, the
111
geogrid mesh was deactivated in order to simulate the case where the geogrid mesh was
removed from the model pile cap system.

Before starting the initial phase, all structural elements were deactivated and the soil
body was activated. Once the initial stage was executed, the stresses in the soil body were
calculated using the K0 procedure and vertical stresses in equilibrium with the soil self-
weight were created. Hence, an initial stress field was generated within the finite element
mesh. Then, in the construction phase the soil, geogrid and structural elements were
activated together and an elastic-plastic deformation analysis was executed (Brinkgreve
et. al, 2012).

In the dynamic response analysis stage, the basic dynamic equation is solved using the
Newmark implicit time integration scheme (Rao , 2005):

[17a]

t=0 [17b]

Where M, C, K and F denote the known n x n mass, damping, stiffness and force
matrices, respectively. The Newmark method solves the above equation by using a step-
by-step procedure in which is obtained at discrete time intervals Δt apart (Rao,
2005).

4.4 Free Field Ground Response Analysis Using Deep

Soil
The free field ground response analysis was carried out using the software DeepSoil
(Hashash. et al., 2002). A harmonic signal having the scaled frequency of the Upland
earthquake with amplitude of 0.06g was used as the input motion for the three layer soil
column in a Deep soil model. The equivalent linear analysis method was carried out to
calculate the soil column response. The stiffness degradation and damping curves of the
surface aggregate layer under light confinement stress was obtained from the curves
developed by Rollins et al. (1998) (Figure 4-12 and 4-13). The stiffness degradation and
112
damping curves of glyben were obtained from Turan et al. (2009b). The Seed and Idress
(1975) stiffness degradation and damping curves of sand (built-in within DeepSoil) was
employed for the base sand layer. Figure 4-14 compares the acceleration response
calculated by DeepSoil and the acceleration measured by accelerometer AC-7 that was
placed at the surface of the gravel layer in the shake table test. The input motion used
was a harmonic signal of 16 Hz and 0.06 g amplitude. Figure 4-14 shows a close
agreement between the measured and calculated responses.

Figure 4-12 : Mean curves defining G/Gmax versus γ relationships for gravelly soils
at various confining pressures along with standard deviation boundaries for
reduced data set (Rollins et al. ,1998).

113
Figure 4-13 : Mean Curves Defining D versus γ relationships for gravelly soils at
various confining pressures along with standard deviation boundaries for entire
data set (Rollins et al. ,1998).

The input parameters used to generate the soil column response are summarized in Table
4-3. The average soil column natural frequency was calculated as 16 Hz which matches a
scaled natural frequency of 3.5 Hz.

114
Table 4-3: DeepSoil input parameters; model soil column.

Layer # Layer Name Model Prototype Unit Weight Model Shear

3
Thickness Thickness (kN/m ) Wave
(cm) (cm) Velocity (Vsm)
(m/s)

1 Aggregate 8 160 16 15

2 Glyben 23 460 14 40

3 Sand 36 720 18 50

0.3
Mesured Response_AC7
Calculated Response_DEEP SOIL
0.2

0.1
Acceleration (g)

0
12 12.5 13 13.5 14 14.5 15

-0.1

-0.2

-0.3
Time (sec)

Figure 4-14: Measured vs. calculated acceleration using DeepSoil.

115
4.5 Numerical Model Calibration Results
The same harmonic input signal was used to calibrate the Plaxis model with geogrid
being embedded within the top aggregate layer. Figure 4-15 compares the measured
results for the pile cap horizontal acceleration time history and the numerical model
predictions. Figure 4-15 shows excellent agreement between the measured and calculated
responses.

0.25
0.2
0.15
Acceleration (g)

0.1
0.05
0
-0.05
-0.1
-0.15
-0.2
-0.25
3.5 3.7 3.9 4.1 4.3 4.5 4.7 4.9 5.1 5.3 5.5
Measured Response_AC8 Plaxis_Undrained Analysis_AC8
Dynamic Time (sec)

Figure 4-15: Calculated vs. measured responses of pile cap reinforced with geogrid

4.6 Dynamic Model and Its Verification

This section describes a prototype scale model developed to study the dynamic behavior
of the geogrid-soil-pile foundation system. It also includes a parametric study carried out
to investigate the influence of geogrid stiffness depth and length on the kinematic pile
soil interaction. The prototype scale model was subjected to a lateral harmonic shaking
signal applied at the base of the supporting soil. The results obtained from the parametric
study were discussed in order to explore the beneficial effects of the geogrid
reinforcement in reducing the acceleration of the pile foundation system.

116
4.6.1 Soil Model
The soil model extends 100 meters from the foundation center in the direction of
dynamic shaking and 25 meters from the foundation system in the transverse direction
(see Figure 4-16). The model boundaries in the perpendicular direction were set as
viscous boundaries to absorb the shaking energy while the boundaries parallel to the
shaking direction were set as free boundaries. The soil finite element volume was
constructed out of 4 m tetrahedron elements with 10 nodes per element and 3
translational degrees of freedom per node.

Figure 4-16 : Soil model general view

As shown in Figure 4-17, the model soil profile is subdivided into three main layers: a
surface granular layer (aggregate); a soft clay layer; and a granular layer. The depth of
the three layers is equivalent to their scaled model depth on a geometric scale of 20. The
granular layers were assumed to follow a drained behavior while the soft clay layer was
assumed to follow the Undrained (A) behavior due to the nature of quick loading
associated with the shaking signal (Brinkgreve et. al, 2012). The stress-strain behavior of
the three layers was modeled using the HSSMALL model. The G0 values of the soil
layers were calculated from Equation 4.2 after assuming practical Vs values so that the
117
fundamental frequency of the prototype soil column was set at 3.7 Hz. The method of
calculating the HSSMALL parameters is similar to that used for the model granular
layer; as explained in section 4.3.3.2. Table 4-4 shows the Vs values and HSSMALL
parameters associated with each soil layer.

(a)

Backfill Layer 1.6 m

Soft Clay Layer

20.0 m 4.6 m

7.2 m
Dense granular
Layer

(b)

Figure 4-17 : a) Soil column stratigraphy without geogrid, b) Soil column stratigraphy
after adding geogrid

118
Table 4-4: The HSSMAL prototype parameters of each soil layer

Parameter Sym Top Backfill Interface Soft clay Dense

bol layer Layer layer Granular
layer

Soil unit weight γ 20 20 17 20

(kN/m3)

Shear wave velocity Vs 220 300 170 220

(m/s)

Secant stiffness in 31.6 X 103 58.8X 103 4.91 X 103 31.6 X 103
standard drained
triaxial test (kN/m2)

Tangent stiffness for 25.3 X 103 47.0 X 103 3.93 X 103 25.3 X 103
primary oedometer
loading (kN/m2)

Unloading/reloading 94.8 X 103 176.0X 103 14.7 X 103 94.8 X 103

stiffness (kN/m2)

Power for stress- m 1 1 1 1

level dependency of
stiffness

Cohesion 5 5 5 NA

Friction angle φ 40 40 25 40

Reference shear 98.8 X 103 184.0 X103 50.1X103 98.8X103

modulus at small
strains (kN/m2)

119
Poisson’s ratio 0.2 0.2 0.2 0.2

4.6.2 Piles-cap-geogrid-foundation model

Taking advantage of symmetry, the foundation model involved two steel piles 12 m long,
0.5 m in diameter with 6 mm wall thickness. The piles were spaced at 2.5 m and
supported a reinforced concrete cap, with dimensions 2 m X 4 m x 0.4 m depth (see
Figure 4-18). A geogrid mesh was embedded within the surface granular layer at a depth
of 1.6 m (as shown in Figure 4-17).

The linear elastic stress-strain model was used to simulate the behavior of the piles and
pile cap. The surfaces of the piles and the pile cap were modeled using 6 nodes triangular
shell elements with 6 degrees of freedom per node, three translational and three
rotational. The geogrid mesh was represented by 6-nodes triangular elements with 3
translational degrees of freedom per node. The geogrid elements were set to carry tension
loads with axial stiffness of 500 kN/m and no bending stiffness. The goegrid elements
were connected to the pile shell elements and the soil tetrahedron elements and share the
translational degrees of freedom at the connecting nodes. In the parametric study, the
geogrid mesh stiffness was increased up to 44,000 kN/m to study the influence of higher
stiffness polymer strips.

Figure 4-18: Pile-Cap–geogrid finite element mesh.

120
4.6.3 Interface model
Soil-pile interface nonlinearity is a major source of pile foundation stiffness degradation.
Due to cyclic loading, gapping and slippage occur at the pile-soil interface and cause
interface nonlinearity. El Nagger and Novak (1995) accounted for interface slippage
using rigid sliders connecting both piles and soil nodes. The pile and soil nodes were
allowed to move relative to each other when the shear force in the slider exceeded the
maximum shear force. Wu and Finn (1996) accounted for gapping by not allowing any
tension to occur between the pile and soil. This was achieved by keeping the normal
stresses at the pile soil-interface smaller than the assigned tensile strength. El Naggar and
Bentley (2000) modeled the pile-soil interface using non-tension springs connecting both
soil and piles elements. In the case of clays, gapping is allowed when these springs is
disconnected under tension loading. In the case of sand, the interface springs do not
allow gapping and sand follows the pile on the tension side with zero soil stiffness.
Maheshwari et al. (2003) considered separation between piles and soils using no tension
elements.

Similarly, for earth retaining structure, the mechanism in which shear forces are
transferred between geogrid and soil is of major importance and influences the overall
seismic wall behavior. Several researchers (e.g. Ling et al., 2004; Burke and Ling, 2004;
Cai and Bathrust, 1995; Helwany et al., 2001; and Ling, 2009) modeled soil-geogrid
interface elements as thin layer elements, thin shell membrane elements or slip elements
The slip elements are assumed to follow the Mohr-Coulomb criterion in which slip
occurs when the shear stress reaches the yield shear strength at the interface The interface
slip is initiated when the applied stress exceeds the yield strength and the shear stiffness
reduced with a fraction (.

In this study, the pile-soil and geogrid-soil interfaces were modeled as surfaces with
frictional slip elements having their strength equivalent to a percentage of the adjacent
soil shear strength, denoted as Rinterface. The interface elements were 6 nodes triangular

121
elements connecting both sides of pile shell and geogrid elements to the soil tetrahedron
elements. These interface surfaces allow relative movement between piles, geogrid and
soil elements (Brinkgreve et al., 2012).

At the surface backfill layer, the Pile-Soil Rinterface strength is assumed to be 10 % of the
adjacent soil strength up to a depth of 1 m to account for the slippage associated with the
dynamic loading. At the geogrid mesh level, the geogrid-soil interface strength was
assumed equivalent to the strength of the adjacent granular soil. In order to simulate the
real strengthening effect of the geogrid interlocking forces, the shear strength of the soil
adjacent to the geogrid mesh was assigned a value higher than that of the top granular
backfill. The soil strength of at the interface was assigned a shear stiffness value
equivalent to 35 % higher than that of the backfill soil.

This stiffness increase was established from matching the results of the experimental
static pull test of the case where geogrid was embedded in deep backfill with the results
of the numerical model of the deep backfill with an interface layer with thickness equal
to approximately 2 times the average particle size of the aggregates (i.e. 2 x 5 mm) x 20
(scale factor). To achieve the best match between the measured and calculated pile cap
responses, the shear modulus of the interface layer was increased by 35 % of the backfill
shear modulus in order to account for the positive effects of interlocking. Figure 4-18
compares the measured lateral displacements of the pile cap with those calculated from 2
numerical models, one considering the microgrid only and one considering the geogrid
and an interface layer with shear modulus increased by 35%. As can be noted from
Figure 4-19, there is good agreement between the calculated response considering the
interface layer and the measured response. This agreement demonstrated the ability of
the numerical model to reproduce the observed behavior of the geosynthetics-reinforced
pile foundation system. This approach was considered in the rest of the numerical models
in the rest of this chapter as it allowed the size of elements within the interface layer to be
smaller to accommodate the large nonlinearity in this region, without increasing the
number of elements of the entire model significantly.

122
 0.9

0.8

0.7

0.6
Cap pull load (kN)

0.5

0.4

0.3

0.2

0.1

0
0 0.5 1 1.5 2 2.5 3 3.5
Lateral cap deformation (mm)
Plaxis_Microgrid only Plaxis_35% Incresead Interface modulus
Static test _Deep Backfill With Microgrid

Figure 4-19: Pile cap translation, Static pull test vs. 35% improved interface
modulus

4.6.4 Ground response analysis

Free field ground response analyses of the prototype soil model were carried out using
the 3D finite element model developed herein and the ground response analysis software
DeepSoil. A harmonic signal with frequency = 3.6 Hz and an amplitude = 0.1g was used
as the input motion for both models. The equivalent linear analysis method was used in
the DeepSoil analysis. The stiffness degradation and damping curves of the surface
aggregate layer under light confinement stress was obtained from the curves developed
by Rollins et al. (1998). The stiffness degradation and damping curves of the clay were
obtained from Vucetic, and Dobry (1991), while the Seed and Idriss (1975) stiffness
degradation and damping curves of sand were employed for the base sand layer. Figure
4-19 shows the acceleration responses obtained from the DeepSoil and the 3D finite
element models at the surface of the backfill layer. Figure 4-20 shows that there is

123
excellent agreement between the two responses. This excellent agreement demonstrated
the ability of the numerical model to simulate the dynamic behavior of the layered soil
profile. The input parameters used to generate the soil column response are summarized
in Table 4-5. The average soil column natural frequency was calculated as 3.7 Hz which
approximately corresponds to the scaled natural frequency of the model soil column as
explained in section 4.4.

Table 4-5: DeepSoil input parameters; prototype soil column.

Layer # Layer name Thickness (cm) Unit Weight Shear Wave

3
(kN/m ) Velocity (Vsp)
(m/s)

1 Aggregate 1.6 20 220

2 Glyben 4.6 17 170

3 Sand 7.2 20 220

0.30

0.20

0.10
Acceleration (g)

0.00

-0.10

-0.20

-0.30
0.00 0.50 1.00 1.50 2.00 2.50 3.00 3.50 4.00 4.50
Free Field_Plaxis 3D Free Field_DeepSoil

Dynamic Time (sec)

Figure 4-20: Ground response analysis, 3D finite element model vs. DeepSoil.
124
4.7 Parametric Study Considering Dynamic Analysis
A parametric study was conducted using the verified numerical model to evaluate the
influence of different parameters on the performance of geosynthetics-reinforced Pile
foundation system. Table 4.6 summarizes the cases considered.

Table 4-6: Summary of numerical model verification analyses

Numerical Test case description Base Excitation Base Excitation

Case No.
Amplitude Frequency

1 Harmonic excitation, 2.0 m backfill 0.1 g 3Hz

with and without geogrid

2 Harmonic excitation, 2.0 m backfill 0.2g 3Hz

with and without geogrid

3 Harmonic excitation, 2.0 m backfill 0.3g 3Hz

with and without geogrid

4 Harmonic excitation, 2.0 m backfill 0.2g 2Hz

with and without geogrid

5 Harmonic excitation, 2.0 m backfill 0.2g 5Hz

with and without geogrid

6 Harmonic excitation, 6.0 m backfill 0.2g 3Hz

with and without geogrid

7 Harmonic excitation, 1.6 m backfill 0.2g 3Hz

with polymer strips

125
4.7.1 Effect of geogrid reinforcement on response to shaking with
varying intensity
This section investigates the effectiveness of geosynthetics reinforcement for pile
foundations subjected to input motion with different amplitude. The base motion
amplitude was increased from 0.1 to 0.3 g, while the excitation frequency was kept
constant at 3Hz.

Figure 4-21 shows the response calculated at the pile cap due to the applied harmonic
excitation for the cases of 2 m backfill with and without geogrid. The input motion was
set at amplitude = 0.1 g. Figure 4-21 shows that the addition of the geogrid has reduced
the acceleration amplitude by an average of 20% when the response reached a steady
state, which indicates beneficial effect in reducing acceleration, as well as lateral
response of the pile foundation.

0.30
0.25
0.20
0.15
0.10
0.05
Cap acceleration (g)

0.00
-0.05
-0.10
-0.15
-0.20
-0.25
-0.30
-0.35
0.00 0.50 1.00 1.50 2.00 2.50 3.00 3.50 4.00 4.50
Dynamic Time (sec)
Cap Response _Wiithout Geogrid Cap Response_With Geogrid

Figure 4-21: Cap motion comparison, Cases of 2 m backfill with and without
geogrid (0.1 g, 3 Hz).

126
Figure 4-22 shows the response of the pile cap with and without the geogrid
reinforcement subjected to input motion with amplitude = 0.2g. It can be noted from
Figure 4-22 that the geogrid reinforcement reduced the cap acceleration by almost 50 %.
This indicates increased effectiveness for stronger ground input motion. This can be
attributed to the stiffening effect of the geogrid-soil interlock which enhanced the lateral
resistance of the pile cap.

0.70
0.60
0.50
0.40
0.30
0.20
Cap acceleration (g)

0.10
0.00
-0.10
-0.20
-0.30
-0.40
-0.50
-0.60
-0.70
0.00 0.50 1.00 1.50 2.00 2.50 3.00 3.50 4.00 4.50
Dynamic Time (sec)
Cap Response _Without Geogrid Cap Response _With Geogrid

Figure 4-22: Cap motion comparison, Cases of 2 m backfill with and without
geogrid (0.2 g, 3 Hz).

Figure 4-23 shows the cap response to an excitation with amplitude = 0.3g. Similar to the
previous case, the geogrid reinforcement resulted in reduced cap acceleration by almost
30 %. It appears that for much increased input motion, the shear stresses at the geogrid-
soil interface exceeded its maximum shear resistance, hence limiting the favorable effect
of the geogrid to an average of 30 %.

127
 1.00

0.80

0.60

0.40
Cap acceleration (g)

0.20

0.00

-0.20

-0.40

-0.60

-0.80
0.00 0.50 1.00 1.50 2.00 2.50 3.00 3.50

Cap Response_Without Geogrid Cap Response_With Geogrid

Dynamic Time (sec)

Figure 4-23: Cap motion comparison, Cases of 2 m backfill with and without
geogrid (0.3 g, 3 Hz).

4.7.2 Effect of geogrid reinforcement on response of foundation

subjected to shaking with different frequency
The effect of the input motion frequency on the behavior of the pile-cap-geogrid system
was investigated by subjecting the pile foundation system to base motion with two
different frequencies, 2 Hz and 5 Hz g, while keeping its amplitude constant at 0.2 g.
This frequency range represents the typical predominant frequencies of medium to high
frequency earthquakes in North America.

Figures 4-24 and 4-25 show the cap response with and without the geogrid reinforcement
for excitation frequencies 2 Hz and 5 Hz, respectively. Figures 4-24 and 4-25, along with
Figure 4-22 reveal that the geogrid advantageous effect was sustained over the studied
range of frequencies. This confirms the effectiveness of the geosynthetics-strengthened
pile foundation system in a wide range of seismic activities.

128
 0.60

Cap acceleration (g) 0.40

0.20

0.00

-0.20

-0.40

-0.60
0.00 0.50 1.00 1.50 2.00 2.50 3.00 3.50 4.00
Dynamic Time (sec)
Cap Response_Without Geogrid Pile Cap Response_With Geogrid

Figure 4-24: Cap motion comparison, Cases of 2 m backfill with and without
geogrid (0.2 g, 2 Hz).

0.60

0.40

0.20
Cap acceleration (g)

0.00

-0.20

-0.40

-0.60
0.00 0.50 1.00 1.50 2.00 2.50 3.00 3.50 4.00
Dynamic Time (sec)
Cap response_Without geogrid Cap Response _With Geogrid

Figure 4-25: Cap motion comparison, Cases of 2 m backfill with and without
geogrid (0.2 g, 5 Hz).

129
4.7.3 Effect of engineered backfill thickness
Figure 4-26 shows the acceleration time history of the pile cap response for the cases of 6
m backfill without geogrid, 1.6 m backfill with geogrid and 1.6 m backfill without
backfill. It can be noted from Figure 4-26 that both cases of 1.6 m backfill with geogrid
and 6 m backfill without geogrid produced similar acceleration time histories. This
demonstrates that the geogrid-reinforcement can reduce the requirements for ground
improvement significantly, while maintaining the same acceptable performance. This can
represent significant savings in cost and construction time. In contrast, the acceleration
time history of the pile cap for the case of 1.6 backfill without geogrid is almost 40 %
higher than the other two cases. These results confirm the effectiveness of the
geosynthetics reinforcement to improve the seismic performance of pile foundations.

0.80

0.60

0.40
Cap acceleration (g)

0.20

0.00

-0.20

-0.40

-0.60

-0.80
0.00 0.50 1.00 1.50 2.00 2.50 3.00 3.50 4.00 4.50
Dynamic Time (sec)
Cap Response _Without Geogrid_1.6m Backfill
Cap Response _With Geogrid_1.6 m Backfill
Cap Response _ Without Gegrod_6 m Backfill

Figure 4-26: Pile cap acceleration time history comparison , ( cases of 1.6 m with
geogrid, 1.6m without geogrid and 6 m backfill without geogrid).

130
4.7.4 Effect of stiffness of geosynthetics reinforcement on the system
performance
The stiffness of the geosynthetic reinforcement used in the aforementioned sections (500
KN/m) represented the stiffness of typical geogrids. However, polymer strips can have a
stiffness of as much as 44000 kN/m. Additional dynamic analyses were carried out to
investigate the beneficial effect of using polymer strips instead of geogrids as reinforcing
element.

Figure 4-27 compares the responses of a pile cap reinforced with polymer strips and
geogrids when subjected to base excitation with amplitude = 0.2g and frequency = 3Hz.
Figure 4-26 shows that the polymer strips resulted in further reduction of the pile cap
response compared to the conventional geogrid. This additional improvement, however,
should be weighed against the extra cost of using the polymer strips. The final design
should be optimized considering the level of seismic intensity and the specified
performance criterion for the pile foundation.

0.60

0.40

0.20
Cap acceleration (g)

0.00

-0.20

-0.40

-0.60

-0.80
0.00 0.50 1.00 1.50 2.00 2.50 3.00 3.50 4.00
Cap Response_Convetional Geogrid Cap Response_Polymer Strips
Dynamic Time (sec)

Figure 4-27: Effectiveness of conventional geogrid vs. polymer strips in reducing

pile cap response

131
4.8 Parametric Study Considering Pseudo-Static Analysis
This section presents the results of a parametric study carried out to evaluate the
performance of the proposed geogrid-strengthened piled foundation system considering
the typical seismic design loads stipulated by the National Building Code of Canada
(NBCC, 2010). The objective of the design exercise is to optimize the ground
replacement effort (i.e. thickness of engineered backfill).

The seismic loading is given by the pseudo-static inertial force generated for 20 m high
(6 stories) reinforced concrete building. The parametric study investigates the effects of
geogrid depth and stiffness on the lateral pile deflection. The soil profile considered in
this study was modified to reflect a practical case of an engineered backfill and a week
native soil layer.

4.8.1 Building, pile cap model and geogrid

The example building was assumed to be located in Vancouver, BC. The total building
height is 20.0 m covering an area of 22.5 m X 22.5 m. The building is supported on a raft
0.5 m thick and supported by 100 steel piles spaced at 2.5 m centre-to-centre. The piles
were 12 m long, 0.5 m in diameter and had 9.5 mm wall thickness. The pile raft was
supported on the soil surface over an interface mesh which allows maximum free
translation in the direction of loading. For the purpose of reducing the analysis time, only
an area of 4 m X 4 m within the core of the building was modeled. The core of the
building was assumed to support part of the lateral seismic load through two shear walls
in the direction of the seismic loading. Therefore, only one pile group (2X2) was
considered in the analysis.

The soil model boundary and geogrid width were extended 4.5 and 4.25 meters in the
lateral direction beyond the line of symmetry , as previously shown in figure 4-17 and 4-
18 .The piles, pile cap, and soil finite element mesh characteristics were all similar to the
ones modeled in the previous prototype dynamic study. Figures 4-28 and 4-29 show the
vertical shear walls, pile cap and geogrid layout plus the soil profile in 2D and 3D views.

132
Figure 4-28 : Pile, cap, geogrid and superstructure model

133
 Figure 4-29 : Plaxis model 3D view

The method of analysis for the nonlinear soil model was similar to that adopted in the
dynamic prototype study. However, for practical consideration, the shear wave velocity
Vs of the backfill used in this study was increased to 300 m/s and a softer clay layer with
Vs equals 113 m/s was used to simulate a weaker native soil. Table 4-7 summarizes the
HSSMALL model parameters used in this investigation. The soil layers depths were
modified so that the total depth of native clay layer is 6 m and the depth of the bedding
layer is 13 m. The backfill layer replaced the native clay layer in three stages: 2, 4 and 6
m.

134
Table 4-7: HSSMALL Soil parameters of each soil layer adopted for the equivalent
static study

Parameter Symbol Top Backfill Interface Soft clay Dense

layer Layer layer Granular
layer

Soil unit weight (kN/m3) γ 20 20 17 20

Shear wave velocity (m/s) Vs 300 405 113 220

Secant stiffness in 58.8 X 103 107 X 103 2.17 X 31.6 X

standard drained triaxial 103 103
test (kN/m2)

Tangent stiffness for 47.0 X 103 85.7 X 103 1.74 X 25.3 X

primary oedometer 103 103
loading (kN/m2)

Unloading/reloading 176.0X 103 321.0 X 103 6.51 X 94.8 X

2 3
stiffness (kN/m ) 10 103

Power for stress-level m 1 1 1 1

Cohesion 5 5 5 NA

Friction angle φ 40 40 25 40

Reference shear modulus 184.0 X103 335.0 X103 22.1X103 98.8X103

at small strains (kN/m2)

Poisson’s ratio 0.2 0.2 0.2 0.2

There were two types of geosyenthetics considered in the pseudo-static analysis. The first
is the polymer strips with stiffness of 44,000 kN/m. The second is a conventional geogrid
135
with a stiffness of 2900 kN/m at 2% strain. The conventional geogrid stiffness was taken
from the technical data sheet of TMP Geosyenthetics – Uniaxial Geogrid GG200PE.
Table 4-8 shows the specification of the conventional geogrid.

Table 4-8 : Uniaxial Geogrid GG200PE specifications

Index Properties Test Method Units MD Value

Polymer - - HDPE

Minimum Carbon Black ASTM D 4218 % 2

Tensile Strength @ 2% Strain ASTM D 6637 kN/m (lb/ft) 58 (3.970)

Tensile Strength @ 5% Strain ASTM D 6637 kN/m (lb/ft) 116 (7,950)

Ultimate Tensile Strength ASTM D 6637 kN/m (lb/ft) 200(13,700)

Strain @ Ultimate Strength ASTM D 6637 kN/m (lb/ft) 11.5

Junction Efficiency GRO GG2-87 kN/m (lb/ft) 90

4.8.2 Pseudo-static seismic lateral loading calculation

The equivalent seismic base shear has been calculated as per the NBCC (2010)
provisions. The site was assumed to be located in Vancouver area with soft soil and site
Class E classification (NBCC 2010, Table 4.1.8.4.A). The strength level design base
shear is given by:

V= W (NBCC 4.1.8.11) [18]

Where

 The Fundamental period of the structure in seconds Ta = 0.05 (hn)3/4 = 0.5 sec
(CI.4.1.8.11.3(c)).

136
  For Vancouver , the 5% damped spectral response acceleration ratios, Sa(T), are
provided in Table-C2 of NBCC2010 as :
Sa(0.2) Sa(0.5) Sa(1.0) Sa(2.0) PGA
0.95 0.65 0.34 0.17 0.47

 The design spectral acceleration value S(T):

From Table 4.1.8.4.b, the value of acceleration-based site coefficient, Fa = 0.9.

From Table 4.1.8.4.c, the value of velocity-based site coefficient, Fv = 1.8.

For T=0.5 sec, S(T) = Fv . Sa(0.5) or S(T) = Fa. Sa(0.2), Smallest S(T) = 0.855.

Therefore, for T=0.5 sec, S(T) = 0.855

 From Table 4.1.8.11, Sa(0.2)/Sa(2.0) = 0.95/0.17 = 5.6 ˂ 8.0, Therefore higher

mode factor Mv = 1.0.

 The seismic importance factor was set at 1.5.

 From Table 4.1.8.9, and considering the building has limited ductility shear walls,
Rd = 1.5, Ro = 1.5.

 Seismic dead load W = 15220 kN.

 Total lateral seismic force is obtained from Eq. [18].

V= 8726.13 kN

The calculated base shear force was applied at the vertical centroid of the distributed
lateral floor forces. The vertical centroid was calculated from the division of the total
seismic overturning moment by the total base shear. Table 4.9 shows the distribution of
the lateral earthquake forces Fi and determines the total overturning moment and the
vertical centroid.

137
Table 4-9: Equivalent static forces and overturning moment calculation
Floor
height W(kN) wiXhi LTFi*(kN) LFi**(kN) OM***(KN.m)
20 3961.25 79225.00 3546.41 141.85 0
15 3961.25 59418.75 2659.80 106.39 709.28
10 3961.25 39612.50 1773.20 70.92 1950.52
5 3336.25 16681.25 746.72 29.87 3546.40
Total 15220.00 194937.50 8726.13 349.05 5291.63
* Total lateral force over 100 piles
**Lateral force over 4 piles
***Overturning moment over 4
piles

The vertical centroid = 5291.63/349 = 15.16 m

4.8.3 Summary of numerical pseudo-static analyses

This section describes the cases investigated in this parametric study. The equivalent
static load was applied to the model geogrid-strengthened foundation system considering
various soil profiles and different configurations for the geosyenthetics reinforcement
including: embedment depth and stiffness. Table 4.10 summarizes the different cases
analyzed.

138
Table 4-10: Equivalent static analysis cases

Case No. case description

1 Native soil, no backfill without geosynthetic reinforcement

2 2 m backfill without geosynthetic reinforcement

3 4.0 m backfill without geosynthetic reinforcement

4 6.0 m backfill without geosynthetic reinforcement

5 2.0 m backfill with one high tensile geogrid mesh placed at 1.0 m depth

6 2.0 m backfill with one high tensile geogrid mesh placed at 1.25 m depth

7 2.0 m backfill with two high tensile geogrid mesh placed at 1.0 and 1.5 m
depth
8 2.0 m backfill with one polymer strips placed at 1.25 m depth

4.9 Results and Discussion

The results of the parametric study are presented and discussed in this section. The
results are discussed with respect to the effect of geosynthetic material stiffness, depth
and length on: the building maximum lateral displacement, pile head maximum
deflection, pile maximum bending moment and pile maximum shear. Figure 4-30 depicts
the deformed shape of the building, pile cap and piles as a result of the equivalent seismic
loading.

139
 Figure 4-30: Deformed model shape

4.9.1 Effect of geosynthetic material stiffness and depth

Table 4-11 provides the results of cases 1-8 with respect to building maximum lateral
drift, pile head maximum deflection, pile maximum bending moment and pile maximum
shear.

140
Table 4-11: Cases 1-8 results comparison.
Max pile cap Max pile cap Max pile Max pile
Case No. vertical lateral bending shear
deflection deflection moment (kN/m)
(mm) (mm) (Nm/m)

1 19.4 19.6 209 208.0

2 17.4 14.3 215.7 268.3

3 16.1 12.2 265.7 279.2

4 15 11.3 259.4 272.6

5 16 11 146 285

6 16.2 12 135 260

7 15.7 10.4 150.4 272.3

8 16.3 9.4 133.1 291

It can be noticed from the results of cases 1-3 that the pile cap lateral deflection decreases
as the conventional backfill depth increases. Also, the results show that increasing the
backfill induced higher bending moment and shear force in the piles.

Cases 4 and 5 compare the 6 m plain backfill results to the results of the 2 m backfill with
high tensile geogrid embedded at 1.0 m depth. It can be noticed that embedding the high
tensile geogrid at 1.0 depth within 2m of backfill resulted in the same performance as the
case of 6m backfill, while reducing the bending moment of the piles by 44%. However,
the shear force increased by 5 %. This comparison indicates that by using the geogrid, the
backfill depth can be reduced by 67% while achieving the same reduction in the lateral
displacement (i.e. 44 % reduction in lateral displacement).

141
Comparing the results of Cases 5 and 6 demonstrates the effect of the geogrid layers
depth on the performance of geosynthetics-reinforced pile foundation system. It can be
noticed that embedding the geogrid mesh at depth of 1.25 m has reduced the bending
moment and shear forces while improving the serviceability of the foundation system.
Case 6 demonstrates that by using the geogrid reinforcement, the backfill depth can be
reduced by 67% (compared to Case 3) while achieving 39 % improvement in lateral pile
foundation performance relative to the native soil case.

In addition, the comparison of Cases 6 and 7 indicates that the lateral performance of the
foundation can be improved further by increasing the number of geogrid layers. Also, the
comparison of Cases 1, 4 and 7 indicates that through using 2 layers of geogrid, the
backfill depth can be reduced by 67% while achieving 47 % improvement in the lateral
pile foundation performance and reducing the bending moment and shear forces in the
piles.

Moreover, comparing Cases 4, 6 and 8 indicates that the lateral performance of the pile
foundation with polymer strips embedded in 2 m backfill was better than that of the pile
foundation system with 6.0 m backfill and the 2 m backfill enhanced with one geogrid
layer. Using the polymer strips, the backfill depth can be reduced by 67 % while
achieving 52 % improvement in the lateral pile foundation performance. In addition,
using the polymer strips as reinforcement induced the least bending moment, but polymer
strips induced bigger shear force in the piles. This behavior can be attributed to the large
stiffness of the polymer strips resulting in high lateral force at the pile shell nodes.

4.10 Summary and Conclusions

This chapter presents the finite-element (FE) model for simulating the dynamic
performance of geogrid-reinforced pile foundation system. A numerical dynamic model
of geogrid-reinforced pile foundation system was calibrated using the results of the
experimental dynamic model. A parametric study was carried out to study the effect of
the base motion amplitude and frequency on the dynamic behavior of the geogrid piled
foundation system. The parametric study also evaluated the efficiency of pile foundation-

142
geogrid against the performance of a conventional pile foundation with ground
replacement (i.e. deep engineered backfill). The following conclusions may be drawn
from the results:

 The numerical results compared well with the experimental results demonstrating
that the HSSMALL model was able to simulate the nonlinear stress-strain
behavior of the experimental soil bed.

 The numerical results revealed that embedding the geogrid mesh has enhanced
the lateral performance of the pile foundation system and reduced its acceleration
response.

 The geogrid reinforcement favorable effect was observed for a range of base
shaking frequencies and amplitudes.

 Increasing the geogrid stiffness further enhanced the lateral performance of the
pile foundation. However, the results indicated that conventional geogrid could
replace high stiffness geosynthetic grids as the performance using both materials
was comparable.

 The dynamic numerical parametric study suggested that it is possible to reduce

the depth of granular backfill by using geogrid while achieving improved lateral
performance of the pile foundation system.

 The pseudo-static analysis showed that, as expected, the lateral performance of

the pile foundation was improved as the thickness of the conventional backfill
increased. However, increasing the backfill thickness induced higher bending
moment and shear force in the piles.

 The pseudo-static analysis indicated that embedding the high tensile geogrid in
2m thick engineered backfill at 1.0 m depth reduced resulted in almost the same
performance of the pile foundation with 6 m backfill. This means using the
143
 googrid reinforcement can reduce the backfill by 67% while achieving the same
improved performance (i.e. 44% reduction in lateral displacement and pile
bending moment). However, the shear force increases by 5 %.

 Embedding the geogrid mesh at a depth of 1.25 m, within 2m backfill, have

reduced the bending moment and shear forces of the piles while improving the
performance of the foundation system (39 % reduction in lateral pile foundation
displacement).

 The lateral performance of the pile foundation can be improved further by adding
another geogrid layer. It was found that using 2 layers of geogrid, the backfill
thickness can be reduced by 67% while achieving 47 % reduction in lateral
displacement of the pile foundation, and reduced bending moment and shear force
in the piles.

 The lateral performance of the pile foundation reinforced with polymer strips
embedded in 2 m thick backfill was better than the performance of the pile
foundation with 6m thick backfill and that with the 2 m backfill strengthened with
one and two geogrid layers.

144
References:
[1] Angelides, D. C., and Roesset, J. M. 1980. Nonlinear dynamic stiffness of piles.
Department of Civil Engineering, Massachusetts Institute of Technology,
Cambridge, Maass. Research Report R80-13

[2] Bathurst RJ, Alfaro MC. Review of seismic design, analysis and performance of
geosynthetic rein- forced walls, slopes and embankments. International Symposium
on Earth Reinforcement, keynote lecture, Kyushu, 1996. p. 23-52.

[3] Bentley , K.J., and El Naggar, M.H. 2000, Numerical analysis of kinematic
response of single piles, Canadian Geotechnical Journal, 37:1368-1382.

[4] Brinkgreve, R., Egin, E. and Swolfs, W. User’s Manual for Plaxis 3D 2012 , Delf
University of Technology & PLAXIS bv, The Netherlands.

[5] Cai, Z., and Bathurst, R.J. 1995. Seismic response analysis of geosynthetic
reinforced soil segmental retaining walls by finite element method. Computers and
Geotechnics, 17 (4), 523-546.

[6] Christopher Burke C, Hoe I. Ling and Huabei Liu, 17th ASCE Eng Mech Conf,
June 13-16, Univ of Delaware, Newark, DE, 2004 SEISMIC RESPONSE
ANALYSIS OF A FULL-SCALE REINFORCED SOIL RETAINING WALL.

[7] El-Emam, M.M., and Bathurst, R.J. 2007. Influence of reinforcement parameters
on the seismic response of reduced-scale reinforced soil retaining walls.
Geotextiles and Geomembranes, 25 (1), 33-49.

[8] Fakharian, K., and Attar, I.H. 2007. Static and seismic numerical modeling of
geosynthetic-reinforced soil segmental bridge abutments. Geosynthetics
International, 14 (4), 228-243.

[9] Faruque, M. O., and Desai, C.S. 1983. 3-D material and geometric nonlinear
analysis of piles. In Proceedings of the 2nd International Conference on Numerical
145
 Methods in Offshore Piling, The Univesity of Texas at Austin, Austin, Tex. pp.
553-575.

[10] Hardin, B.O., and Drnevich, V.P. (1972). Shear modulus and damping in soils:
Design equations and curves. Proceedings. ASCE: Journal of Soil Mechanics and
Foundations Division, 98(SM7), 667-692.

[11] Hashash, Y., Park, D.,Tsai, C., and Groholski, D. 1-D Wave Analysis Program for
Geotechnical Site Response Analysis of Deep Soil Deposits. 2002.V5.0. University
of Illinois at Urbana-Champaign.

[12] Helwany, S.M.B., Budhu, M., and McCallen, D. 2001. Seismic analysis of
segmental retaining walls. I: Model verification. Journal of Geotechnical and
Geoenvironmental Engineering, 127 (9), 741-749.

[13] Ju, S. H., and Ni, S. H. 2007. Determining Rayleigh damping parameters of Soils
for finite element analysis. International journal of numerical and analytical
methods in geomechanics, 31, 1239-1255.

[14] Ling, H.I., Liu, H., Kaliakin, V.N., and Leschchinsky, D. 2004 .Analyzing dynamic
behavior of geosynthetic-reinforced soil retaining walls. Journal of Engineering
Mechanics, 130 (8), 911-920.

[15] Liu, H. 2009. Analyzing the reinforcement loads of geosynthetic-reinforced soil

walls subject to seismic loading during the service life. Journal of Performance of
Constructed Facilities, 23 (5), 292-302.

[16] Michalowski R L. 1998. Soil reinforcement for seismic design of geotechnical

structures. Computers and Geotechnics, 23 (1-2), 1-17.

[17] Randolph, M. F. 1981, Response of flexible piles to lateral loading. Geotechnique,

31(2), 247-259.

146
[18] Rao, S. S., 2005, The finite element method in engineering, Elsevier Butterworth
Heinemann, Amsterdam.

[19] Rollins, K. M., Evans, M. D., Diehl, N. B., and Daily, W. D. 1998. Shear Modulus
and Damping Relations for Gravels. Journal of Geotechnical and
Geoenvironmental Engineering, 124 (5), 396–405.

[20] Seed, H. B. and Idriss, I. M. Soil moduli and damping factors for dynamic response
analysis, report EERC 70-10, Earthquake Research Center, University of
California, Berkeley; 1969.

[21] Surarak, C., Likitlersuang S., Wanatowski D., Balasubramaniam A., Oh E., and

Guan H. 2012. Stiffness and strength parameters for hardening soil model of soft

and stiff Bangkok clays. Soils and Foundations, 52 (4), 682-697.

[22] Trochanis, A., Bielk, J., and Christiano, P. 1988. A three dimensional nonlinear
study of piles leading to the development of a simplified model. Department of
Civil Engineering, Carnegie Institute of Technology. Report R-88-176.

[23] Trochanis, A.M., Bielak, J., and Christiano, P, 1991, Simplified model for analysis
of one or two piles. Journal of Geotechnical Engineering, ASCE, 117(3),448-466.

[24] Turan, A., Hinchberger, S.D., and El Naggar, M.H. 2009. Design and
commissioning of a laminar soil container for use on small shaking tables. Soil
Dynamics and Earthquake Engineering, 29 (2), 404-414.

[25] Turan, A., Hinchberger, S.D., and El Naggar, M.H. 2009. Mechanical
characterization of an artificial clay. Journal of Geotechnical and
Geoenvironmental Engineering, 135 (2), 280-290.

[26] Vucetic, M., and Dobry, R. 1991. Effect of soil plasticity on cyclic response.
Journal of Geotechnical Engineering, ASCE, 117(1),89-107.

147
[27] Wu, G., and Finn, W, 1997. Dynamic nonlinear analysis of pile foundations using
finite element method in the time domain, Canadian Geotechnical Journal, 34(1),
44-52.

[28] Yegian, M., and Wright, S. 1973. Lateral soil resistance displacement relationships
for pile foundations in soft clays. In Proceedings of the 5th Offshore Technology
Conference, Houston. Tex, Vol.2, pp. 663-676. OTC Paper 1893.

148
 Chapter 5

5. Summary and Recommendations for Further Research

This thesis investigated the effect of geogrid reinforcement on the lateral response of pile
foundations. The investigation was carried out on two model superstructure mounted on a
model piles-cap foundation and subjected to 1-G shaking on a shaking table. The
research also investigated the static lateral response of the geogrid-piled foundation
composite system through subjecting the foundation cap to a static lateral pull. The
results derived from the composite system are compared with the base case where the
model piled foundation with ground replacement using thick engineered backfill is not
reinforced with geogrid and another case where the granular backfill is reduced
significantly. The thesis presents static and dynamic numerical models of the piled-
geogrid reinforcement composite system that could be used for the analysis and design.
The numerical models were verified against the results of the reduced scaled model
tested on 1-G shaking table.

5.1 Summary of thesis findings

The following conclusions may be drawn from the thesis findings.

Considering the configuration of the foundation system investigated in this study, and
based on the results of reduced scale physical pull tests and corresponding static
numerical analyses:
 The lateral resistance of the foundation system was increased by 15 % due to the
addition of the model microgrid mesh.
 The lateral resistance of the foundation system decreased by 20 % due to reducing
the thickness of the surficial granular backfill layer.
 The numerical results show that the addition of the polymer strips would allow
reducing the required soft ground replacement by 50%, while providing improved
lateral performance.

149
  The numerical results showed that the addition of the geosynthetic reinforcement
reduced the bending moment by 8% and reduced the shear force by 2%.
 The numerical results also showed that extending the geosynthetic reinforcement
increased the lateral stiffness of the foundation system and reduced the lateral
displacement by 12 %.
 The numerical results indicated that extending the microgrid mesh farther from
the pile foundation bending decreased the bending moment by around 5%.
 The parametric study indicated that the extent of the geosynthetic reinforcement
influences the level of enhancement it provides. Thus, the length of the
reinforcement should be optimized for the specific case considered.
 Overall, the model scale experimental and numerical results showed that the
beneficial effects of the geosynthetic reinforcement increased as the applied load
increased. Thus, further improvement of the lateral performance of the
geosynthetic-reinforced foundation is expected if the foundation is allowed to
experience larger displacement.

The results of a series of reduced scale shaking table tests performed to study the
influence of a model polymer strips, the microgrid, reinforcement on the dynamic
response of a SDOF superstructures and a supporting piled foundation are summarized
herein. The experimental results presented here provide physical evidence on the
effectiveness of the proposed foundation concept and can be used to calibrate numerical
models. Following are a summary of the conclusions from this experimental study;

 As expected, increasing the thickness of the backfill layer reduced the ground
motion amplification and the maximum dynamic response of the soil column to
harmonic loading. The addition of the microgrid mesh further reduced the lateral
response by 23%.
 The scaled earthquake and sine sweep tests indicated that the microgrid
reinforcement resulted in reducing lateral response of the pile foundation, even
for the case with thicker engineered backfill. The maximum lateral response of
the pile cap decreased farther for the case of the geosynthetics-reinforced backfill.
150
  The lateral response of the SDOF systems to the scaled earthquake and sine
sweep tests decreased for the case of microgrid-reinforced backfill. The sine
sweep tests results indicated that the thick backfill reduced the maximum lateral
response of the low frequency SDOF structure and the microgrid reinforcement
reduced it further. The effectiveness of the microgrid reinforcement in reducing
the dynamic response was more pronounced at higher inertial interaction
associated with the low frequency SDOF due to the larger dynamic loads
associated with larger deformations in the microgrid.
 The rocking vibrations of the pile cap due to the sine sweep tests and harmonic
loading were reduced due to the geosynthetics reinforcement. The sine sweep
tests revealed that increasing the backfill thickness from 3 cm to 8 cm reduced the
peak cap rocking motion by 15 % and the addition of microgrid reinforcement
further reduced the peak rocking motion by 9 %.

Finite element analyses were carried out to simulate the dynamic performance of
geogrid-reinforced pile cap system. A numerical dynamic model of geogrid-reinforced
pile cap system was calibrated against the experimental dynamic model. A parametric
study was carried out to study the effect of the base motion amplitudes and frequencies
on the dynamic behavior of the geogrid piled foundation system. The study also included
comparing the efficiency of pile cap-geogrid embedded in relatively shallow backfill
against the performance of a conventional pile foundation embedded in deep backfill
using pseudo-static analysis. The following points present the summary of results:

 The numerical results compared well with the experimental results demonstrating
that the HSSMALL model was able to simulate the nonlinear stress-strain
behavior of the experimental soil bed.

 The numerical results revealed that embedding the geogrid mesh has enhanced
the lateral performance of the pile foundation system and reduced its acceleration
response.

151
 The geogrid reinforcement favorable effect was observed for a range of base
shaking frequencies and amplitudes.

 Increasing the geogrid stiffness further enhanced the lateral performance of the
pile foundation. However, the results indicated that conventional geogrid could
replace high stiffness geosynthetic grids as the performance using both materials
was comparable.

 The dynamic numerical parametric study suggested that it is possible to reduce

the depth of granular backfill by using geogrid while achieving improved lateral
performance of the pile foundation system.

 The pseudo-static analysis showed that, as expected, the lateral performance of

the pile foundation was improved as the thickness of the conventional backfill
increased. However, increasing the backfill thickness induced higher bending
moment and shear force in the piles.

 The pseudo-static analysis indicated that embedding the high tensile geogrid in
2m thick engineered backfill at 1.0 m depth reduced resulted in almost the same
performance of the pile foundation with 6 m backfill. This means using the
googrid reinforcement can reduce the backfill by 67% while achieving the same
improved performance (i.e. 44% reduction in lateral displacement and pile
bending moment). However, the shear force was increases by 5 %.

 Embedding the geogrid mesh at a depth of 1.25 m, within 2m backfill, have

reduced the bending moment and shear forces of the piles while improving the
performance of the foundation system (39 % reduction in lateral pile foundation
displacement).

 The lateral performance of the pile foundation can be improved further by adding
another geogrid layer. It was found that using 2 layers of geogrid, the backfill
thickness can be reduced by 67% while achieving 47 % reduction in lateral

152
 displacement of the pile foundation, and reduced bending moment and shear force
in the piles.

 The lateral performance of the pile foundation reinforced with polymer strips
embedded in 2 m thick backfill was better than the performance of the pile
foundation with 6m thick backfill and that with the 2 m backfill strengthened with
one and two geogrid layers.

5.2 Recommendations for Future Research

It is recommended to evaluate the performance of a full-scale geogrid-reinforced piled
foundation system. A small group of piles, 2X2, can be driven in soft soil and lateral
dynamic and static forces can be applied to the pile cap. The lateral displacements of the
pile cap and the deformations in the piles can be recorded with sensitive accelerometers,
LVDT’s and strain gages. The surface clay layer can then be removed in stages and
replaced with engineered backfill. The foundation can be tested for each backfill case
with the same loading conditions and instruments. For each backfill case, a high tensile
geogrid mesh can be embedded within the backfill layer and the testing procedure can be
repeated to evaluate the effect of geogrid reinforcement. The data gathered from the field
tests can be used for numerical model calibration. The calibrated model can be used to
execute a parametric study that establishes a detailed design guidelines and procedure for
this novel foundation system.

153
 Appendix A

Direct shear tests results:

Mohr Coulomb Envelope for Sand

300

250
Shear Strength (kPa)

200
y = 0.8284x
150

100

50

0
0 50 100 150 200 250 300 350
Normal Stress (kPa)

Mohr Coluomb Envelpe for Aggregate

250

200
Shear Strenght (kPa)

y = 1.4186x
150

100

50

0
0 20 40 60 80 100 120 140 160

Normal stress (kPa)

154
 Curriculum Vitae

Name: Ahmed Mohammed A Taha

Post-secondary King Fahd University of Petroleum and Minerals

Education and Dhahran, Kingdom of Saudi Arabia

Degrees: 1992-1997 B.Sc.

The University of Ottawa

Ottawa, Ontario, Canada 2008-2010 M.Sc.

The University of Western Ontario

London, Ontario, Canada 2010-2013 Ph.D.

Related Work The Saudi Arabian Oil Company

Experience: Project Engineer 1998-2008

Teaching Assistant

The University of Western Ontario 2012-2013

Publications:

Taha A, and Fall M, (2013) “ Shear Behavior of Sensitive Marine Clay-Concrete

Interfaces” Journal of Geotechnical and Geoinvironmental Engineering, Vol 139, No. 4,
PP: 644-650.

155