Citation
Facing of geosynthetic reinforced soil structures

Material Information

Title:
Facing of geosynthetic reinforced soil structures
Creator:
Beauregard, Melissa Stewart ( author )
Language:
English
Physical Description:
1 electronic file (205 pages) : ;

Subjects

Subjects / Keywords:
Soil stabilization ( lcsh )
Geosynthetics ( lcsh )
Geosynthetics ( fast )
Soil stabilization ( fast )
Genre:
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )

Notes

Review:
A study was undertaken to explore the role of facing in Geosynthetic Reinforced Soil (GRS) structures from a variety of perspectives and to develop a new design protocol that considers environmental and aesthetic considerations alongside traditional strength and service performance metrics. This topic of facing is an important one as industry moves closer to accepting the inherent differences between internally-stabilized GRS structures and externally-stable geosynthetic mechanically stabilized earth (GMSE) structures. While construction materials and methods are similar for the two design options, a major difference is that the facing elements for GRS are not structures and exist mostly to aid in construction and to prevent sloughing. GMSE structures require a strong connection between reinforcement and facing for the system to remain stable.
Review:
To achieve the goal of a more holistic design protocol, the following three tasks were completed:
Review:
1) Establish the role of facing on load-deformation behavior of GRS structures: The role of facing was established by: a) summarizing existing results from full-scale, laboratory-scale, and analytical models; and b) developing a new experimental approach to determine how facing stiffness during and post-construction can affect behavior.
Review:
Major conclusions: The load-deformation response of GRS structures differ greatly from that of GMSE. With or without facing, structures built with spacing less than 0.3 meters exhibits a stiffer response than GMSE structures with similar geometry.
Review:
2) Define sustainability of GRS structures: A new Life Cycle Assessment (LCA) framework was developed for comparing GRS and GMSE structures. Additionally, the feasibility of incorporating recycled materials into GRS facing is presented.
Review:
Major conclusions: The flexibility in GRS design leads to opportunity for engineers to create a more sustainable end product, not just in materials used but also in required equipment and construction methods.
Review:
3) Establish relevant aesthetic considerations: Both strictly aesthetic and functional aesthetic considerations are presented. Recommendations are made based on the location of the structure (i.e. natural, urban, transportation corridor).
Review:
Major conclusions: A wider variety of facing materials results in an aesthetic treatment specific to the location of a wall instead of proprietary systems not catered to individual projects. Additionally, there is opportunity to incorporate some functionality to the aesthetic, specifically in highway corridors.
Review:
The end result of this study is a new design method that includes LRFD factors calibrated to an existing design method, a method for quantifying the environmental impact of a GRS structure, and framework for determining the most appropriate facing treatment.
Thesis:
Thesis (Ph.D.)-University of Colorado Denver
Bibliography:
Includes bibliographic references
System Details:
System requirements: Adobe Reader.
Statement of Responsibility:
by Melissa Stewart Beauregard.

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
954003874 ( OCLC )
ocn954003874
Classification:
LD1193.E553 2016d B43 ( lcc )

Downloads

This item has the following downloads:


Full Text
FACING OF GEOSYNTHETIC REINFORCED SOIL STRUCTURES
by
MELISSA STEWART BEAUREGARD
B.C.E., University of Delaware, 2010
M.S., University of Colorado Boulder, 2012
A thesis submitted to the
Faculty of the Graduate School of the
University of Colorado in partial fulfillment
of the requirements for the degree of
Doctor of Philosophy
Engineering and Applied Science
2016


This thesis for the Doctor of Philosophy degree by
Melissa Stewart Beauregard
has been approved for the
Engineering and Applied Science Program
by
Caroline Clevenger, Chair
Jonathan T.H. Wu, Advisor
Peter Hoffman
Mohsen Tadi
Melanie Short


Beauregard, Melissa, Stewart (Ph.D., Engineering and Applied Science)
Facing of Geosynthetic Reinforced Soil Structures
Thesis directed by Professor Jonathan T.H. Wu
ABSTRACT
A study was undertaken to explore the role of facing in Geosynthetic Reinforced Soil
(GRS) structures from a variety of perspectives and to develop a new design protocol that
considers environmental and aesthetic considerations alongside traditional strength and
service performance metrics. This topic of facing is an important one as industry moves
closer to accepting the inherent differences between internally-stabilized GRS structures
and externally-stable geosynthetic mechanically stabilized earth (GMSE) structures.
While construction materials and methods are similar for the two design options, a major
difference is that the facing elements for GRS are not structures and exist mostly to aid in
construction and to prevent sloughing. GMSE structures require a strong connection
between reinforcement and facing for the system to remain stable.
To achieve the goal of a more holistic design protocol, the following three tasks were
completed:
1) Establish the role of facing on load-deformation behavior of GRS structures: The
role of facing was established by: a) summarizing existing results from full-scale,
laboratory-scale, and analytical models; and b) developing a new experimental approach
to determine how facing stiffness during and post-construction can affect behavior.
in


Major conclusions: The load-deformation response of GRS structures differ greatly
from that of GMSE. With or without facing, structures built with spacing less than 0.3
meters exhibits a stiffer response than GMSE structures with similar geometry.
2) Define sustainability of GRS structures: A new Life Cycle Assessment (LCA)
framework was developed for comparing GRS and GMSE structures. Additionally, the
feasibility of incorporating recycled materials into GRS facing is presented.
Major conclusions: The flexibility in GRS design leads to opportunity for engineers to
create a more sustainable end product, not just in materials used but also in required
equipment and construction methods.
3) Establish relevant aesthetic considerations'. Both strictly aesthetic and functional
aesthetic considerations are presented. Recommendations are made based on the location
of the structure (i.e. natural, urban, transportation corridor).
Major conclusions'. A wider variety of facing materials results in an aesthetic
treatment specific to the location of a wall instead of proprietary systems not catered to
individual projects. Additionally, there is opportunity to incorporate some functionality to
the aesthetic, specifically in highway corridors.
The end result of this study is a new design method that includes LRFD factors
calibrated to an existing design method, a method for quantifying the environmental
impact of a GRS structure, and framework for determining the most appropriate facing
treatment.
The form and content of this abstract are approved. I recommend its publication.
Approved: Jonathan T. H. Wu
IV


DEDICATION
I dedicate this work to my wife, Virginia, who encourages me daily to do what makes me
happy.
v


ACKNOWLEDGEMENTS
First and foremost, I would like to thank my advisor, Dr. Jonathan T.H. Wu, for
allowing me to be a part of this important body of work. I could not have asked for a
more knowledgeable and supportive mentor on this academic journey. I would also like
to thank Peter Hoffman always being available for questions, regardless of the hour, and
for injecting a good deal of passion into this work. Thank you as well to the remainder of
my committee: Caroline Clevenger, Mohsen Tadi and Melanie Short for lending their
time and expertise to this research effort.
The experimental portion of this study would not have been possible if it werent for
the help of two visiting undergraduate students, Tiago Bueno and Felipe Cardoso, who
graciously spent their last summer before returning to Brazil in the geotechnical lab at the
University of Colorado assisting me.
I am eternally grateful to the National Science Foundation, the Geo-Synthetics
Institute, and the Department of Civil Engineering for the financial support necessary to
complete this work.
Finally, I would like to thank my father, Gary Stewart, for being excited for all of my
successes and unwaveringly optimistic about the occasional failure.
vi


TABLE OF CONTENTS
Chapter
1. Introduction.................................................................1
1.1 Problem Statement............................................................1
1.2 Research Objectives..........................................................6
1.3 Methods of Research..........................................................6
2. Literature Review............................................................9
2.1 History of Reinforced Soil..................................................9
2.2 Current Approach to Design.................................................11
2.3 GRS Bridge Abutments.......................................................18
2.4 Case Studies of GRS Structures.............................................20
2.4.1 Japan Full-Height Rigid...................................................20
2.4.2 Grand County, CO..........................................................22
2.4.3 Fox Wall, Denver, CO......................................................24
2.4.4 Aluminum Faced Wall, Canada...............................................27
2.4.5 Tongass National Forest, Alaska...........................................29
2.5 Role of Facing in GRS Performance.........................................30
2.6 Analytical Approach to GRS Behavior........................................40
2.6.1 Pham, 2009 ...............................................................40
2.6.2 Hoffman, 2015.............................................................44
2.6.3 Summary of Behavior.......................................................47
2.7 Load and Resistance Factor Design..........................................49
2.7.1 Allowable Stress Design versus Load and Resistance Factor Design..........49
vii


2.7.2 LRFD in Geotechnical Engineering............................................51
3. Aesthetics of Reinforced Soil Facing............................................54
3.1 Current State of Practice.....................................................54
3.2 Human Impacts and Aesthetics..................................................56
4. Sustainability of GRS..........................................................70
4.1 Effect of Recycled Content on Concrete Blocks.................................70
4.2 Wu and Payeur Analysis........................................................73
4.2.1 Results......................................................................76
4.3 Additional Recycled Materials.................................................78
4.4 Comparative Life Cycle Assessment.............................................80
4.4.1 Background...................................................................80
4.4.2 Goal and Scope..............................................................85
4.4.3 Life Cycle Inventory........................................................94
4.4.4 Impact Assessment...........................................................97
4.4.5 Sensitivity Analysis.......................................................103
4.4.6 Results, Discussion, Analysis..............................................106
5. A New Analytical Approach to GRS Behavior.....................................108
5.1 Jewell-Milligan Model.........................................................108
5.2 Modified Jewell-Milligan Model...............................................116
5.2.1 Verification................................................................119
5.2.2 Sensitivity Analysis........................................................128
6. A New Experimental Assessment of the Role of Facing...........................135
6.1 Apparatus....................................................................136
6.1.1 Facing......................................................................139
viii


6.1.2 Loading...............................................................140
6.1.3 Instrumentation.......................................................140
6.2 Materials...............................................................140
6.2.1 Backfill..............................................................140
6.2.2 Geosynthetic..........................................................142
6.3 Procedure...............................................................142
6.3.1 Compaction............................................................142
6.3.2 Test Procedure........................................................144
6.4 Results.................................................................145
6.4.1 Rigid Facing..........................................................145
6.4.2 Flexible Facing.......................................................146
6.4.3 Comparison............................................................150
6.5 Conclusions.............................................................152
6.6 Future Work.............................................................153
7. Load and Resistance Factor Design for GRS-NLB............................154
7.1 Background..............................................................154
7.2 Development of Design Equations.........................................156
7.3 LRFD Design Method for GRS-NLB.........................................160
8. Concluding Remarks.......................................................178
8.1 Summary.................................................................178
8.2 Conclusions.............................................................179
References..................................................................180
Appendix A..................................................................186
ix


LIST OF TABLES
Table
2.1: Comparison of measured to calculated failure loads for timber-faced GRS structure
in plane strain testing device...................................................12
2.2: Summary of results from series of tests involving facing effects for varying
reinforcement spacing (Nicks et al., 2013).......................................35
2.3: Comparison of analytical versus experimental results for ultimate strength of GRS
structure (Pham, 2009)........................................................43
2.4: Recommended Resistance Factors for Geosynthetic Reinforced Structures..........50
4.1: Parameters used for recycled concrete block analysis (Wu and Payeur, 2014).....76
4.2: Total material requirements for GMSE and GRS abutments......................94
4.3: Emissions into air for production of PET and PP resin (LCI, NREL)...........95
4.4: Emissions into air for production of aggregate (Marceau et al., 2007).......96
4.5: Emissions into air for production of abutment facing (Marceau et al., 2007).97
4.6: Lost or damaged CMU blocks required for GRS to have same impact as GMSE
facing.......................................................................106
5.1: Summary of test conditions for GCSC tests (Pham, 2009).........................124
5.2: Parameters for modeled structure used in sensitivity analysis..................128
6.1: Overview of Testing Conditions.................................................135
6.2: Summarized results from testing program....................................152
7.1: Some recommended factors of safety for ASD methods (AASHTO, 1996)..............155
7.2: Some recommended LRFD factors for geotechnical applications................155
7.3: Recommended resistance factors for relevant strength and service limits states.155
x


7.4: Factors of safety recommended by Wu (2012) for GRS.......................159
7.5: Values for k-factor to determine required strength of geosynthetics......159
7.6: Calibrated resistance factors to be used with GRS-NLB LRFD method........160
7.7: Bearing capacity factors (AASHTO, 2012)..................................167
xi


LIST OF FIGURES
Figure
1.1: Schematic of major differences between GRS and GMSE walls...................3
1.2: Variety of facings available for GRS structures (Wu, 1994)..................5
2.1: Photo of one of the first geosynthetic reinforced soil walls built: Washington State,
1975 (Powell etal., 1999).....................................................10
2.2: GRS wall constructed along the Interstate 25 corridor in Denver, CO (Wu, 2010)...11
2.3: Schematic of potential failure surface propagation through a reinforced soil mass...14
2.4: Comparison of 2x closer spacing vs. 2x stronger reinforcement...............17
2.5: FHWA Performance test data for comparison of spacing and facing (Nicks et al.,
2013)..........................................................................18
2.6: GRS-IBS Section (Adams et al., 2012).........................................19
2.7: Construction sequence for full-height rigid GRS structures (Tatsuoka, 1997).21
2.8: Photograph of completed full-height rigid wall along high-speed train corridor in
Japan (Tatsuoka, 2008).........................................................22
2.9: Photo of 55 Grand County, CO GRS Wall (Wu, 2010)...........................23
2.10: Photographs of Grand County, Colorado GRS wall: a) built-up ice cliff during
winter season and b) lack of damage to wall face due to freeze-thaw cycles over 18
years..........................................................................24
2.11: Typical section for independent full height facing (IFF) wall...............25
2.12: Photograph of completed Fox Wall in Denver, Colorado (Wu, 2010)............26
2.13: Photograph of Nadahini Creek pipe arches in Haines Junction, Yukon Territory
(Wadey and Indrees, 2014)......................................................27
xii


2.14: Finished abutment in Haines Junction, Yukon Territory a) during construction of
the superstructure and b) after construction was completed (Wadey and Indrees, 2014)
..................................................................................28
2.15: Finished product of timber-faced bridge abutment in Tongass National Forest,
Alaska........................................................................30
2.16: Bin pressure diagram (Wu, 2007)...........................................31
2.17: Picture of test completed in Commerce City, Colorado with geosynthetics detached
from facing and cantilevered load placement (Wu, 2001)........................33
2.18: Photo of bed sheets experiments (Photograph by Robert Barrett)............34
2.19: Results of study completed at Turner-Fairbanks (Nicks et al., 2013).......36
2.20: Dimensioned schematic of GRS specimen used for large-scale testing
(Pham, 2009)..................................................................38
2.21: Graph of results for confined versus unconfined GRS pier testing (Pham, 2009) ...39
2.22: Mohrs Circles with graphical depiction of increase in apparent cohesion and
confining pressure................................................................41
2.23: Comparison of analytical versus experimental results for ultimate strength of GRS
piers (Hoffman and Wu, 2015)......................................................44
2.24: Plot to determine stress-strain behavior of GRS structure by relating vertical strain
to the reinforcement mobilization factor, M (Hoffman and Wu, 2015)................45
2.25: Calculated versus measured stress-strain curves comparing Hoffman (2014)
analytical method to test results from FHWA pier tests (Hoffman and Wu, 2015; Nicks
etal., 2013)......................................................................46
2.26: The quad chart proposed by Hoffman (2015)......................................47
xiii


2.27: Ratio of horizontal versus vertical deformation for reinforced soil.............48
2.28: Load and resistance distribution for determining probability of failure (Bathurst et
aL 2008)...........................................................................52
3.1: Photographs of various facing types used for reinforced soil structures with relatively
small loads: a) geotextile-wrapped wall; b) timber-faced; c) gabion-faced; d) simulated
stone (Powell et aL 1999; I IIWA. 2005)..........................................55
3.2: Visual effect of different wall facings commonly used for a variety of applications 58
3.3: Photographic examples of each facing category using stabilized earth walls from
around Colorado: a) vertical, b) horizontal, c) mixed horizontal and vertical and d)
interlocking...................................................................59
3.4: Photographs of the types of patterns found naturally in rural settings.......60
3.5: Photographs of the types of patterns often found in man-made urban settings..61
3.6: Artistic finishes done on a wall in Riverside, CA............................62
3.7: A schematic of how parts of a wall move through various fields of vision as a car
moves through a corridor...................................................................64
3.8: Visual effect of varying focus (Intriligator and Cavanagh, 2001).........................64
3.9: Photograph of an approach to toll booths which utilized bright yellow lines on a
roadway to give the illusion of speed to drivers and remind them to slow down
(I I IWA)............................................................65
3.10: Photograph from FHWA showcasing different roadway markings........67
4.1: Summary of results for studies on how replacing natural aggregate with recycled
72
concrete aggregate affects: a) unit weight and b) compressive strength of concrete
(Xiao et al., 2005; Poon and Lam, 2008)..................................
xiv


4.2: Free body diagram of reinforcement frictional connection with blocks (Wu and
Payeur, 2014).....................................................................74
4.3: Results from analysis to determine if friction offered by lightweight recycled CMU
blocks is adequate to resist lateral pressures: a) no grout and b) top three courses
grouted for added stability.......................................................77
4.4: Schematic depicting section view of reinforced soil wall with whole tire facing
(Jayawickrama, 2000)..............................................................79
4.5: Flow of LCI results to midpoint categories to damage categories.................83
4.6: Schematics of retained earth walls compared in previous LCA study (Frischknecht et
aL 2013)..........................................................................84
4.7: Section of: a) GMSE abutment and b) GRS abutment...............................86
4.8: Flow diagram outlining three major phases where sustainability was considered..88
4.9: Example of Life Cycle Assessment diagram........................................90
4.10: Example of PET reference flow calculation......................................92
4.11: Final reference flows for materials............................................93
4.12: Comparison considering each major construction material required for GMSE and
GRS abutments considering: a) global warming potential; b) acidification; c)
eutrophication and d) smog........................................................99
4.13: Comparison using Traci of midlevel impact for GMSE and GRS abutments for 4
midpoint categories..............................................................100
4.14: Comparison using Impact 2002+ considering each major construction material
required for GMSE and GRS abutments considering: a) human health and b)
ecosystem quality................................................................101
xv


4.15: Comparison of normalized damage level impact for GMSE and GRS abutments. 102
4.16: Sensitivity analysis for 20% error in backfill volume at midpoint level using Traci:
a) global warming potential; b) acidification and c) smog........................105
5.1: Mohrs Circles of Stress and Incremental Strain................................109
5.2: Critical zones in Jewell-Milligan model........................................111
5.3: Jewell-Milligan design chart for uniform spacing and ideal length conditions...114
5.4: Jewell-Milligan design chart for uniform spacing and truncated length conditions. 115
5.5: Force diagram depicting forces present on single layer of an idealized reinforced soil
structure with block facing....................................................116
5.6: Photograph of testing setup by Ehrlich and Mirmoradi (2013): a) CMU block facing
and b) geosynthetic wrapped facing........................................120
5.7: Comparison of modified Jewell-Milligan analytical method to Ehrlich and
Mirmoradi (2013) results: a) block faced and b) wrapped-face.............122
5.8: Comparison of modeled vs measured results for Pham (2009) testing program: a)
GSGC Test 2; b) GSGC Test 3; c) GSGC Test 4; d) GSGC Test 5..............127
5.9: Comparison of effect of facing resistance shown for CMU block facing versus no
facing....................................................................130
5.10: Plot of change in vertical deformation of the reinforced soil as a function of facing
type and distance from the wall face.....................................131
5.11: Load-deformation plots shown for different facing conditions for reinforcement
spacing of: a) 0.2 m; b) 0.4 m; c) 1.0 m.................................133
6.1: Dimensioned schematic of reinforced soil mass...........................137
6.2: Photograph of specimen in metal frame with gridded membrane visible from side 138
xvi


6.3: Photograph of side view showing facing type for (a) rigid and (b) flexible facing.. 139
6.4: Particle size distribution for Smith Road Backfill...........................141
6.5: Standard proctor curve for Smith Road backfill...............................142
6.6: Photograph of initial moisture conditioning procedure........................143
6.7: Load-deformation curve for unreinforced and reinforced rigid facing test.....145
6.8: Load-deformation curve for unreinforced and reinforced flexible facing test..146
6.9: Lateral deformation data for flexible facing specimens: (a) time-series for reinforced;
(b) time-series for unreinforced; (c) deformation profiles for reinforced; (d)
deformation profiles for unreinforced...........................................149
6.10: Comparison of load-deformation curves for reinforced and unreinforced test
specimens that are (a) rigid-faced and (b) flexible-faced.......................151
7.1: Minimum and maximum particle size distribution curves allowable for LRFD GRS-
NLB design method...............................................................162
7.2: Schematic depicting loads present on NLB reinforced soil structures (AASHTO,
2012)...........................................................................164
xvii


1. Introduction
1.1. Problem Statement
Since reinforced earth structures first came into existence in the mid-20th century, their
use has been expanded from the forest roads, 1 to 2 meters in height, first built by the
United States Forest Service (USFS) to wall and abutments greater than 15 meters in
height and carrying loads as large as 200 kPa along highway corridors. As these
structures have evolved, so have the efforts to build them using cost-effective and
efficient methods. For the most part, such efforts have resulted fewer layers of
geosynthetics used, which, theoretically, would increase the rate of construction. For this
approach, as vertical spacing of reinforcing layers is increased, the tensile strength of the
reinforcing is increased as well. The thought process for designers and engineers is if
fewer layers of reinforcement are present, then construction is faster and fewer materials
are needed, which would theoretically result in a more efficient structure. However, this
may not be the case. The class of reinforced soil structures with relatively large spacing is
here referred to as geosynthetic mechanically stabilized earth (GMSE) structures.
Mechanically stabilized earth structures can also be constructed using metallic
reinforcement, but this study is limited to those structures reinforced with geosynthetics,
either geo-grids or geotextiles.
Although GMSE structures are quite common and exist all over the world, they have
been known to have a reported failure rate of between 2-4% (Berg, 2010; Valentine,
2013; Koerner and Koerner, 2012), which doesnt include those instances when failure
was caught before the structure was completed and mitigation measures applied without
reporting. Additionally, there is no assurance that failures occurring in private industry
1


are reported. These structures have failed for a variety of reason, though rarely was a
single mechanism found responsible for the failure (Collin, 2001; Koemer and Koemer,
2012). With such a high rate of failure, one must wonder why these structures continue to
be built, or if more robust options are available.
One such alternative is here referred to a geosynthetic reinforced soil (GRS) structure.
These structures differ from their GMSE counterparts in that they are defined as
structures with relatively small reinforcement spacing, usually 0.3 m or less. Although
construction materials and methods are similar for the two structures, there is a major
difference in the mechanical behavior of GMSE versus GRS structures. In short, GMSE
structures are externally-stabilized and GRS structures are internally-stabilized. A
schematic of the differences between GMSE and GRS is presented in Figure 1.1. The
external stability of GMSE means that a mass of soil, defined by strength parameters and
geometry, actively moves away from the retained soil and is held in place by the length
of reinforcement that extends beyond the potential failure plane. The structure is held
together by the tensile strength of the geosynthetic, the friction between geosynthetic and
soil behind the moving mass of soil, and by the connection of the geosynthetic to the wall
facing.
For GRS structures, the reinforced zone of the structure acts as a composite material,
which results in a structure than can achieve stable conditions without the aid of facing.
This approach to reinforced soil can be thought of in terms of area of influence of the
reinforcing geosynthetics. When spacing is relatively small, as in GRS structures, the
area of influence associated with each individual reinforcing layer overlaps with that of
the layers above and below it. The result is reduced lateral pressures on the wall face and
2


a reinforced mass that acts as a single body, confined by the relatively close spacing. The
interaction between reinforcement and soil is a function of applied loads, geometry and
the properties of both the soil and geosynthetic.
In general, there are a variety of improvements made to a soil mass by the inclusion of
reinforcing layers, including increased lateral confinements, suppressed dilation,
enhanced compaction-induced stresses, and restrained lateral deformations (Wu, 2015).
These improvements exist in GMSE structures as well, but the spacing is large enough
that these improvements to the soil do not dominate the behavior of the structures. GRS
structures tend to perform well due to the fact that the smaller the spacing, the greater the
influence of these soil improvements.
Figure 1.1: Schematic of major differences between GRS and GMSE walls
Since GRS structures are internally stabilized, the most important interaction is
between the geosynthetic and the soil; the facing connection is not a concern, particularly
when seismic loads are not expected. In fact, there is no special facing connection
required for GRS structures (Wu and Payeur, 2014), and most are constructed by dry-
3


stacking standard concrete masonry unit (CMU) blocks with layers of geosynthetic
sandwiched between courses. This frictional connection is the only connection between
reinforcement and facing. Although strong reinforcement connection to the facing is not
necessary, it may be one reason why the structures are not more commonly used.
Designers accustomed to a dependency on the reinforcement-facing connection may have
a hard time accepting a design alternative where this relationship is virtually obsolete.
Additionally, using dry-stacked CMU blocks could leave a finished aesthetic that appears
less permanent than some facing types commonly used in GMSE structures. This
perception of impermanence is incorrect, but should not be ignored as public confidence
in engineering solutions should always be considered.
When seismic loads are a concern, it may be necessary to engineer a strong connection
between geosynthetic and facing, even for those GRS structures, to ensure adequate
resistance to the relatively large loads applied over a short period of time during an
earthquake event (Tatsuoka et al., 2009).
Another reason why adoption of GRS has been slow could be that designers and
engineers working outside of the research community tend to adhere to methods they are
familiar with, which is understandable since familiarity leads to greater confidence in an
engineering solution as well as increased efficiency in both design and construction
phases. Mistrust in the internally-stabilized nature of the structures could lead to over-
designed GRS walls and abutments, which results in an expensive, inefficient end
product. There could also be some confusion concerning which facing types and
geosynthetics are compatible with GRS since GMSE walls are often bundled into
propriety systems, which usually involve pre-chosen facing elements bundled with a
4


certain type of geosynthetic. These proprietary systems are beneficial to the
manufacturers that bundle their products, but take the decision-making away from the
designer.
In GRS design, there are no propriety panels needed or special connections required
on the back of the facing elements. The frictional connection described previously is
adequate for most applications. The result of this approach to connection, or lack thereof,
is that virtually any geosynthetic fabric is compatible with a wide variety of facing types.
Figure 1.2 shows some of the facing types possible for GRS structures, including
recycled tire, segmental block, and full height rigid walls. The facing used is completely
at the discretion of the designer and can be altered to meet the needs of individual project
performance needs and site location.
Wropped-foced wall
Wropped-foc ed wall
With ihotcrete cover
C
GRS wall with full-height
concrete facing
Modular block woll
Full-height concrete
MSB woll
GRS woll with articulated
concrete facing
Timber-faced woll
M
Gobion-foced woll
mr
Figure 1.2: Variety of facings available for GRS structures (Wu, 1994)
5


In the current GMSE industry, a designer certainly has options when it comes to
materials and aesthetic. However, since facing for GMSE structures requires carefully
engineered connection to the reinforcing medium, there is typically a set number of
options for a designer to choose from, particularly for exceptionally tall structures or
those with atypically high loads.
While a custom design developed from scratch for each site and loading condition
results in a better end product, it could negatively affect efficiency. Also, such a wide
variety of materials and aesthetics to choose from means that a single option that is
objectively better than other options may not exist. This study aims to streamline the
process of designing GRS structures. The end result of this project is a new design
protocol that can be used to choose the best facing type, material, and aesthetic without
sacrificing efficiency or performance.
1.2. Research Objectives
The primary objective of this study is the development of a new design protocol that
can be used by the geotechnical community to design efficient, effective and
environmentally conscious GRS structures.
1.3. Method of Research
In order to achieve the primary research objective of developing a new design protocol
for GRS structures, the following intermediate tasks were identified and completed in
support of a holistic design method.
6


Task 1: Literature Review (Chapter 2)
This task is a review of previously completed works to summarizing the current body
of knowledge relevant to facing effects of GRS structures. This review includes full-
scale, laboratory-scale, and analytical models developed for describing load-deformation
behavior of GRS structures as a function of both spacing and facing.
Task 2: Review of Aesthetics (Chapter 3)
For this task, concepts relating to human interaction in built environments are applied
to GRS structures as a function of structure location, including natural, urban, and
transportation corridor environments. Additionally, the concept of functional aesthetic is
explored in the context of GRS facing.
Task 3: Sustainability Analysis (Chapter 4)
For this task, a method for quantifying environmental impact of GRS structure using a
Life Cycle Assessment (LCA) is developed to determine the sustainable potential of GRS
considering typical materials and construction methods. Additionally, a discussion of the
potential for incorporating recycled materials into facing elements is presented in the
context of strength and service performance metrics.
Task 4: Analytical Model Describing Load-Deformation Behavior (Chapter 5)
For this task, an existing analytical model is modified to include facing stiffness in
determining load-deformation behavior of GRS structures. The task includes preliminary
verification of the model and a sensitivity analysis exploring how various facing elements
may affect the response of the structure under an applied vertical surcharge.
7


Task 5: Experimental Analysis of Role of Facing (Chapter 6)
This task involves the development and implementation of a new experimental
approach for determining how the stiffness of the facing during both construction and
services phases may affect the load-deformation behavior of a reinforced structure. The
testing program consisted of rigid and flexible facing testing with and without
reinforcement in a laboratory-scale experiment. The results differ from the analytical
approach in that the experiments are focused on how construction methods may affect
mobilization of geosynthetic resistance.
Task 6: Design Protocol and Recommendations (Chapter 7)
The final task incorporates the previous tasks into a single design method for GRS
non-load bearing (GRS-NLB) walls. The method is calibrated to fit a load and resistance
factor design (LRFD) approach, and LRFD factors are calibrated using an existing design
method as well as American Association of State Highway and Transportation Officials
(AASHTO) recommendations for LRFD design methods. In addition to strength and
service metrics, the design method includes instructions for environmental assessments
and for choosing the right aesthetic for a specific project.
8


2. Literature Review
2.1. History of Reinforced Soil
In a time before reinforced soil structures were as commonly used as they are today,
designers and engineers did not have the option to cut soil at a vertical face safely and
were forced to cut long slopes. This approach was not ideal as it reduced the land area
available for development on a site. The next generation of earthwork technology was
gravity walls, defined as heavy structures, usually constructed from concrete that retained
soil in sliding and overturning by way of sheer mass. Engineers realized they could create
strongly retained soil by reinforcing the soil mass with a material which could offer better
tensile resistance than soil. In the 1970s, the U.S. Forest Service began constructing
reinforced soil walls with geosynthetic reinforcement. The first geosynthetic reinforced
wall was built in 1974 and the second following soon after in Washington in 1975
(Steward, 1992). A photograph of the second geosynthetic reinforced wall ever built is
presented in Figure 2.1. This photo shows a worker hanging down the front of the
sandbag facing applying gun-ite (also called shotcrete) to protect the facing from UV
exposure.
9


Figure 2.1: Photo of one of the first geosynthetic reinforced soil walls built:
Washington State, 1975 (Powell etal., 1999)
One major difference between these early structures and the structures we build today
are the expected design life and applied service loads. From their humble roots as non-
aesthetic, non-load bearing, sometimes temporary walls in the remote corners of our
national parks these structures have moved into the spotlight. As shown in Figure 2.2,
they line our highways, shape our commercial zones and allow us to build structures in
places that would have otherwise been considered impossible to develop.
10


Figure 2.2: GRS wall constructed along the Interstate 25 corridor in Denver, CO (Wu,
2010)
2.2. Current Approach to Design
Although design methods currently used for GRS structures have resulted in a safe,
reliable structure, they have been shown to be exceptionally conservative in terms of
measured versus predicted performance. In 1992, a study was completed which compared
a measured failure load on a timber-faced GRS structure to the predicted failure load
using a variety of design methods (Clayboume and Wu, 1992). Although the methods
had various recommended factors of safety associated with a final design, for this study
the factor of safety was assumed to be 1 so that differences between the design
assumptions and measured structural response could be identified. The measured quit
determined by large-scale plane strain testing at the University of Colorado Denver was
200 kPa. It is worth noting here that the reinforcement spacing for these structures was
11


0.28 m, which is a spacing that is on the cusp between GRS and GMSE behavior. In fact,
if the spacing were smaller one might expect this differences between calculated and
measured to be even greater. The results of this study are presented in Table 2.1.
Table 2.1: Comparison of measured to calculated failure loads for timber-faced GRS
structure in plane strain testing device
Method Type of Analysis Calculated quit (kPa) % difference
Forest Service Method Tie-back Wedge 4.9 -97.6%
Brams Method Tie-back Wedge 43.2 -78.4%
Collin Method Tie-back Wedge 50.9 -74.6%
Bonaparte et al. Method Tie-back Wedge 6.3 -96.9%
Leshchinsky and Perry Method Fapplied = Fresisting 36.3 -81.9%
Schmertman et al. Method Fapplied = Fresisting 41.9 -79.1%
Although the six methods listed in Table 2.1 are all limiting equilibrium methods of
analysis, the first four utilize a tie-back wedge approach to reinforced soil design and the
last two assume a failure surface then calculate applied and resisting forces for that
assumed geometry. It is obvious from this data that current methods of calculating
maximum capacity of GRS structures are not appropriate. While undeniably robust, these
overly conservative design methods can lead to walls that are inefficient and could make
it more difficult for GRS technology to be adopted on a larger scale.
Current approaches to designing with GRS are similar or identical to methods for
designing GMSE walls, with the exception of bridge abutment designs. A major
assumption for the majority of these approaches is lateral earth pressures can be
calculated by assuming Rankine earth pressure theory. There are several issues with
applying Rankine earth pressures to GRS structures. What follows is a list of assumption
12


integrated into the Rankine theory and how those assumptions compare to a GRS
structure:
1) Soil is cohesionless\ this assumption is most likely accurate for GRS structures
since freely-draining backfills are typical specified.
2) Wall is frictionless'. The back part of the wall facing would most likely have
some frictional interaction with the adjacent soil. The magnitude of the
interaction would be a function of facing type and the properties of the backfill
material, but would likely be small enough to be considered negligible.
3) Soil-wall interface is vertical'. While the target design of the facing of GRS
structure is often vertical, the movement of the facing that occurs during
compaction most likely leaves the blocks at least a little staggered, and
therefore no longer perfectly vertical. However, unless the wall facing is
intentionally designed to be off the vertical plane, the effects of this technically
incorrect assumption are likely negligible.
4) Homogeneous soil. This is another assumption which is clearly false for
reinforced soil structures. The backfill cannot be homogeneous since
reinforcing layers are present. As vertical spacing of reinforcing layers is
decreased, the heterogeneity of the reinforced mass is increased. This
discrepancy could have a large effect on how a wall may behave differently
than the design of the wall intended.
5) Failure surface is planar: Due to the heterogeneity of GRS structures as
compared with GMSE structures, there is still question as to where a failure
13


plane would occur in a GRS structure, though it is most certainly not where
traditional Rankine analysis would place it and equally as unlikely to be planar.
As can be seen in Figure 2.3, for relatively large reinforcement spacing it may be
appropriate to assume a Rankine plane failure wedge could develop between reinforcing
layers. However, as vertical spacing decreases it become more difficult to justify this
assumption.
Figure 2.3: Schematic of potential failure surface propagation through a reinforced soil
mass
Although conservative, Rankine earth pressures may be a valid method of analysis for
reinforced soil structures with relatively large vertical spacing. However, it is not valid
for GRS structures where the heterogeneity of the reinforced zone is more pronounced.
Although the Rankine assumption could be a major contributor to discrepancies
between design methods and actual behavior of GRS structures, additional
misunderstandings of GRS versus GMSE behavior are also present. One such commonly
14


repeated error is the assumption that spacing and facing are of equal importance in all
reinforced soil structures. That is, modem design methods dictate that if a designer would
like to increase the spacing of their reinforcing layers, they need only increase the tensile
strength of the reinforcing fabric by the same factor.
The fundamental design equation for all reinforced soil structures as recommended by
AASHTO is presented in Equation 1:
Treq = crH Sv Fs [Eq. 1]
where Treq is the required tensile strength of the geosynthetic reinforcement, aH is the
horizontal stress in the soil at a depth z, Sv is the vertical spacing of the geosynthetic
reinforcement, and Fs is the factor of safety. Note that this equation could easily be
manipulated to determine the required reinforcement spacing for a given allowable
tensile strength of reinforcement.
Since horizontal stress and factors of safety are treated as constants at a given depth, it
can be seen that a major assumption of this fundamental design equation is that the
vertical spacing of reinforcement and tensile strength are linearly proportionate, which
implies that spacing and tensile strength have equal contributions to wall strength. It is
because of this assumption in design that engineers have been rewarded by designing
walls with larger spacing and geosynthetics with higher tensile strength. When fewer
layers of reinforcement are utilized, there are fewer times when earthwork must stop in
order to lay the reinforcement. Also, it makes it possible to place larger loose lifts of soil
for compaction. This has, in turn, pushed the geosynthetic manufacturing industry to
15


produce fabrics with greater tensile strengths instead of focusing on other considerations,
such as environmental impact.
The design implications of Equation 1 have had an impact on the growth of reinforced
soil structures. However, recent tests have shown that this ratio between tensile strength
and reinforcement spacing is not a unique indicator of structure performance. In fact,
these tests show that decreases spacing can have a much greater positive affect on
performance than increasing tensile strength.
In 2009, Pham completed a series on tests on a 2 m tall GRS structure in plane strain
conditions (Pham, 2009), with one of the goals of the study identified as examining the
relationship between reinforcement strength and reinforcement spacing. Two of the tetss
completed by Pham (2009) were set up to determine this relationship. One test utilized a
geosynthetic with an ultimate tensile strength of 70 kN/m and 0.2 m spacing, the other
contained a geosynthetic with 140 kN/m strength at 0.4 m spacing. The results from those
tests are presented in Figure 2.4.
16


Figure 2.4: Comparison of 2x closer spacing vs. 2x stronger reinforcement
Based on the assumption that reinforcement spacing and tensile strength of reinforcing
layers contribute equally to the strength of a structure, one would expect that the stress-
strain curves in Figure 2.3 would be equal. Its obvious that they are not. In fact,
decreasing reinforcement spacing by half results in an increase in ultimate strength of
over 60% and an increase in initial stiffness of the structure of 35 percent.
Another study which supports this theory was completed by the Federal Highway
Administration (FHWA) at their Turner-Fairbanks Highway Research Center in McLean,
Tf
Virginia (Nicks et al., 2013). In this study, the ratio of was explored to determine if
sv
this ratio is a unique indicator of wall performance. Three different tests were completed
with varying vertical reinforcement spacing and tensile strength of geosynthetic, but with
Tf
a constant ratio of 180 kPa (3,800 psf). The results from this study are presented in
sv
Figure 2.5.
17


Figure 2.5: FHWA Performance test data for comparison of spacing and facing (Nicks
et al., 2013)
As can be seen in Figure 2.4, similar ratios of vertical spacing to tensile strength do
not result in similar stress-strain behavior. The smallest spacing had the stiffest response
to vertical loading despite also involving the geosynthetic with the lowest tensile
strength.
2.3. GRS Bridge Abutments
Although the language so far has combined all reinforced soil structures with
reinforcement spacing < 0.3 m as GRS structures, it is worth mentioning that a separate
design recommendation exists for those reinforced soil structures used at bridge
abutments (Wu, 2006, Adams, 2011; Adams et al. 2012). In 2012, the FHWA made
recommendations for implementing GRS as bridge abutment structures. These structures
are often called Geosynthetic Reinforced Soil Integrated Bridge Systems (GRS-IBS)
because the method allows the roadway to be blended with the superstructure. A
18


schematic of a section of GRS-IBS is presented in Figure 2.6 showing how the
components of the structure are integrated. The bearing bed labeled in Figure 2.6 is also
an important component of a GRS abutment. The closely spaced reinforcing layers in this
zone add additional strength to the structure directly under the beam seat, which assists in
limiting deformation.
Figure 2.6: GRS-IBS Section (Adams et al., 2012)
By incorporating the roadway into the superstructure, the problematic differential
settlement that often occurs at the roadway-superstructure interface is avoided, thus also
avoiding costly repairs down the road. This system also has the advantage of being
quickly constructed in all types of weather, not requiring special equipment like cranes or
19


boom trucks, and being easily altered during construction should on-site conditions differ
from expected.
The initial tested performance of GRS-IBS was so promising that the FHWA chose
GRS-IBS for the Every Day Counts initiative in order to accelerate the usage of these
structures. These structures are recommended for single-span bridges not greater than 10
m in height with a span of less than approximately 40 meters. The applied vertical stress
is currently limited to 200 kPa, though GRS-IBS designed outside of these boundaries are
possible as long as conditions are thoroughly tested using the FHWA Performance Test
method (Nicks et al., 2013).
2.4. Case Studies of GRS Structures
Although GRS has not been as widely accepted as GMSE in the United States, there
are certain markets where GRS has been accepted as a legitimate, and often superior,
engineering solution. What follows is a discussion of several different case studies, both
domestic and international, that showcase the large range of performance objectives and
aesthetic treatments that can be achieved with GRS technology.
2.4.1. Japan Full-Height Rigid
One location where the performance of GRS has been embraced is in Japan. By the
year of 2000, over 30 km of GRS structures were built including walls and bridge
abutments, some specifically built for the high-speed trains Japan is famous for (Tatsuoka
et al., 1997). These structures in Japan are typically constructed with full-height rigid
facing, which poses a unique engineering dilemma. If the walls are not permitted to
deform during construction, then the strength of the geosynthetic cannot be mobilized
and the wall becomes a gravity retaining wall instead of a reinforced soil structure. To
20


address this problem, engineers in Japan introduced a novel construction method. A
schematic of this process is presented in Figure 2.7.
Figure 2.7: Construction sequence for full-height rigid GRS structures (Tatsuoka et al.,
1997)
The process involves allowing the wall to deform during construction by utilizing a
primary facing that is flexible. Once the structure is complete, a secondary cast-in-place
full-height facing is constructed directly over the flexible facing. The end result is a
structure that meets the exceptionally rigid deformation tolerances associated with bridge
abutments built for high-speed trains, yet allows deformation during construction so that
21


the strength of reinforcing layers can be mobilized. The aesthetics of the finished product
are presented in Figure 2.8.
Figure 2.8: Photograph of completed full-height rigid wall along high-speed train
corridor in Japan (Tatsuoka, 2008)
2.4.2. Grand County, CO
In 1997, a GRS wall with a maximum height of 16.8 m was constructed in Grand
County, Colorado. The wall was construction with a woven polypropylene geosynthetic
and a well-compacted granular backfill. The facing of this wall is made up of dry-
stacked, split faced concrete blocks. This wall is interesting because if the Rankine earth
pressure assumption was accurate for GRS structures, it would be physically impossible
for this wall to be standing. It is still a functioning wall 18 years after being constructed
therefore it is clear that Rankine does not apply.
22


Figure 2.9: Photo of 55 Grand County, CO GRS Wall (Wu, 2010)
Figure 2.10 presents two photographs taken of this GRS wall. Figure 2.10a is a
photograph taken during winter when an ice cliff forms on the fact of the wall coming out
of the large-diameter pipe at the top of the wall. Figure 2.10b is a photograph taken in
July 2015 and highlights the lack of damage to the wall, despite 18 years of annual
freeze-thaw cycles.
23


(a) (b)
Figure 2.10: Photographs of Grand County, Colorado GRS wall: a) built-up ice during
winter season and b) lack of damage to wall face due to freeze-thaw cycles over 18 years
2.4.3. Fox Wall, Denver, CO
In 1996, the Colorado Department of Transportation (CDOT) required a new segment
of retaining wall along the 1-25 corridor in Denver, CO (Wu, 2010). One restriction on
this project was that, due to being located along a busy highway corridor, CDOT required
that the highway remain open during construction. In order to keep the highway open, the
area influenced by the construction needed to remain small. They decided to go with a
GRS wall with an independent full height rigid facing. A typical section for this type of
structure is presented in Figure 2.11.
24


Figure 2.11: Typical section for independent full height facing (IFF) wall
As discussed previously in the Japan case study, it is important for reinforced soil
structure to have some flexibility in order to mobilize resistance in the reinforcing layers.
The independent full height facing (IFF) wall concept is convenient in that the wall
panels are originally placed at a negative batter so that deformation can take place during
construction, pushing the panels into a vertical position. There was also a mechanism
25


built into the panels that allowed for adjustment such that contractors can ensure panels
line up and the aesthetic of the finished product is not marred by staggered panels.
The wall height varied from 1.7 m to 5.7 m and the total length of the wall was 426
meters. The facing panels were either 1.2 m or 2.4 m wide and the reinforcement used
was a welded wire mesh. The vertical spacing was 0.3 meters. A photograph of the
finished wall is presented in Figure 2.12.
Figure 2.12: Photograph of completed Fox Wall in Denver, Colorado (Wu, 2010)
Rebar face anchors were instrumented to monitor the performance of the wall post-
construction at two different stations, designated Station 3116 and Station 3119. At
Station 3116, they were located at the ground surface and at 1.8 and 3.4 m above the
ground surface. At Station 3119, they were located at the ground surface and 0.9 and 2.1
m above the ground surface. Each of the forces of the rebar was very small, and in some
cases they were in compression, indicating that the lateral earth pressure along the wall
were small.
26


2.4.4. Aluminum Faced Wall, Canada
In 2013, the Department of Transportation in Yukon, Canada (YKDOT) constructed
the first ever GRS bridge abutment built by the Canadian government (Wadey and
Indrees, 2014). The project was located in Haines Junction in Yukon Territory. The
original culverts are pictured in Figure 2.13 and were constructed in 1968, serving as a
bridge of Nadahini Creek.
Figure 2.13: Photograph of Nadahini Creek pipe arches in Haines Junction, Yukon
Territory (Wadey and Indrees, 2014)
During routine maintenance it was discovered that the bottom portion of the steel pipes
had been worn through and the structural integrity of the arches undermined. After
looking into several different options, including replacing the existing structure with a
similar culvert system, installing H-piles and concrete box girders, and an open-bottom
27


pipe arch. In the end, YKDOT decided to implement the bridge abutment design outlined
by the FWHA in the US (Adams et al., 2011). One major design consideration was the
remoteness of the area was and how difficult it would be to get materials to the site. The
original design called for steel panels, however the contractor recommended using
corrugated metal panels which would be lighter. This helped with both the cost of
transportation and ease of construction since the panels were light enough to be lifted by
hand into place. Photograph of the structure during superstructure construction and after
construction was completed are presented in Figures 2.14a and 2.14b, respectively. The
reinforcement used was a geotextile with a wide-width tensile strength of 70 kN/m
spaced at 0.2 m just below the girder and at 0.275 m elsewhere in the abutment.
(a) (b)
Figure 2.14: Finished abutment in Haines Junction, Yukon Territory a) during
construction of the superstructure and b) after construction was completed (Wadey and
Indrees, 2014)
28


The corrugated aluminum sheeting recommended by the contractor proved its worth as
it was easy to construct and offered an aesthetically pleasing end product. The total time
for the project from start to end was 45 days, which included erecting and removing a
detour bridge necessary to keep the route open. Benefits of using the GRS-IBS system
were reported by YKDOT to be: minimal materials required, long service life, ease of
installation, and minimized likelihood of settlement.
2.4.5. Tongass National Forest, Alaska
Early in the development of reinforced soil technology, the United States Forest
Service built a timber-faced structure in Tongass National Park, Alaska. Although the
USFS wasnt using GRS as a way to describe structures, they recommended that
relatively small spacing be used (Powell et al., 1999). Geogrid reinforcements were used
with a treated timber facing. The reinforcement spacing was reported at 0.3 m near the
bottom of the abutment and 0.15 mm near the top of the abutment. The backfill used
was a freely-draining backfill with a maximum particle size of 0.025 m. The timber
facing used for the abutment was treated to avoid UV degradation and meet the
requirements of the 50-year design life specified for the structure. Additionally, an
additional, 0.05 m thick timber facing was applied to the primary facing in such a way
that it could be maintained or replaced without replacing the entire structure. A
photograph of the completed bridge is presented in Figure 2.15.
Similar to the project discussed in Yukon Territory, this timber-faced, small spacing
reinforced structure was convenient to build in this remote area due to being constructed
quickly and the use of readily available materials.
29


Figure 2.15: Finished product of timber-faced bridge abutment in Tongass National
Forest, Alaska
2.5. Role of Facing in GRS Performance
While externally stabilized systems (retained soil) rely on a strong connection between
reinforcement and wall facing and require that that the facing contribute to the overall
stability of a structure, internally stabilized systems (reinforced soil) only require facing
to act as a construction aid, to prevent sloughing, and as an aesthetic facade (Wu, 2010).
Since the soil and reinforcing layers are stable without facing, the only forces the facing
needs to be designed to withstand are the compressive forces from the weight of the
blocks above and from some lateral pressure exerted on the facing from the soil behind.
These lateral pressures, however, have been measured to be much lower than one might
expect by applying the familiar Rankine earth pressure theory.
Using a Rankine analysis could result in significantly higher earth pressures calculated
than are actually present. In fact, these pressures have been found to be nearly
30


independent of wall height and much more related to reinforcement spacing (Wu, 2001;
Wu, 2007; Wu, 2010). In theory, where no soil is moving against the back of the facing,
no pressure is present. This would indicate that, particularly for structures with a flexible
facing, at each reinforcement layer the pressure exerted on the back of the wall is zero,
though realistically there may be some deformation of the geosynthetic that would cause
some force to act laterally against the facing. This approach to lateral pressure is called
the Bin Pressure theory. A recommended bin pressure diagram is presented in Figure
2.16 which accounts for some movement of geosynthetic and may be a more reasonable
assumption than an ideal bin pressure scenario.
1/3 ah
T~ s M
0.7 S
y 0.3 S
k-1
2/3 ah
Reinforcement
Reinforcement
Figure 2.16: Bin pressure diagram (Wu, 2007)
Since the pressure exerted on the facing is relatively small, there is not substantial
performance required from the facing. Although this theory idealizes the state of a GRS
structure, there have been some physical models developed which indicate that this bin
pressure theory may be more accurate than assuming Rankine earth pressure. One way
for this theory to be tested is to compare the behavior of GRS structures with and without
31


facing. Where little difference is present, it can be inferred that minimal pressure is being
applied to the facing elements.
There were two testing programs completed by CDOT which aimed to determine how
facing, or lack thereof, affected the performance of reinforced soil structures with
relatively close spacing. The first was completed in 1994 and consisted of a large
reinforced structure built with modular blocks as the facing. Metal wires were attached to
the geosynthetic just behind the facing and, after loading the structure with 180 kN of
concrete jersey barriers, a current was sent through the metal wires which created enough
heat to melt the reinforcing sheets and sever the connection between geosynthetic and
facing. There was no increase in rate of deformation of the wall when the geosynthetic
connections were severed and it was determined that, extrapolated to 100 years, the total
movement of the wall would be around 0.04 m. A picture of the test is presented in
Figure 2.17.
32


Figure 2.17: Picture of test completed in Commerce City, Colorado with geosynthetics
detached from facing and cantilevered load placement (Wu, 2001)
In 1996, CDOT conducted an experiment using a local silty gravel road base soil
reinforced with simple bed sheets, the type which can be purchased at a local market
(Wu, 2001). In this test, reinforcing sheets were spaced at 0.2 m and jersey barriers were
used temporarily to assist in creating an initial vertical face of the structure. After
construction was completed, the jersey barriers were removed, exposing the facing, and
33


21 total jersey barriers were place on top of the structure with no visible lateral
deformation occurring in the soil. A photo of this experiment is presented in Figure 2.18.
Figure 2.18: Photo of bed sheets experiments (Photograph by Robert Barrett)
The results of both of these tests indicate that the lateral pressures occurring on the
wall facing is small, and that a lack of facing does not significantly affect the
performance of these structures. These earliest experiments were positive proof that
34


engineers were severely underestimating the strength of reinforced soil walls with small
reinforcement spacing. In order to better quantify the strength of these structures,
laboratory testing was required.
In 2013, Nicks et al. conducted a series of tests that included a study on how
reinforcement facing affected the contribution of wall stiffness or confining pressure. The
reinforced specimens were piers with facing on all four sides. Variables in the testing
program include vertical spacing of reinforcing geosynthetics, tensile strength of
geosynthetic, and facing. The piers were constructed with dry-stacked CMU blocks as the
facing material. For half of the tests, these blocks were removed prior to testing. A
summary of test conditions and results is presented in Table 2.2.
Table 2.2: Summary of results from series of tests involving facing effects for varying
reinforcement spacing (Nicks et al., 2013)
Facing Spacing Initial Stiffness Quit
Test ID [type] [inches] [ksf] [psf]
TF-2 CMU 7.625 710 25260
TF-3 None 7.625 330 17491
TF-6 CMU 7.625 750 43763
TF-7 None 7.625 320 26546
TF-9 CMU 15.250 550 22310
TF-10 None 15.250 260 10330
TF-11 None 3.8125 390 23249
TF-12 CMU 3.8125 810 29030
TF-13 None 11.25 220 12960
TF-14 CMU 11.25 460 23562
. For the majority of the tests, the height of the specimen was approximately 1.9 m and
the base of the specimen was 1 meter. For the test in which the reinforcement spacing
35


was 0.29 m, the height of the specimen was 2 m to account for spacing which did not
coincided with typical CMU dimensions.
A plot of the results of these experiments is presented in Figure 2.19. The most drastic
difference in behavior based on facing versus no facing was seen for the ultimate strength
of the structures. The trend for the range of reinforcement spacing studied can be seen to
be linear and range from a facing to no facing ratio of 1.25 to as high as 2.20 for the
largest reinforcement spacing.
2.5
a
2.0
£
U
1 5 --
p4


a Ultimate Strength
Initial Stiffness
Reinforcement Spacing (inches)
Figure 2.19: Results of study completed at Turner-Fairbanks (Nicks et al., 2013)
In 2009, Pham completed an experimental study of, among other things, the effect of
confinement on the stress-strain behavior and ultimate strength of a GRS pier (Pham,
2009). The soil mass was tested in plane strain conditions inserting a layer between the
soil and outside plastic wall which consisted of a 1 mm layer of silicon grease and a 0.5
mm thick latex membrane. This method of reducing side wall friction was first developed
36


by Dr. Fumio Tatsuoka at the University of Tokyo and reduced the friction angle between
soil and wall to less than 0.5 degrees. The exterior of the soil mass was instrumented with
linear variable differential transducers (LVDT) in both the vertical and lateral
dimensions. The geotextile used as a reinforcement was a Geotex 4x4, which is also
identified as US4800, manufactured by US Fabrics. Wide-width tensile tests were
completed and the fabric was found to have an ultimate tensile strength of 70 kN/m.
The specimen was first constructed by using dry-stacked CMU blocks as the facing.
After construction was complete, the blocks were removed and the entire soil mass was
wrapped with a 0.5 mm thick latex membrane. Suction was applied to the airtight system
to maintain a constant confining pressure throughout testing.
A dimensioned schematic of this test is presented in Figure 2.20. The reinforcing
geosynthetics were spaced at 0.2 m and the specimen was failed by applying a vertical
load. The two confining conditions tested were 34 kPa and 0 kPa.
37


SIDE VIEW
FRONT VIEW
T9
: If -sl
_ d s
w 9
i! -:b
JLl
Cross Seetjon A-A
Figure 2.20: Dimensioned schematic of GRS specimen used for large-scale testing
(Pham, 2009)
The results from this study are presented in Figure 2.21. It can be seen that a confining
pressure of 34 kPa applied to the specimen did have a positive impact on the overall
stress-strain behavior. However, even without a confining pressure the pier was able to
withstand an applied vertical stress of 1,900 kPa before failure, which is quite large
considering the FHWA limits bearing stress in abutment design to approximately 200
kPa.
38


3000
Figure 2.21: Graph of results for confined versus unconfined GRS pier testing (Pham,
2009)
Additionally, it can be seen that the small-strain behavior of both the confined and
unconfined specimens are similar within the range of strain expected a full-scale GRS
structure. This result seems to disagree with results from Nicks et al. (2013) as presented
in Figure 2.19, however this difference may be due to the membrane used by Pham
(2009) assisting in preventing sloughing of soil from between reinforcement layers as the
structure deformed and likely also has to do with the difference between the plane-strain
tests and the three-dimensional piers.
While these tests continue to offer evidence to the fact that small reinforcement
spacing can decrease the need for stiff facing, there is still need for testing program which
further inspect the role of facing in the strength and stress-strain behavior of these
structures. Particularly, there is a need to fully define how the stiffness of a facing in both
construction and service phases can affect the performance of a wall. Design methods
39


leaving out the effects of facing may be overly conservative and miss both small-strain
and ultimate strength behavior of these structures.
2.6. Analytical Approach to GRS Behavior
Current design methods typically use similar techniques for designing GRS and
GMSE walls. As discussed in Section 2.2, this may not be an accurate approach to GRS
design, resulting in a final project with is over-designed to the point of inefficiency. The
FHWA allows design of GRS-IBS outside of their recommended boundaries, but doing
so requires a large-scale testing program to be completed to predict stress-strain behavior.
It would be beneficial to adopt some analytical method of analysis to avoid the time and
labor intensive experimental process. Recently, attempts have been made to develop
analytical models to describe stress-strain GRS behavior. Two of these methods are
discussed here, one presented by Pham in 2009 and another developed by Hoffman in
2015.
2.6.1. Pham, 2009
Pham (2009) developed a new analytical model for determining strength properties of
GRS structures. For this model, the theory of reinforcement causing an increase in both
apparent confining pressure and apparent cohesion was applied to GRS structures
(Schlosser and Long, 1972). Figure 2.22 is a graphical representation of how both
confining stresses and cohesion can be enhanced by the inclusion of reinforcing elements
in a soil mass.
40


Figure 2.22: Mohrs Circles with graphical depiction of increase in apparent cohesion
and confining pressure
From this Figure, the apparent cohesion due to the presence of reinforcement can be
calculated to be:
£ _ (Ap3R*/KiQ £
R 0
[Eq. 2]
where CR is the apparent cohesion in a reinforced soil mass, Act3R is the increase in
confining pressure do the reinforcement in the soil mass, KP is the Rankine coefficient of
passive earth pressure and C is the cohesion of the soil alone. The increase in the
confining pressure due to presence of tensile reinforcement in a soil mass was defined by
Schlosser and Long to be the ratio of tensile strength of the reinforcement to vertical
spacing. As presented with Equation 1 in Section 2.2, this linearly proportionate
relationship between spacing and reinforcement strength has been shown through
41


extensive testing to not be a valid assumption. Therefore, Pham introduced a new way of
defining this increase in confining pressure:
Acr3
[Eq. 3]
where
W = r
[Eq. 4]
where r is a dimensionless factor which is a ratio of average force in the reinforcement to
maximum force in the reinforcement and Sref is a reference spacing developed
empirically and equal to 6 times the maximum particle size of the reinforced soil. For a
variety of vertical pressures and reinforcement lengths, Pham determined 0.7 to be a
reasonable value for the parameter r.
Therefore, substituting in values for r and Sref we can now define the W-factor as:
Substituting Equation 5 into Equation 4, the result can be plugged into Equation 3 to
determine a modified equation for calculating apparent cohesion in reinforced soil
structures:
[Eq. 5]
42


[Eq. 6]
CR = 0.7
which, from the graphical solution presented in Figure 2.22, results in a maximum
vertical load capacity of a1R equaling:
By substituting apparent increase in both cohesion and confining pressure into the
model presented by Schlosser and Long (1972), Pham was able to develop a model that
predicts ultimate capacity of a full-scale GRS composite specimen tested in plane strain
conditions to an accuracy of with 10% of the measured data. This accuracy is much
greater than with any other model developed to analyze GRS behavior. Table 2.3 presents
a summary of how Pham (2009) analytical results compare with the accompanying
experimental program.
Table 2.3: Comparison of analytical versus experimental results for ultimate strength of
GRS structure (Pham, 2009)
Test ID Reinforcement Spacing Reinforcement Strength quit Measured quit Analytical % Difference
Test 2 0.2 70 kN/m 2700 kPa 2460 kPa -8.9%
Test 3 0.4 140 kN/m 1750kPa 1900 kPa 8.6%
Test 4 0.4 70kN/m 1300 kPa 1250kPa -3.8%
It can be seen in Table 2.3 that there is exceptionally good agreement between the
Pham (2009) analytical model and the experimental results. Comparison was also made
[Eq. 7]
43


between the analytical results and 4 other available data sets from GRS experiments. The
results from that comparison is presented in Figure 2.23.
Figure 2.23: Comparison of analytical versus experimental results for ultimate strength of
GRS piers (Hoffman and Wu, 2015)
It is apparent from Figure 2.23 that the analytical approximation developed by Pham
(2009) is an excellent approximation of ultimate capacity of GRS structures.
2.6.2. Hoffman, 2015
Although the load-carrying capacity equation presented in the previous section has
been proven to be an adequate method for calculating ultimate vertical capacity of a GRS
structure, there is still the question of stress-strain behavior in order to fully define
strength characteristics. Hoffman and Wu (2015) developed a new analytical method for
predicting load-deformation behavior based off of results from tests completed by the
FHWA on their GRS-IBS performance test. This approach utilizes the W-equation
44


similar Pham (2009) as well as the reinforcement mobilization factor, M, which is
defined as the fraction of reinforcement strength that can be mobilized for load carrying.
As a soil mass is loaded stresses increase, however deformation of a soil mass relieves
those stresses. In order to derive a method for determining M, Hookes law must be
assumed for the composite structure. While Hookes law may seem like a stretch to apply
to materials such as soil, when the behavior of the soil is determined by sheets of metallic
or polymeric reinforcement and when discussing in terms of relatively small strains it
may be appropriate.
Hoffman and Wu (2015) introduced a variable called the M-factor, defined as the
fraction of reinforcement strength mobilized for load-carrying, which describes how the
horizontal and vertical stresses vary as deformation occurs in the soil mass. A plot
defining the M-factor versus vertical strain behavior for a modeled structure is presented
in Figure 2.24.
t-t
4 1-4-
4-
, 25% S 20% X
v |. f t 4- _
MU. X
a 35 ,
-h-Nt |
§ 10% ~r
-rrH- '1 V 4-
;> tt _U
TTT 4- - 1
5% - 1

0% -1 ( ! 1 1 0 1 0 2 R el 0. nl b rc 0, en 4 n< r it 0. ft 5 4< 3.( li 7n 0 tio 7 n. ft c 4 1.8 C r 19
Figure 2.24: Plot to determine stress-strain behavior of GRS structure by relating vertical
strain to the reinforcement mobilization factor, M (Hoffman and Wu, 2015)
45


Hoffmans model was verified by comparing the calculated stress-strain results to data
published by Nicks et al. in (2013). The results of this comparison are presented in Figure
2.25.
Figure 2.25: Calculated versus measured stress-strain curves comparing Hoffman (2014)
analytical method to test results from FHWA pier tests (Hoffman and Wu, 2015; Nicks et
al., 2013)
It can be seen in Figure 2.25 that the calculated stress-strain curves for GRS behavior
are reasonably representative of the measured values from the series of tests completed
by the FHWA (Nicks et al., 2013). Results are especially similar in the realm of small
strain, which would be expected considering the assumption of linear response to loading.
Beyond the ultimate capacity of the GRS mass, this model may not apply. However, the
46


behavior of a structure is most relevant to designers is the stress-strain behavior prior to
ultimate capacity, so this model retains its usefulness in the realm of design.
Figure 2.25 is the so-called Quad Chart proposed by Hoffman (2015) as a graphical
representation of the complicated material and geometric dependencies that control
internally-stabilized (spacing-based) versus externally-stabilized (facing-based) response
to loading.
2.6.3. Summary of Behavior
Throughout Chapter 2.6, both experimental and analytical approaches to modeling
GRS behavior have be presented. For the most part, the data is in agreement that the
response to vertical loading of a GRS structure differs from that of a GMSE structure.
47


However, it can be difficult to quantify what that difference is. The response of
reinforced soil structures is also depending on the load applied, as the load-deformation
behavior for these structures, particularly those built with geosynthetic reinforcement, is
only linear for the relatively small deformations usually associated with service loads.
Although it has been shown that GRS structures with or without facing are able to
withstand applied stresses much larger than expected strength limits, there is still the
question of how deformation is affected by internal versus external stability, in both
vertical and lateral directions. Utilizing results from GSGC Tests 2 and 4 (Pham, 2009),
the ratio of horizontal to vertical deformation was calculated for each of the structures at
various applied surcharges and plotted against the fraction of ultimate capacity for each
test. These results are presented in Figure 2.27.
Figure 2.27: Ratio of horizontal versus vertical deformations for reinforced soil
48


It is clear in Figure 2.27 that the spacing directly effects the stiffness of the structure.
GSGC Test 2 was testing with 0.2 m spacing, and GSGC Test 4 was testing with 0.4 m
spacing. Not only did the smaller spacing reduce the initial deformation ratio, but the
ratio remained constant until a large enough load was applied for the structure to
transition into external-stability dependent behavior. On the Quad Chart presented in
Figure 2.26, this represents movement from Quadrant III (spacing-dependent plastic
behavior) to Quadrant IV (facing-dependent plastic behavior).
Is summary, GRS structures are inherently more robust structures than their GMSE
counterparts. Under typical service loads, GRS response to loading is not facing-
depending, as evidenced by Figure 2.27 compared with Figure 2.26, and therefore facing
for these structures is non-essential.
2.7. Load and Resistance Factor Design
2.7.1. Allowable Stress Design versus Load and Resistance Factor Design
For many years, designers used Allowable Stress Design (ASD) methods for designing
reinforced soil structures. The ASD approach is applying a single factor of safety to the
calculated design load, regardless confidence in the performance of individual
components. An example of this type of approach is presented in Equation 8:
FS Fappiied < Fresisting [Eq. 8]
where FS is the factor of safety, Fapplied represents the sum of applied forces, and
Fresisting represents the resisting forces. Although a single factor of safety is applied to
the entire structure, there are times when factors of safety for various failure mechanisms
49


are different. For example, for reinforced soil structures the FHWA recommends a factor
of safety of 1.5 for sliding and global stability, and a factor of safety of 2.0 for
overturning.
Load and Resistance Factor Design (LRFD) applies a factor of safety to individual
components of both applied and resisting forces. An example of this type of approach is
presented in Equation 9:
y0D + yLL + YwW < (|>tT + (J)cC + 4>bB [Eq. 9]
where D is the dead load, L is the live load, and W is the wind load associated with the
structure and Ym Yl> ar|d Yw are their respective load factors. On the right side of the
equation, T represents tensile members, C represents member in compression, and B
represents beams present in the structure, while c|)D, c|)L, and c|)w are their respective
resistance factors. Examples of LFRD factors recommended by AASHTO are presented
in Table 2.4. Its important to note that these factors are often grouped into various load
combinations, with different factors depending on which loads are being analyzed
together.
Table 2.4: Recommended Resistance Factors for Geosynthetic Reinforced Structures
Component Strength Extreme Event
Tensile Resistance of Geosynthetic Reinforcement 0.9 1.2
Pullout Resistance of Geosynthetic Reinforcement 0.9 1.0
Sliding at Soil Reinforcement Interface 0.8 1.0
50


2.7.2. LRFD in Geotechnical Engineering
The effort to move geotechnical engineering into the realm of LFRD is an important
step for the industry. Not only because the field of structural engineering been
successfully using LRFD for the past 3 decades, but also because the effort of calibrating
LRFD factors forces a better understanding of the materials being used and the associated
risks. Geotechnical engineers have struggled with LRFD factor calibration because the
materials associated with geotechnical engineering have quite a bit more variability than
structural elements. For example, in geotechnical engineering the unit weight of the soil
shows up on both the load and resistance side of the equation. Equate this to a steel beam
in a structural building which has both a strength and a weight. The beam has a precisely
known weight and strength and is manufactured in a controlled environment. Soil, on the
other hand, can have a 20 pcf variation depending on compaction effort, how close the
soil used is to the soil that was specified, and how much water was added during
construction. Additionally, should it rain during the construction process, the weight of
the soil would change drastically. Although there are geotechnical design standards that
apply principles of LRFD design, they have mostly been developed by calibration by
fitting, which is adjusting LRFD factors such that the end result is the same as ASD
methods.
Some work has been done to encourage LRFD factor calibration through statistical
analysis (Allen et al., 2005; Bathurst et al., 2008; Bathurst et al., 2011a; Bathurst et al.,
2011b; Kim and Salgado, 2011; Huang et al., 2011), though often these efforts are
capable of calibrating just a single factor, such as pullout failure of steel strip
reinforcement, and to date there has not been a complete LRFD calibration for a reinforce
51


earth structure, with the biggest contributor to this shortfall being variability in soil.
However, there is some effort to reduce variations in soil by a method called intelligent
compaction which incorporated a feedback system that is capable of adjusting
compaction efforts based on density measured as compaction is occurring (Briaud and
Sau, 2003; Anderegg and Kaufmann, 2004).
Bathurst et al. (2008) presented a general method for calculated probability of failure
for components of reinforced earth wall. In their example, the failure mode was pullout of
steel reinforcement. A graphical representation of distributions and probability of failure
is presented in Figure 2.28.
Figure 2.28: Load and resistance distribution for determining probability of failure
(Bathurst et al., 2008)
In Figure 2.28, random distributions of load (Q) and resistance (R) are presented on
the left. The distribution of R-Q (g) is presented on the right. The probability of failure is
the region highlighted which is one standard deviation from the mean of g (u) multiplied
by the reliability index, p. If the distributions of Q and R are normal then the reliability
index is calculated as the reciprocal of the coefficient of variation for g:
52


R-Q
[Eq. 10]
In this method, it is possible to easily calculate the probability of failure for a given
distribution of load and resistance. There do exist some nuances in calculating loads and
resistance of a materials. For example, if one were to attempt to establish LRFD factors
for a geosynthetic, the first thing to do is determine which failure mode is being studied,
for example pullout or rupture. If pullout is the failure mechanism, then it gets
complicated because soil unit weight affects both loads applied and ability for the fabric
to resist. If rupture is the assumed failure mechanism, assumptions must be made about
stresses induced on the geosynthetic in the soil mass, which can be difficult given the
dynamic nature of polymeric reinforcement.
The current state of LRFD in geotechnical engineering is a state of discontinuity.
Some materials, such as steel reinforcement, have undergone multiple studies and have
LRFD factors calibrated as well as can be expected given variability of soil properties.
Although there has been effort to decrease variations in soil compaction by methods such
as intelligent compaction, these methods are likely only valid for soils with well-known
properties that are engineered to behave in a certain way.
53


3. Aesthetics of Reinforced Soil Facing
3.1. Current State of Practice
There exists an almost limitless variety of wall facing materials and construction
methods that can be used with reinforced soil walls. For walls where facing is a primary
component of the structural integrity of the wall, these finished are limited to those
materials which can also incorporate some method for attaching a reinforcing layer.
Additionally, as heights and loads increase, those facing materials which are practical to
implement become fewer and fewer. For the purpose of this discussion on aesthetics, soil
structures will be broken into two different groups: small load and large load. Structures
in the small load category include those involving low-volume roads and relatively short
non-load bearing (NLB) walls. The large load category includes high-volume roadways
supported by reinforced soil, bridge abutments and relatively tall NLB walls.
Reinforced structures involving relatively small loads have a much wider variety of
facing aesthetics available since smaller loads result in facing which not as critical. The
United States Forest Service (USFS) has both developed and adopted reinforced soil
technology and has reported as many as 500,000 km of roads built using soil
reinforcement (Powell et al., 1999). The USFS is in the unique situation of building a
large amount of structures in remote areas of the country. Along with this comes the
responsibility to not negatively affect the surrounding landscape by introducing an
abruptly different aesthetic with their reinforced structures. For small loads, almost any
material can be used as facing since there is little lateral pressure behind the wall and
since vertical loads above the bottom course of facing, which is the critical course, never
54


get excessively high. Figure 3.1 shows four examples of the types of aesthetic finishes
which are common for reinforced soil structures.
relatively small loads: a) geotextile-wrapped wall; b) timber-faced; c) gabion-faced; d)
simulated stone (Powell et al., 1999; FHWA, 2005)
For walls with relatively large vertical loads, increasingly large horizontal loads are
expected. Due to these increased loads, extra consideration must be given to the rigidity
55


of the wall facing. However, as demonstrated with the Grand County wall presented in
Chapter 2, lateral pressure on walls is not nearly as high as the assumption of Rankine
earth pressures would result in. While it may be tempting with large structures or
structures with large loads to apply a rigid facing to offer additional strength, it is the
flexibility of the wall in moving slightly that allows the pressure behind the wall to
dissipate. Therefore, the major considering with large walls in ensuring that the facing at
the bottom of the wall has a high enough compressive strength to withstand the weight of
the blocks stacked above it. This weight of block above can be alleviated some by
including a batter in the wall face, therefore limiting the number of courses above any
given block.
3.2. Human Impacts and Aesthetics
Thus far this study has consisted mostly of hard engineering evidence pertaining to the
role of facing in GRS structures. As engineers, we tend to think this is the most important
aspect of the project. Although engineers and contractors often walk away from a project
after it is built, there is one function of a structure besides performance that still needs to
be considered throughout the remainder of its design life. That is the aesthetic of the
structure. Depending on where the structure is built, people in the surrounding
community may interact with it on a daily basis. If it is a small wall along a pedestrian
path, the aesthetic function may be to blend in as much as possible and the purpose of the
facing may be to ensure that nobody notices it. If it is a large structure along a highway
corridor, it will certainly be noticed and attention needs to be given to the question of
whether or not the human interaction with the structure is a positive one.
56


This portion of the study attempts to determine how the aesthetics of reinforced soil
walls are functioning in current practice and how this practice can be improved to get
more out of these structures. Due to differences in intended aesthetic, the current state of
practice is broken down into rural and urban settings. Special attention is given to urban
settings and how highway engineers and roadway designers can maximize the potential
of large reinforced soil walls to be used to assist in readability and way-finding. In
general, the result of this study is that blending of wall patterns in rural setting is good
practice, but more can be done in urban setting to maximize the public potential of the
wall facing.
The FHWA categorizes reinforced soil structures by the facing which was used
(FHWA, 2005), having more to do with material than appearance. However, in order to
have the language to discuss these structures in terms of how human users experience
them, they will be organized by the spaces in which they are used (urban or rural) and the
visible pattern on the face of the wall. This study will deconstruct the visual effect of the
wall finish and compare how popular fa9ade appearances differ from region to region. In
addition, specific attention will be paid to the scale of the walls, with special
consideration given for varying usage. Finally, this chapter will explore how these walls
function in visible space and how they affect legibility and way-finding.
Fig. 3.2 show the four different categories of patterns being discussed in this study:
vertical, horizontal, a mix of vertical and horizontal, and interlocking. Interlocking
patterns will from here on be defined as those patterns which give the appearance of the
motion of an individual shape being restricted in at least two dimensions. There are other
facing types used that are not considered because the materials are natural and therefore
57


the pattern is blended enough with the surrounding area to be considered irrelevant, such
as grass cover or other vegetated facing finishes.
Mixed vertical
and horizontal
Figure 3.2: Visual effect of different wall facings commonly used for a variety of
applications
Fig. 3.3 presents photographic examples of the patterns presented in Figure 3.2. Figure
3.3a is a vertical pattern achieved through cast-in-place concrete and is located along 1-25
in Denver, Colorado. Figure 3.3b is a horizontal pattern achieved through timber facing
which also has some natural stone facing towards the bottom to help the facing blend into
the background. This wall is located in Colorado as well. Fig. 3.3c is a wall made from
modular blocks which results in a mixed vertical and horizontal pattern and is located in
Grand County, Colorado. Fig. 3.3d is a photograph of an interlocking pattern achieved
through pre-cast concrete panels on a bridge abutment.
58


(a)
(c)
(d)
Figure 3.3: Photographic examples of each facing category using stabilized earth walls
from around Colorado: a) vertical, b) horizontal, c) mixed horizontal and vertical and d)
interlocking
In order to determine how these different patterns are being put to use in different
scenarios, 50 different reinforced structures from around the country were studied and
59


categorized by the following metrics: urban or rural, height, and fa9ade pattern. An
assessment of the length of each wall show the range of facing used to be so varied that
there is little to no statistical significance. Comparing the height of the walls in rural and
urban environments, it is interesting to note that the walls were built with a similar range
of heights. The average height for both rural and urban was around 20 feet tall. This is of
particular interest because special attention was given to fa9ade for much more of the
rural walls, with additional funding spent to use a construction method or material that
would help blend the walls into the natural surroundings. In the urban setting, some
special attention was given to walls through historic areas and college campuses but
outside of that there was little aesthetic consideration. This difference in aesthetic
consideration could be at least partially attributed to differing performance requirements,
particularly in terms of allowable deformations.
Each of the walls in this study used in rural environments were of an interlocking
pattern. This is due mainly to the focus placed on blending these walls into the natural
environment surrounding them. The vast majority of the interlocking patterns were
achieved by utilizing natural stone facing or modular block facing which was stained to
look like the surrounding area. Photographic examples of how interlocking patterns
mimic natural environments are presented in Figure 3.4.
60


For urban environments, a majority of the walls in this study were a mix of horizontal
and vertical lines. This is similar to the approach to walls in rural settings in that it
mimics the surrounding patterns. Figure 3.5 presents three different photographs of the
types of patterns one may expect to find in an urban setting. In fact, one of the walls in
the study was specifically designed to mimic the brick on the building behind it, therefore
blending and making the structure less distracting when viewed along with the
surrounding elements.
Figure 3.5: Photographs of the types of patterns often found in man-made urban settings
Unlike meandering stone walls in a rural setting, reinforced earth structures in urban
setting have the additional requirement of assisting users in finding their way, or at the
very least not hindering that effort. There needs to be some balance between utilizing the
large area of space that is the wall facing and having a pattern that is not distracting at
high speeds. These walls can be large, in fact one particular wall in Riverside, California
is almost a mile long and over 40 feet high. When the structures become large, it may be
advisable to strive for more than just blending. In the particular case just mentioned, the
designers decided to use an artistic design on the wall, as shown in Figure 3.6. These
designs are spread through the length of the wall and depict the history of the surrounding
61


area, therefore allowing this large, intrusive structure to be a part of local culture and
scenery.
Figure 3.6: Artistic finishes done on a wall in Riverside, CA
Thus far in practice, the language for these types of structures typically has to do with
either the material the wall is made of (timber, concrete, etc) or the pattern on the wall as
seen by someone viewing the structure standing still. While the latter approach works for
those structures built in low-speed environments, such as walking trails and low-traffic
dirt roads, it does not take into consideration how the visual effect of the wall may be a
function of the speed at which it is view, for example the difference between standing and
looking at a wall and driving by it at 60 mph along a highway. In order to design walls
for high-speed environments, where it is important to ensure the driver is not distracted, it
is important to understand how humans process patterns within their visual field. When a
62


person is standing still and looking out in front of them, their eyes move across a pattern,
each movement called a fixation until the pattern is processed. The human brain is
capable of processing 4 or 5 fixations in a single second (Noton, 1970). It makes sense
that simple patterns have fewer points of fixation needed to complete and are therefore
processed faster than complicated patterns. Also, patterns which are very familiar to the
person required fewer points of fixation for the viewer to put together. An example of this
is the difference between encountering an animal you have never seen before and coming
across a human in your travels. It would take much longer to process the unknown animal
than to recognize the human form.
The challenge for the singular-pattern approach to wall facade is that the person in the
car driving over 60mph is not standing still. A schematic of how a single point on a wall
moves through multiple visual fields as a car moves through a corridor is presented in
Figure 3.7. In this figure, the must focused field of vision is shown in the darkest gray
color. The transition from focused to peripheral is shown in the medium gray and the
zone of peripheral visual is depicted in the lightest shade of gray. As the walls patterns
moves from focus to peripheral, it needs do so without confusing or distraction the driver.
This is complicated because, for example, a pattern that is confusing and disorienting in a
narrow field of visual could be completely washed out and inconsequential in periphery.
Or a pattern that is easily discernible in a focused field of vision could become confusing
and disorienting in periphery.
63


Figure 3.7: A schematic of how parts of a wall move through various fields of vision as a
car moves through a corridor
Another potential issue to consider is the transition zone between an object in focus as
it moves into the periphery. Figure 3.8 is an example of how objects and shapes can
become confusing and disorienting when a complicated pattern is not directly in the eyes
focal point. Focusing on the cross to the left, the vertical lines on the right are only
somewhat legible. You can tell what they are, but it is not possible to focus on a single
line, and if your brain tries to, it becomes distracting and stressful.
Figure 3.8: Visual effect of varying focus (Intriligator and Cavanagh, 2001)
64


Vertical Wall Patterns
This section addresses vertically patterned wall faces. Firstly, imagine yourself driving
in a car down a highway and having vertical lines on the wall zooming by every so often.
As the lines move through your periphery, they give the appearance of moving quickly.
In fact, lines perpendicular to the direction of movement are often used by highway
designers as a method for making drivers feel as though theyre moving much faster than
they are, which subconsciously makes them want to slow down (FHWA, 1999). Most
people have experienced this subtle approach to slowing traffic in approaches to full-stop
toll booths, where lines are used on the roadway as shown in Figure 3.9.
Figure 3.9: Photograph of an approach to toll booths which utilized bright yellow lines on
a roadway to give the illusion of speed to drivers and remind them to slow down (FHWA,
1999)
65


For the reasons listed above, sharp vertical lines would not be recommended for
highway corridors due to causing some distraction in the periphery. This could lead to an
uncomfortable driving experience at high speeds. However, they could be used at
transition zones to encourage traffic to slow down, much like they are currently used in
toll plazas. Where there exists a decrease is driver site distance, a dangerous curve, or
traffic entering or exiting the highway, vertical lines of varying spacing could and should
be used to encourage slower, safer driving speeds.
Horizontal Wall Patterns
Called horizontal lines because of how they appear on the wall face, what these lines
actually represent are lines in the direction that the car is moving. This is an important
distinction to make because drivers are already familiar with lines moving with traffic,
since that is how all lines that mark the roads appear to the person driving. Lines on the
road have specific colors to let drivers know that they are on the correct side of the road
and specific patterns to let the driver know where the shoulder begins and when it is and
is not safe to pass. It is also possible to see when the road is curving by looking ahead and
noticing the direction in which the horizontal line is moving relative to your current
position. All of these components are shown in Figure 3.10. In this figure, the yellow line
to the left of the drivers field of vision indicates that they are traveling in the correct
lanes. The dashed white line indicates two things, that the adjacent lane is traveling in the
same direction and that it is safe to move between lanes to pass. The curvature in the road
is obvious in the background as the white line can be seen to move towards the right.
66


Figure 3.10: Photograph from FHWA showcasing different roadway markings
This information leads indicates that there is good potential for horizontal lines to be
used on reinforced structures along highway corridors without distracting the drivers,
since drivers are used to these types of patterns in their field of vision already and also in
helping with way-finding.
Interlocking Wall Patterns
These patterns are being grouped with mixed horizontal and vertical because at high
speeds they would generally be perceived as the same; which is a mix of non-uniform
lines. The effectiveness of this pattern is scale-dependent. If the scale were too large, the
problem of distractingly large vertical lines zooming by would be present. If the scale
were small, the pattern would be lost in the speed of movement and would result in a
difficult-to-process blur, given the ability to only process 4 or 5 fixation points per
second. With vertical lines mixed in, it would be difficult to read the movement of
67


horizontal lines and be warned of upcoming changes in road direction or breaks in the
wall that may indicate exiting or entering traffic. If vertical lines exist along the length of
the wall, it would be impossible to use them at interchanges to slow the speed of traffic.
Based on the above information, reinforced soil walls along highway corridors should
consist of mostly horizontal lines, with walls being cut into thirds or fourths to ensure
proper scaling. When encountering a transition area, whether it is an on or off ramp, toll
plaza or simply a dangerous or blind curve, vertical lines can be used to give the
impression that the car is moving too fast which will encourage the driver to slow down.
Additionally, current design standards fall short on another helpful way to increase
legibility along highways, which is the use of color. Currently, traffic signs have varying
colors to indicate whether a sign is informational, for tourism, etc. Recent studies have
looked into the use of the color purple to indicate that the driver is on a toll road
(Funkhouser et al., 2008). By slightly dyeing a concrete mix the wall could have a green
tint in long straight of highway, a touch of yellow when curves or entrance ramps are
present and have a red hue as the driver is approaching a toll booth. While this colors
would be visible from a distance, they would quickly move out of the main focal point of
the field of vision and into the zones where no color is readable, therefore reducing
distraction in the periphery.
This portion of the study was focused on the current state of practice for reinforced
soil wall facade in the United States. It was found that a majority of walls in rural setting
mimic nature by utilizing an interlocking pattern. In urban setting, a majority of the walls
mimicked urban and industrial structure by utilizing a mix of horizontal and vertical
lines, much like brickwork often seen in cities. This blended approach to aesthetic in rural
68


settings is non-critical and seems to function well as-is. The patterns and colors blend and
do not distract from the peaceful tranquility of walking and bike paths.
Although the blended approach to urban settings may work in terms of not being a
distraction, designers could do more to design the walls such that they actually help with
wayfinding. It is no longer adequate to settle for a structure that does no harm when it
is possible to build a structure that actually assists in way-finding and readability.
69


4. Sustainability of GRS
One benefit of utilizing an internally-stabilized reinforced soil system, like GRS, is the
number of design options available to the design engineer. This benefit is not limited to
facing, in fact reduced tensile strength requirements for the reinforcing fabric also leads
to a greater variety of potential reinforcement solutions.
Selection of appropriate facing may lead to a more sustainable design even without
incorporating recycled materials, as the CMU blocks typical for current GRS design can
be lifted into place by hand, with no need for special equipment to lift heavy blocks or
panels. This reduces both the economic and environmental cost as compared with the
GMSE design option.
Although possibilities for incorporating recycled materials into GRS facing are
essentially limitless, this chapter focuses on CMU blocks manufactured with recycled
concrete aggregate in place of virgin coarse aggregate. This approach to reusing crushed
concrete into new concrete mix is well-established and engineers are likely already
familiar with this product, making it a solution that could be incorporated quickly.
Other considerations for sustainability and GRS are addressed in this chapter by
developing the framework for a Life Cycle Assessment (LCA). In this way, the
environmental impact of various components of the structure such as materials,
construction methods and maintenance, can be summed and directly compared with
competing design alternatives.
4.1. Effect of Recycled Content on Concrete Blocks
There is good potential to re-use waste concrete as aggregate for new concrete mixes.
However, this substitution does have an effect on the properties of the concrete. As
70


recycled content in concrete increased, weight and compressive strength tend to decrease.
This decrease in compressive strength may result in a product unsuitable for structural
application. For GRS structures, the compressive strength need only be sufficiently high
to withstand the weight of facing above bottom course. However, the geosynthetic
connection to the facing is frictional which would be negatively affected by a decrease in
the unit weight of facing blocks. Additionally, constructability of the structure may
become more difficult if the weight of the blocks is such that compacting behind them is
difficult without additional external supports.
In order to explore these potential issues, previously published works on how recycled
content affects both weight and compressive strength were summarized (Poon and Lam,
2008; Xiao et al., 2005). In each study, virgin coarse aggregate was replaced with
recycled concrete aggregate in various percentages by weight of total coarse aggregate in
the concrete mix. The results from both studies and presented in Figure 4.1. Figure 4.1a is
a plot of the unit weight of concrete as a function of recycled content. Figure 4. lb is a
plot of the normalized compressive strength as a function of recycled content.
71


26
CO
£24
-a
bD
^
£
^22
^
5
20
0%
Xiao et al. 2005
Poon and Lam 2008


20% 40% 60% 80% 100%
Recycled Coarse Aggregate (%)
(a)
Recycled Coarse Aggregate (%)
(b)
Figure 4.1: Summary of results from studies done on how replacing natural aggregate
with recycled concrete aggregate affects a) unit weight and b) compressive strength of
concrete (Xiao et al., 2005; Poon and Lam, 2008)
72


Unit weight trends from the two studies agree; a decrease in unit weight of
approximately 5% was measured for a concrete mix with 100% recycled coarse
aggregate, as compared with completely virgin materials.
The trend for compressive strength is also similar for each study. The largest drop in
compressive strength was measured for between 0 and 50% recycled content. Beyond
50% recycled coarse aggregate little change in compressive strength was measured. At
100% recycled coarse aggregate, the compressive strength was measured to be no less
than 80% of the compressive strength for a mix with virgin materials.
4.2. Wu and Payeur Analysis
Wu and Payeur (2014) developed an analytical method for determining the threshold
of failure between the reinforcing layer and the facing blocks for GRS structures. This
model is based on static equilibrium and assumes Rankine earth pressures for calculating
lateral thrust against the wall face. A literature review of how recycled course aggregate
affects density and compressive strength of concrete was completed and the result used in
conjunction with the Wu and Payeur model to study how recycled content in concrete
blocks may affect the stability of a GRS structure, specifically for the geosynthetic-block
pullout failure mode. A free body diagram of the problem is presented in Figure 4.2.
73


top of wall face

N
Fm
Ni-i
Figure 4.2: Free body diagram of reinforcement frictional connection with blocks (Wu
and Payeur, 2014)
From the diagram in Figure 4.2 and assuming static equilibrium, the vertical forces are
summed as:
Wi + Nj_! Ni+1 + Fbi = 0
[Eq. 11]
and the horizontal forces can be summed as
Ti + Fi+1 Fi_! Pi = 0 [Eq. 12]
where W represents the weight of the block, N represented the normal force, F represents
the shear force between the block and soil mass, T is the tensile force in the
reinforcement connection, and P represents the lateral thrust from the soil behind the
facing element. For this study, the force acting on the wall face was assumed to be
Rankine earth pressure and can be written as:
74


Pi = Ka(ysZi +q)(2h)
[Eq. 13]
where Ka is the Rankine active earth pressure coefficient, ys is the unit weight of soil, z is
the depth, q is the surcharge load applied and h is the height of each facing block.
From these assumptions and applying Coulomb friction theory, the following
equations were developed to determine driving and resisting forces for geosynthetic-
block frictional failure.
where n is the ratio of spacing to block height, 5gb is the friction between geosynthetic
and block, B is the depth of the block, and 5sb is the friction between the block and the
retained soil.
The wall modeled in this study is NLB wall, 4.65 meters in height with a dry-stacked
CMU facing with nominal block dimensions of 0.2 m x 0.2 m x 0.4 meters. There was no
batter and the geosynthetic is assumed to be a woven polypropylene with 0.2 m vertical
spacing of reinforcement. The unit weight of the backfill was 21.6 kN/m3 and the friction
angle of the soil was assumed to be 40 degrees. A full list of parameters and values used
for this study is presented in Table 4.1.
[Eq. 14]
[Eq. 15]
75


Table 4.1: Parameters used for recycled concrete block analysis (Wu and Payeur, 2014)
Parameter Value Unit
n 1
h 0.19 m
Ka 0.217
$ 40 degrees
ys 21.6 kN/m3
B 0.19 m
yb 1032 kN/m3
5gb 30 degrees
5sb 25 degrees
q 12 kPa
Height: 4.65 m
4.2.1. Results
Equations 14 and 15 were applied to the model wall described in Table 4.1 to
determine whether resisting forces were capable of resisting the driving forces for CMU
block facing with recycled aggregate content. For this study, the worst case scenario is
the lightest block, which coincides with a block manufactured with 100% of the coarse
aggregate replace with recycled concrete aggregate.
Designing with CMU blocks generally requires grouting the top three courses of block
to ensure adequate friction between the blocks to resist movement. Figure 4.3a is a plot of
resisting and driving forces through the full height of the model wall without grouting.
Figure 4.3b is a plot of the results with grouting.
76


Force (kN)
(a)
Force (kN)
(b)
Figure 4.3: Results from analysis to determine if friction offered by lightweight recycled
CMU blocks is adequate to resist lateral earth pressures: a) no grout and b) top three
courses grouted for added stability
77


Looking at the results presented in Figure 4.3a, there may be some question as to the
stability of the structure when the top three courses of block are not grouted. In Figure
4.3b the effect of the grouting can be seen in the non-linearity of the results up to the 4th
course of block at a depth of approximately 0.6 meters. The added weight of the grout
ensures that driving forces do not overcome resisting forces at any point along the height
of the wall. It should also be considered that the lateral earth pressures assumed for this
model are quite possibly much higher than actual earth pressures, and the plots in Figure
4.3 represent a worst case scenario.
Despite the obvious decrease in both strength and weight due to the increase in
recycled content, this study suggests that blocks with large amounts of recycled content
still have adequate weight and strength for use as facing in GRS structures. Additional
considerations include variability in recycled CMU blocks as compared with those made
with virgin materials and the potentially high energy cost of manufacturing the recycled
concrete aggregate, which requires not only transportation of old materials, but crushing
and sorting before they can be used in engineering applications. Additionally, when
incorporating these recycled materials into a sustainability analysis some consideration
must be given to differences in block loss due to breakage or mishandling, as these
difference could have a large effect on how many blocks are required for construction.
4.3. Additional Recycled Materials
Depending on performance requirements and where the wall is being used, there is a
potential to incorporate a wide variety of sustainable, recycled, or reused materials. This
effort can be as simple as filling whole tires with soil to build up a crib facing
(Jayawickrama, 2000) or as complicated as incorporating specially engineering
78


corrugated metal facing made with recycled materials, an aesthetic that was presented in
Chapter 2 of this paper (Wadey and Idrees, 2014). Although the use of metallic facing
could help reduce the environmental cost of a project by incorporating some recycled
metal content, there are a wide variety of uses for recycled metal, so from an
environmental impact standpoint, it may be more beneficial for a project such as a short,
NLB GRS wall to integrate materials like tire where its possible. A schematic of a
typical wall section for a reinforced soil structure with a tire facing is presented in Figure
4.4.
Figure 4.4: Schematic depicting section view of a reinforced earth wall with whole tire
facing (Jayawickrama, 2000)
79


For this type of structure, the tops are cut off of the tires and the tires are filled with
soil then covered with some other type of facing finish, such as shotcrete or a
geosynthetic fabric. This use of whole recycled tire may be preferable over shredded tire
incorporated into the backfill since little to no energy is required to prepare the material.
Using whole tire as a facing element is a good match for GRS structure as compared with
GMSE since lateral pressure on the facing is small and no connection to the reinforcing
elements is required. However, large whole tires can be heavy, which could make them
difficult to ship and cumbersome to construct. While this method is an excellent way to
reuse the ever-growing stockpiles of retired tires, they would not be appropriate for large
or high-load walls and may be difficult to transport to remote sites.
Recycled metallic facing such as the bridge abutments in Yukon Territory presented in
Chapter 2 may be more appropriate for large-load and other critical structures because the
strength properties of the facing would be easier to determine. However, there is some
energy that goes into processing the recycled metal as well as the manufacturing of the
corrugated plates. Its not always accurate to assume that a recycled material
automatically results in a sustainable structure. In order to determine the actual
environmental impact of a material or process, the best option is to perform a life-cycle
assessment (LCA) as discussed in Section 4.4.
4.4. Comparative Life Cycle Assessment
4.4.1. Background
Despite the current discussion worldwide concerning global warming, the civil
engineering community has been slow to embrace environmental considerations for civil
projects. This is likely due to the fact that performance of a civil structure the primary
80


design concern, but could also be affected by the fact that the majority of structures are
built using existing design standards which may take some time to adjust for the current
state of knowledge. However, there has been some movement towards sustainability of
civil structures, as evidenced by new environmental rating systems geared towards these
types of projects such as Envision and Greenroads. Envision focuses heavily on
incorporating environmental considerations early in a project and can be used outside of
the actual rating system as a guideline for incorporating environmental considerations
into projects at a variety of project stages (ISI, 2012). Greenroads is a system developed
in Washington State (Soderlund et al, 2008) and is similar to Envision in that early
environmental considerations are encouraged and a positive score is based more on the
effort put into incorporating sustainable components and less on quantifying the
environmental impact of individual projects. While green rating systems can be helpful in
showing that the environmental implications of a project were considered, a quantitative
approach may be better due to the ability quantify environmental impact in such a way
that multiple projects could be compared regardless of their participation in specific
rating systems.
As evidenced by the American Society of Civil Engineers report on the state of our
national infrastructure released in 2013, our infrastructure in this nation is in disrepair. A
failing grade of D was awarded to our transportation infrastructure specifically and the
recommendation from the ASCE was that major upgrades and repairs need to be
completed to keep our infrastructure safe for public use. This assessment is likely due to
the age of our infrastructure and the fact that much of it was put into place in the mid-
twentieth century. It was in 1956 that President Eisenhower signed into law the Federal -
81


Aid Highway Act which allotted $25 billion to build 41,000 miles of our national
highway system. The transportation infrastructure built during this period of rapid growth
is currently approaching the end of its design life and requires investment or order to
remain functional. In some cities with disproportionate growth compared to other parts of
the country, such as Denver, Colorado, the issue with the current infrastructure is not
related to age of the structures so much as the age of the design. In these cases, additional
lanes may be required to handle the volume of traffic in multiple corridors as they move
through the metro area. As corridors are widened and roadways are repaired, there is
potential to replace these structures with a more sustainable design option. The question
remains, however, about how to determine which design option is the most sustainable.
For example, one option may use more concrete and another more backfill. One may
require more water or another may require more maintenance.
A method for quantifying the environmental impact of different materials and
processes is a Life Cycle Assessment (LCA). Though not widely used for civil
engineering applications, LCAs are commonly used in environmental assessments and in
industries such as alternative fuel development when environmental impact is a major
design consideration. An LCA takes into account both inputs (i.e. materials and energy)
and outputs (i.e. emissions into air or water and byproducts) and determines to total
environmental impact. This includes both upstream considerations, such as transportation
of raw materials, and downstream considerations, such as recyclability of materials at end
of life. Figure 4.5 is an example of how emissions results from a life cycle inventory
analysis may be incorporated into both mid-level and damage categories.
82


Midpoint
categories
Damage
categories
Human toxicity
Respiratory effects
Ionizing radiation
Ozone layer depletion
Photochemical oxidation
Aquatic ecotoxicity
Terrestrial ecotoxicity
Aquatic acidification
Aquatic eutrophication
Terrestrial acid/nutr
Land occupation
Global warming
Non-renewable energy
Mineral extraction
Human Health
Ecosystem Quality
Climate Change
(Life Support Systems)
Resources
Figure 4.5: Flow of LCI results to midpoint categories to damage categories
There have been previous life cycle assessments completed on reinforced and retained
soil walls (Fraser et al., 2012; Frischknecht et al., 2013, Damians et al., 2015). However,
these studies focused on how geosynthetic reinforced walls compare to concrete gravity
walls. Figure 4.6 is a schematic of each structure used in this comparison.
83


Full Text