REINFORCED SOIL STRUCTURES

Volume II.

o
u.s.

of

~po.r'bwn-t

Transpor-t~~on

F.~r~l "'ghWQY

Adl'\Inls-trQ ~on

Summary of Research and

Systems Information

REINFORCED SOIL STRUCTURES

VOLUME II. SUMMARY OF RESEARCH AND SYSTEMS INFORMATION

BY
BARRY R. CHRISTOPHER
SAFDAR A. GILL
JEAN-PIERRE GIROUD
ILAN JURAN
JAMES K. MITCHELL
FRANCOIS SCHLOSSER
JOHN DUNNICLIFF

PREPARED FOR

Federal Highway Administration

Office of Engineering and Highway Operations
Research and Development
McLean, virginia

BY
STS CONSULTANTS, LTD.
NORTHBROOK, ILLINOIS

TECHNICAL REPORT STANDARD TITLE PAGE

I. REPORT NO.
FHWA-RD-89-043

2. GOVERNMENT ACCESSION NO.

3. RECIPIENT'S CATALOG NO.

5. REPORT DATE
November, 1989

4. TITLE AND SUBTITLE

REINFORCED SOIL STRUCTURES
VOLUME II. SUMMARY OF RESERACH AND
SYSTEMS INFORMATION

6. PERFORMING ORGANIZATION CODE

7. AUTHOR(S) Barry R. Ch ristopher, Safdar Gi II,

Jean-Pierre Giroud, llan Juran, James K. Mitchell,
Francois Schlosser and John Dunnicliff

8. PERFORMING ORGANIZATION REPORT NO.

9. PERFORMING ORGANIZATION NAME AND ADDRESS

lO. WORK UNIT NO.

NCP3E3bOO22

STS Consultants, Ltd.

III Pfingsten Road
Northbrook, Illinois 60062

11. CONTRACT OR GRANT NO.

DTFH61-84-C-00073
13. TYPE OF REPORT & PERIOD COVERED
Final Report (9/84 through 10/88)

12. SPONSORING AGENCY NAME AND ADDRESS

Office of Engineering and Highway Operations R&D
Federal Highway Administration
6300 Georgetown Ave.
McLean, Virginia 22lO 1-2296

14. SPONSORING AGENCY CODE

15. SUPPLEMENTARY NOTES

FHWA Contract Manager (COTR): A. F. DiMillio (HNR-30)
FHWA Technical Consultant: J.A. DiMaggio (HNG-31)
16. ABSTRACT
Volume II was essentially prepared as an Appendix of supporting information for Volume I.
This volume contains much of the supporting theory and a summary of the research used to
verify the design approach contained in Volume I, as well as general information
concerning proprietary reinforced soil systems. The information provided in this volume
is not required for design evaluation and as such Volume I can stand alone.
The basis for this volume was the NCHRP 290 report on "REINFORCEMENT OF EARTH
SLOPES AND EMBANKMENT" (Mitchell and Ville!, 1987) and the research program performed
as part of the contract to develop the design guidelines contained in Volume I.
A summary of that research program is contained in the Introduction section of this
volume.
17. KEY WORDS
Abutments, Construction, Design,
Embankment. Instrumentation, Nailing.
Reinforcement, Slopes, Soil
Specifications, Stabilization, Walls

19. SECURITY CLASSIF.

Unclassified

18. DISTRIBUTION STATEMENT

No restrictions. This document is available
to the public through the National Technical
Information Service. Springfield.
Virginia 22161

20. SECURITY CLASSIF.

Unclassified

21. NO. OF PAGES

158

22. PRICE

REINFORCED SOIL STRUCTURES

VOLUME II. SUMMARY OF RESEARCH AND SYSTEMS INFORMATION
TABLE OF CONTENTS
PAGE
INTRODUCTION

CHAPTER 1. DESCRIPTION OF SYSTEMS

1.0 INTRODUCTION
2.0 PLACED SOIL REINFORCED SYSTEMS
a.
Reinforced Earth
b.
Hilfiker Retaining Wall
c.
VSL RETAINED EARTH System
d.
Mechanically Stabilized Embankment (MSE)
e.
Georgia Stabilized Embankment (GASE)
f.
"Websol" System
g.
The York Method
h.
The AndaAugmented Soil System
i.
Tensar Geogrid System
j.
Other Polymeric Geogrid Systems
k.
Geotextile Reinforced Systems
1.
Gabion Reinforced Fill Wall System
IN-SITU REINFORCED SYSTEMS
3.0
a.
Soil Nailing
b.
Micro-piles or Reticulated Root Piles
c.
Composite In-Situ Systems
4.0
MULTIANCHORED SYSTEMS
a.
American Geo-Tech System
b.
Tension Retaining Earth System (TRES)
c. Anchored Earth
d.
Ladder Wall
5.0 ALTERNATIVE SYSTEMS
a.
Gravity Walls
b.
Cantilever Walls
c.
Grouted Anchor Walls
d.
Deadman Anchored Walls
e.
Composite System

6
6
6

CHAPTER 2. PULLOUT RESISTANCE EVALUATION

1.0 INTRODUCTION
2.0 ESTIMATE OF FRICTIONAL PULLOUT RESISTANCE
a.
Inextensible Linear Strip Reinforcements
3.0 ESTIMATE OF PASSIVE PULLOUT RESISTANCE
a.
Dete rll\ina ti on of cx",
b.
Determination of F
c.
Determination of t~e scale effect
correction factor cx
4.0 ESTIMATE OF THE PULLOUT CAPACITY OF
COMPOSITE REINFORCEMENTS COMBINING
FRICTIONAL AND PASSIVE SOIL RESISTANCES
a.
Determination of F* for composite
reinforcements
5.0 SUMMARY OF PULLOUT DESIGN PARAMETERS

ii

11
13
16
16
19
22
24
24
27
27
30
30
30
33
34
34
34
36
36
40
40
40
42
46
46
46
50
50
51
51

59
63
63

67
67
70
73

TABLE OF CONTENTS (continued)

PAGE
6.0
7.0

8.0

EVALUATION OF ALLOWABLE DISPLACEMENTS

FROM PULLOUT TESTS
SOIL NAILING - ESTIMATE OF PULLOUT CAPACITY
a.
Grouted nails
b.
Driven nails
c.
Jet grouted nails
d.
Estimate of the pullout resistance
LONG TERM PERFORMANCE AND CREEP
CONSIDERATIONS
a.
Soil Nailing

CHAPTER 3. REINFORCED FILL WALLS

1.0 INTRODUCTION
2.0 LIST OF RECENT RESEARCHES, INCLUDING
FHWA PROGRAM
a.
General reports on soil reinforcements
b.
FHWA research program on reinforced
soil wall
c.
Reinforced Earth Company's Research
d.
Research on polymeric reinforcement
3.0 SUMMARY OF FHWA RESEARCH RESULTS
a.
Small-scale models
b.
Full scale field wall
c.
Summary of parametric study on
reinforced soil walls
4.0 INCLINATION OF THE THRUST AT THE BACK
OF THE WALL
5.0 DISTRIBUTION OF THE VERTICAL STRESS a v
AT THE BASE
6.0 INFLUENCE OF SYSTEM STIFFNESS ON THE
HORIZONTAL STRESS IN THE REINFORCED
SOIL SECTION
7.0 INFLUENCE OF THE LENGTH OF THE
REINFORCEMENTS
8.0 LATERAL DEFORMATION AT WALL FACE
9.0 RECENT RESULTS ON SEISMIC BEHAVIOR OF
REINFORCED SOIL WALLS

73
74
74
77
77
77
80
80
91
91
91
91
91
92
92
93
93
93
96
106
108
108
113
116
116

CHAPTER 4. REINFORCED ENGINEERED SLOPES

1.0 INTRODUCTION

119
119

CHAPTER 5.
1.0
2.0
3.0
4.0
5.0

127
127
127
130
135

SOIL NAILING
INTRODUCTION
EMPIRICAL DESIGN EARTH PRESSURE DIAGRAMS
FINITE ELEMENT ANALYSES
KINEMATICAL LIMIT ANALYSIS DESIGN METHOD
STABILITY ANALYSIS OF SOIL NAILED
RETAINING STRUCTURES
a.
Local stability analysis
b.
Global stability analysis
c.
Evaluation of global stability
analysis procedures

REFERENCES

139
139
141
147
151

iii

LIST OF FIGURES

PAGE

1
2
3
4
5
6

8
9

10
11
12
13
14
15
16

17
18
19
20
21
22
23
24
25
26
27
28
29

30
31
32
33

34
35

36
37

Principal elements of reinforced fill walls

Segmented facing panels used for
Reinforced Earth walls
Hilfiker Welded Wire Wall system
Hilfiker's Reinforced Soil Embankment Wall system
VSL Retained Earth Wall system
Facing panels for mechanically stabilized
embankment wall system
Georgia stabilized embankment facing panel
and reinforcement attachment
The Websol system
Reinforcement and facing assembly for
the websol system
The York system
The Anda augmented soil system
Tensar geogrid reinforced soil wall
Geotextile reinforced wall
Types of geotextile reinforced walls
Maccaferri gabion reinforced fill
and gravity wall systems
Uses of soil nailing and micropiles
American Geo-Tech system
Tension Retaining Earth system
TRRL achored earth system
Ladder wall sloped face
Ladder wall with vertical face
Crib lock retaining wall
"Doublewal" retaining wall
Typical element of Evergreen retaining wall
"Evergreen" retaining wall - front view
Stresswall details
Grouted anchor wall
Deadman anchored wall
Influence of reinforcement type and overburden
stress on apparent friction coefficient
Reinforced earth desing ~* values for smooth
and ribbed strips
Stress shear displacement curve from
direct shear test
Variation of displacemens along a woven
polyester strip during a pullout test
Experimental procedure to determine a
for geotextile sheets
Numerical procedure simulating pullout tests
to establish a-L curve for specific soil type
and reinforcement properties
Definition of bearing stresses on
transverse elements
Theoretical relationship and experimental
results bearing stress vs. soil friction angle
Pullout test results - VSL bar mats and
anchorage factor design values

10
12
14
15

17
18
20
21
23
25
26
28
29
31
32
35
37
38
39
39

41
43
44
44

45
47
48

52
53

55
56

58
60

62

65
66

LIST OF FIGURES (continued)

PAGE

38
39

40

41

42
43
44
45

46
47

48
49
50
51
52
53a
53b
54
55
56
57
58
59

60
61
62
63

64
65

Pullout test results - welded wire meshes

Relation between pullout force and
vertical stress
Peak pullout displacement for geosynthetic
reinforcements vs. reinforcement length
Interpretation procedure for pullout tests on
extensible inclusion
Pullout test results
Pullout test results on driven nails
in granular soils
Comparison between measured and estimated
values for ultimate lateral shear stress
Anchor tension test for determination of
critical creep load
Modeling creep of anchors in clays
phases of creep for a typical geotextile tested
without soil confinement at constant load
and temperature
Load versus total creep strain for Tensar SR2
Creep strain rate against total creep strain
Confinement effect on creep behavior of
non-woven geotextiles
Effect of the extensibility of the reinforcements
on the K coefficient
Influence of extensibility on distribution of
maximum tension
Geometry and F.E.M. results for the baseline case
F.E.M. results continued for the baseline case
Maximum reinforcement tensions for strip loading
cases by F.E.M. and conventional methods
Thrust at the back and vertical stress
distribution at the base of reinforced soil walls
Variation of K/K versus depth for some
field wall
a
Variation of K/Ka versus depth for Sr =
20-100 k/ft2
Variation of K/K versus depth for S
500-2000 k/ft2
a
r

~~~i~~io~e~~f~~~edv~~~~~ ~:l~h

for Sr > 2500 k/ft2

Horizontal stress comparison for high wall
Influence of change in reinforcement length on
lateral deformation at the face of the wall
anticipated during construvction
Distribution of maximum tension with depth in
embankment 1
Distribution of maximum tension with depth in
embankment 2
Distribution of maximum tension with depth in
embankment 3

68
69
72
75
78

79
82

84
85

86
88
89

90

94
95

102
103
105
107
109

110
III

112
114
115
117
121

122
123

LIST OF FIGURES (continued)

PAGE
66
67
68
69
70
71
72

73
74

75
76
77
78
79

80

Distribution of maximum tension with depth in

embankment 4
Distribution of tension along different levels
of reinforcement in embankment 1
Deformation at 12.5 feet from the toe of
embankment 1
Horizontal displacement of nailed soil walls
Empirical earth pressure design diagram
Experimental data and theoretical predictions
of tension forces
Effect of the bending stiffness and the
inclination of reinforcement on the facing
displacements
Effect of bending stiffness of the inclusions
on nail forces
Kinematjcal limit analysis approach
Horizontal subgrade reaction as a function of
the soil shear parameters
Bending of a rigid inclusion
Force equilibrium method for global stability
analysis of nailed soil retaining structure
Location of critical failure surface
Multicriteria slope stability analysis method
Predicted and observed locus of maximum tension
forces in nails

124
125
126
128
129
131
133
134
137
138
142
144
145
148
149

LIST OF TABLES
PAGE

1
2
3

4
5

Summary of reinforcement and face panel

details for various reinforced soil systems
ultimate lateral shear stress data for
preliminary design of soil nailing
List of full scale reinforced soil walls
constructed for the FHWA program
Comparison of FEM analysis cases
Full scale reinforced slope experiments
constructed for the FHWA program

81
92
97

119

INTRODUCTION
1.0

SCOPE AND ORGANIZATION

This volume of the Reinforced Soil Structures manual provides

supporting information for the different types of soil
reinforcement systems and the design approaches contained in
volume I, Design and Construction Guidelines.
Following a
review of the contents of volume II, a brief summary of the
research performed and primary conclusions will be provided. The
remaining sections were prepared to coincide with the chapters in
volume I for ease of reference.
The Description of Systems
section, which follows the Introduction section, provides
additional information on the specific reinforced soil systems
reviewed in chapter 1 of volume I.
The next section, Pullout
Resistance Evaluation, provides supporting information for
determining pullout design parameters for specific reinforcement
types using both empirical relations and pullout test results.
The final three sections provide support information used to
develop the design guidelines for the different reinforced soil
systems, reinforced fill walls, reinforced engineered slopes and
soil nailing for in-situ reinforcement, respectively.
2.0

SUMMARY OF RESEARCH

The "Behavior of Reinforced Soil" project sponsored by the

Federal Highway Administration was performed to develop
comprehensive guidelines for evaluating and using soil
reinforcement techniques in the construction of retaining walls,
cut slopes, and roadway embankments.
The work encompassed a
literature review, laboratory tests, full scale field tests,
analytical evaluation, confirmation of design parameters and
equations, and the preparation of construction procedures and
practices.
The lab phase of the study included measuring reinforcement
variables and stress distribution patterns, determining the types
of materials suitable for soil reinforcement and developing the
standard lab test procedures for obtaining the design parameters.
This task was carried out by centrifuge tests at the University
of California, Davis, reduced scale model tests at Ecole
Nationale de Pont et Chaussees, France, and pullout, direct
shear, and triaxial methods for evaluation of design parameters
at STS Consultants, Ltd.
Centrifuge tests on small reinforced soil models using scaled
down reinforcements based on similitude requirements were carried
out to study the behavior of reinforced soil walls as it is
affected by such factors as reinforcement extensibility, external
loading, full height facing panels, and foundation
compressibility.
From centrifuge tests on 47 small models, it
was learned that the behavior of walls reinforced with a wide
variety of materials, both extensible and inextensible, is

similar at failure, and

over y conservative.

rior conventional desi n methods rna

be

Reduced scale model tests were performed on model reinforcements

representing metal strips, plastic strips, plastic grids, woven
and nonwoven geotextiles and anchors.
The models were
constructed using a step by step method similar to actual wall
construction. Models were constructed to heights necessary to
fail internally be breakage or deformation to qualitatively
evaluate failure conditions.
Some of the models were
instrumented to evaluate the distribution of stress. The model
test results were used to evaluate the location of the failure
surface and the magnitude of lateral deformation for
reinforcements of different extensibility and surface
characteristics. The results indicated that onl
materials such as nonwoven geotext1 es an
19 y e orma e
lastic trul acted as extensible reinforcement modeled b a
Ran ine stress distri ution. A
other extens1 e mater1als,
woven geotextiles, plastic grids, and high tenacity plastic
strips were found to behave more like inextensible reinforcement
which is similar to the findings of the small scale centrifuge
test program.
Pullout tests were conducted to evaluate soil reinforcement
interaction for various types of reinforcement under varying
normal load and soil conditions.
Based on a literature review
and tests in a small pullout box on model reinforcement, a
relatively large (4.4 ft x 2.3 ft x 1.5 ft) pullout box was
developed which significantly reduced some of the boundary
influences of previous devices.
consistent test procedures were
also developed that were subsequently used as a model to prepare
an ~STM standard for pullout testing.
The procedures were used
to evaluate 12 different reinforcement materials, including metal
and fiber strips, bar mats, wire mesh (both welded and woven),
extruded and welded geogrids, slit film and coarse woven
geotextiles and needle punches and heat bond~d nonwoven
geotextiles. Tests were also performed on epoxy coated
reinforcement to evaluate the influence of epoxy on pullout
resistance.
Iterative strain measurements were made along the
length of extensible reinforcement to evaluate in soil strain
response and stress transfer.
The results were used to develop
more consistent pullout evaluation procedures for the various
types of reinforcement, as was presented in chapter 2 of volume
I. The approach uses a single pullout coefficient F* and a
geometric factor a to evaluate pullout for any type of
reinforcement. The procedur~s allow for continual updating of
the interpretive procedures for the pullout factors as more data
is developed without modifying the design approach.
The
determination of F* and a will be covered in detail in the
Pullout Resistance Evaluation section of this volume.
Full scale field tests were constructed and monitored by STS
Consultants, Ltd.
The field tests include construction and
monitoring of eight walls, each 35 ft long and 20 ft long and 20

ft high, and four slopes, 50 ft wide and 25 ft high.

The field
instrumentation program was developed to evaluate important
internal stability design parameters.
These include locating
maximum stresses in the reinforcement, lateral stress
distribution, lateral movement of the faces during and after
construction, stress distribution from surcharge and footing
loads, and stress relaxation.
Inclinometers were installed in
the active zone of all slopes and walls.
The base was also
optically surveyed.
Stress in the reinforcement was evaluated
through bonding resistant strain gages mounted on the
reinforcement.
Two pullout tests were also performed on each
wall.
The field wall results are discussed in detail in the Reinforced
Fill Wall section of this volume.
In summary, they indicated
that all reinforcement when designed using a unified approach
behave in a similar predictable manner. When the density of
reinforcement (amount of reinforcement per area of the reinforced
section) is similar, the principal difference in performance can
be attributed to the extensibility of the reinforcement.
The
construction of a wall to failure using the same design approach
as was used for the other structures indicated the conservative
nature of the current design procedures. Existing design methods
for reinforced embankment slopes were similarly found to be
conservative.
However, all things considered, variability in
construction procedures, fill and backfill, foundation material
and construction control would suggest only moderate changes in
current design procedures at this time. Rather, this information
should be used to improve design consistency.
Five large centrifuge models (1:12 scale factors) were tested to
model the behavior of four of the full scale instrumented walls
built as a part of the project.(62)
Good agreement was obtained
between reinforcement tensions developed in the centrifuge models
and those in the prototypes. This agreement adds credibility to
the centrifuge modeling technique for study of reinforced soil
structure.
The results showed that the maximum tensions
developed in the reinforcements at working stress levels depend
both on the reinforcement stiffness and the relative movement
between the soil and reinforcement.
Finite element analyses gave
good predictions of the reinforcement tensions in the five
different large model walls.
A discrete finite element program,

SSCOMP, was used to conduct a

parametric study of reinforced soil walls. The effects of
variations in structure geometry, loading, foundation soil type,
wall facing type, and soil compaction on internal stresses and
deformations were determined. The computer program was used with
some confidence, because it was successfully used to predict the
stress in reinforcements of a number of full scale walls for
which measured values were available and in a number of large
centrifuge test models that were tested during another phase of
this project.

The results of the finite element program are discussed in detail

in the Reinforced Fill Wall section of this volume. Among the
most significant findings from these analyses were that: (1)
prior design methods may underpredict reinforcement tensions when
there is significant compaction of the backfill during
construction; (2) predictions by FEM and conventional analysis
methods do not always give comparable results, especially for
nonstandard wall conditions: and (3) wall face deformations can
be significantly changed through variations in reinforcement
length and spacing.
The results of the parametric study provided insights and
understanding that were helpful in the development of the design
method and recommendations presented in chapter 3, volume I.
In addition to the finite element method parametric study, a
simplified analytical method was developed for estimating the
lateral earth pressure coefficient in reinforced soil structures
as a functio~ of the reinforcement stiffness, the soil stiffness f
and the shear transfer between the soil and the reinforcement.(1
Computer programSSCOMP was evaluated for prediction of stresses
and deformations in reinforced soil structures when a more
complete analysis is needed than can be obtained by using the
simplified method.
The simplified analytical method successfully predicted the
reinforcement tensions in eleven full scale reinforced soil walls
and agreed with the results obtained using the finite element
method.
Good FEM predictions were made of the reinforcement
tensions developed in the full scale instrumented walls built as
part of the FHWA project.
Predictions of deformations were less
successful.
The final task was to use the research to verify and unify
existing design methods and incorporate them into the Design and
Construction Guidelines volume. As will be discussed in more
detail in the Reinforced Fill Wall section of this manual, the
research found that external design could be modified by
inclining the thrust at the back of the wall, at least for
inextensible reinforcement.
Insufficient data was available to
justify this approach for extensible reinforcement. This
modification will allow for shorter base widths in the reinforced
zone.
For internal stability, a simplified approach was
developed around the stiffness of the reinforced zone.
The
approach allows the influence of extensibility and density of
reinforcement to be directly analyzed while decreasing the
complexity of some of the previous models in terms of the
distribution of stress in the reinforced-zone.
Finally, a first
order approximation method of the anticipated lateral deformation
in the wall was developed empirically based on the extensibility
of the reinforcement and the reinforcement length to height
ratio. A simple method with a good experimental base was not
previously available.
The proposed deformation response method

could later be theoretically improved by incorporating stiffness

factor into the analysis.
These procedures were incorporated
into a step by step design approach.
As will be reviewed in the Reinforced Engineering Slope section
of this volume, the research substantiated the use of a limit
equilibrium approach for design of reinforced engineered slopes.
A step by step method is given based on classical rotationally
stability analysis with a chart procedure used for a rapid check
of the results.
The info~mation used to develop the method for evaluating nailed
soil retaining structures contained in chapter 6 of volume I was
developed in a separate FHWA project.(SO)
Comments and
supporting information pertaining to the design recommendations
in volume I are included in the Soil Nailing section of this
volume.

CHAPTER 1
DESCRIPTION OF SYSTEMS
1.0

INTRODUCTION

Reinforced soil systems can be classified in three categories:

placed soil reinforced systems, in-situ reinforced systems, and
multianchored systems. Systems belonging to these categories are
described below in section 2, 3, and 4, respectively.
Many of
the systems of reinforced soil walls are patented or proprietary.
Many companies provide a complete package of services including
design, preparation of plans and specifications for the structure
and supply of the manufactured wall components.
They may also
provide erection assistance to the contractor during start up of
construction.
The various systems which are being offered have different
performance histories, and this sometimes creates difficult in
adequate technical evaluation. Methods for handling the matter
of specification and obtaining the most cost competitive and
technologically acceptable system are covered in the Manual.
Nevertheless, it should be recognized that some systems are more
suitable for low walls, and some are applicable for remote areas,
while others are more suited for urban areas with more rigid
requirements.
Brief descriptions of the various proprietary systems under each
category are included herein.
Systems that are not discussed for
the reasons given in chapter 1 of Volume I are succinctly
described in section 5 devoted to alternative systems.
2.0

PLACED SOIL REINFORCED SYSTEMS

Reinforced fill structures embody three basic components, namely:

engineered fill, also called engineered fill (reinforced soil
volume, or backfill), reinforcement and facing elements which
prevent surface erosion and give an aesthetically pleasing face
(figure 1).
There are a variety of systems marketed by different
specialty companies which use different types of reinforcements
and facing elements.
In practically every case, a granular
material is used within the reinforced soil volume with little
variation in its specified quality or gradation.
The mechanism
of stress transfer between the reinforcement and the backfill is
another variable in the different systems; for example, by
friction or by passive resistance or combinations thereof.
Some
systems are similar in principle but different in reinforcement
materials used; for example, steel strips and plastic strips.
Descriptions of the presently available and commonly known
systems and materials are given below.
Table 1 provides a
summary of the various systems.

Reinforced Soil
Volume

Finished Grode

Original
Ground
Surface
: . _ ...............

.,.

_ J' . . . . ....

Limits of
Construction Excovotioo

Levelling Pod

Reinforcing
Strips

ex Footing

Figure 1.

principal elements of reinforced fill walls.

Summary of reinforcement and face panel details for

Table 1.
various reinforced soil systems.
System Name

Reinforcement Detail

Typical Face Panel Detail 1

Reinforced Earth: (The

Reinforced arth Company
1700 N. Moore St.
Arlington, VA 22209-1960)

Galvanized Ribbed Steel strips:

0.16 in (4 mm) thick; 2 in (50 mm)
wide. Epoxy coated strips also

Facing panels are cruciform

shaped precast concrete 4.9 ft
x 4.9 ft x 5.5 in (1.5 m x 1.5
m x 14 em). Half size panels
used at top and bottom.

VSL Retained arth

(VSL Corporation,
101 Albright Way,
Los Gatos, CA 95030)

Rectangular
plain steel
(61 em x 15
may have 4,
bars. Epoxy
available.

Mechanically Stabilized
mbankment. (Dept. of

Rectangular grid, nine 3/8 in (9.5 mm)

diameter plain steel bars on
24 in x 6 in (61 em x 15 em) grid.
Two bar IIIilts per panel. (connected to
the panel at four points).

Precast concrete; rectangular

12.5 ft (3.81 .) long, 2 ft
(61 em) high and 8 in (20 cm)
thick.

Rectangular grid of five 3/8 in

diameter (9.5 rnrn) plain steel
bars on 24 in x 6 in (61 cm x 15 em)
grid 4 bar mats per panel

Precast concrete panel;

rectangular 6 ft (1.83 m)
wide, 4 ft (1.22 m) high with
offsets for interlocking.

Welded wire mesh, 2 in x 6 in

grid (5 em x 15 cm) of W4.5 x
W3.5 (.24 in x .21 in diameter),
W7 x W3.5 (.3 in x .21 in), W9.5
x W4 (.34 in x .23 in), and W12 x
~ (.39 in x .25 in) in 8 ft wide

Welded wire mesh, wrap around

with additional backing mat and
1.4 in (6.35 mml wire screen at
the soil face (with geotextile
or shotcrete, if desired).

Transportation, Div. of

ngineering Services,
5900 Folsom Blvd.,
PO Box 19128
Sacramento, CA 95819).
Georgia Stabilized
Embankment (Dept. of
Transportation,

State of Georgia,
No. 2 Capitol Square
Atlanta, GA 30334-1002)
Arlington, VA 22209-1960
Hilfiker Retaining Wall:
(Hilfiker Retaining walls,
PO Drawer L
Eureka, CA 95501)

grid of WI1 or mo
bars, 24 in x 6 in
em) grid. Each mesh
5 or 6 longitudinal
coated meshes also

Precast concrete panel. Hexagon

shaped, (59-1/2 in high, 68-3/8
in wide between apex points,
6.5 in thick (1.5. x 1.75 m k
16.5 em).

mats.

Reinforced Soil Embankment

(The Hilfiker Company
3900 Broadway.
Eureka, CA 95501)

6 in x 24 in (15 em x 61 em)
welded wire .... sh: W9.5 to mo
.34 in to .505 in (8.8 mm to
12.8 mm) diameter.

Websol: (Soil Structures

International, Ltd.)
58 Highgate High St.
London N65HX England)

5.3 in (135 mm) wide Paraweb:

made from high tenacity polyester
fibers by Imperial Chemical
Industries.

T-shaped precast co~crete panel

34.4 sq. ft. (3.2 m ) area,
6.3 in (160 mm) thick.

York Method: (Transport and

Road Research Laboratory,
Crowthorne, Berkshire, ngland)

Galvanized mild steel or stainless

steel or glass fiber reinforced
plastic or Faraweb or Terram.

Hexagonal; glass fiber

reinforced cement; 24 in
(600 m) across the flat;
9 in (225 m) deep.

Mda Augmented Soils

(Anda Augmented Soils
Ltd. Oaklands House,
Solarton Road, Farnborough
Hants GU14 7QL ngland)

Fibretain straps (pultruded

fiberglas reinforced plastic strip,
developed by Pilkington Brothers,
1.6, 3.1 or 6.3 in wide, .08, 0.10
or .16 in thick (40, 80, or .160 mm
wide 2, 2.5 or 4 mm thick).

Precast concrete crib units

with 12 in (30 em high) headers
4 ft (1.2 m) apart.

Tensar Geogrid System

(The Tensar Corporation
1210 Citizens Parkway,
Morrow, GA 30260)

Non-metallic polymeric grid mat

made from high density polyethylene
of polypropylene

Non-metallic polymeric grid mat

(wrap around of the soil
reinforcement grid with
shotcrete finish, if desired),
precast concrete units.

Miragrid System
(Mirafi, Inc.
PO Box 240967
Charlotte, NC 28224)

Non-metallic polymeric grid made

of polyester multifilament yarns
coated with latex acrylic.

Precast concrete units or

grid wrap around soil.

Maccaferri Terramesh System

(Maccaferri Gabions, Inc.
43A Governor Lane Blvd.
Williamsport, MD 21795

continuous sheets of galvanized

double twisted woven wire mesh
with PVC coating.

Rock filled gabion baskets

laced to reinforcement.

Precast concrete unit 12 ft

6 in (3.8 m) long, 2 ft (61 cm)
high. cast in place concrete
facing also used.

1Many other facing types as compared to those listed, are possible with any specific system.

a.

Reinforced Earth

Reinforced Earth is a registered trademark of The Reinforced

Earth Company.
Reinforced Earth, which uses metal strip reinforcement, has been
by far the most widely used to date.
The Reinforced Earth
Company is headquartered in the united States in Arlington,
Virginia and markets, designs and supports construction through
regional offices around the united States.
Facing panels for the early Reinforced Earth Walls consisted of
metallic half cylinders of a semielliptic section, fabricated
from galvanized steel sheets. A typical facing element had a
length of 33 ft (10 m), an effective height of 13.1 in (33.3 cm)
and a thickness of 0.12 in (3 mm).
Each element weights 253 lb
(115 kilograms).
Reinforcing strips are connected to the unit by
bolts passing through the strip and the interlocking
(overlapping) edges of the facing units.
Shorter units and
specials are supplied to form corners and bends.
The metal facing unit has now been largely superseded by a
precast concrete unit which is cross shape (cruciform) in front
elevation (figures 2a and 2b). A standard unit weighs lighter
0.8 or 1.1 tons and is 4.9 ft (1.5 m) wide 4.9 ft (1.5 m) high
with total thickness of 7.1 in (18 cm) or 5.5 in (14 cm). All
edges of the unit are rebated to prevent any straight through
joints.
These rebates also facilitate visual alignment of the
units during construction. A further aid to alignment is in the
form of a dowel bar extending from the upper and lower edges of
one arm of the cross element. These dowels are also used as
pivot points for the construction of curved walls. A
compressible material is placed in horizontal joints between the
panels and it allows vertical deformation.
Each panel contains
lifting anchors to facilitate handling and placing.
Each unit is furnished with embedded connector tables, called tie
strips, cast in place during manufacture.
These tie strips which
are usually 2.5 ft (0.75 m) apart horizontally and 2.5 ft (0.75
m)( connection with the steel strip reinforcement.
other special
types of panels which are used to obtain desired overall geometry
are also available.
These include half panels for use at the
base and special panels with varying heights in 8 in (20 cm)
increments for use at the top of the wall to give an upper line
of the facing any desired inclination. Angle elements for
changes in directions are also available.

2~ course of
full-panels

(0 )
I!.! course of
holf-ponels

r1"
IN.

TIE

3 1/2"

...
.

1/2" ,0 A325 BOLT

(b)

FRONT
BACK FACE
OF PANEl:

..

I'

"
#

..

. :t. .

.. ..:

',-

CONNECTION DETAIL
NO SCALE
Figure 2.

Segmented facing panels used for

Reinforced Earth walls,
10

The reinforcement used in the Reinforced Earth walls is

exclusively metal, usually galvanized steel strips. The strips
are generally 0.157 in (4 mm) thick and 2 in (50 mm) in width.
Until about 1975 plain strips were in common use.
Now the
surface of the strips is ribbed in order to improve the apparent
friction.
The ribbed strips are called "High Adherence
Reinforcement." Epoxy coated ribs are also used.
A free draining, nonplastic soil fill is required for the
Reinforced Earth structure in order to achieve the necessary
interaction between the fill and the reinforcing strips.
The
following gradation limits are usually specified:
Particle Size

Percent Passing

6 in (250 mm)
3 in (76 mm)
No. 200 (75 pm)

100
100 - 75
15 max.

The Reinforced Earth system is subject to patents owned by its

inventory, Henry Vidal.
The Reinforced Earth Company of
Arlington, Virginia is the exclusive licensee for this system in
the United States.
.
b.

Hilfiker Retaining Wall

This system, also called Welded Wire Wall (WWW) and Reinforced
Soil Embankment (RSE) is manufactured and marketed by Hilfiker
Retaining Walls of Eureka, California.
Welded Wire Wall (WWW) Description:
This system employees a
welded wire mesh grid within the backfill to serve as
reinforcement for the soil.
The face end of each mesh is bent
upward to provide the facing and then attached to the facing of
the next upper layer (figure 3). The mesh is fabricated to ASTM
185 Standards and is fabricated in 8 ft (2.44 m) wide mats of
varying lengths and can be ordered according to project
requirements.
The mats are placed in alternating layers with
compacted backfill to produce a composite structure. The
thickness of compacted material between the reinforcing mats is
generally 18 in (46 cm).
The mats initially used were 9 gauge
(W1.7) wire laid in 2 in by 6 in (5 cm x 15 cm) mesh oriented
such that the wires spaced at 2 in (5 cm) are perpendicular to
the wall face.
(The W size is the area of the wire in hundreths
of square inches, i.e., the area of WI.7 = 0.017 in 2 ).
Presently, heavier wire mesh W4.5 x W3.5, W7 x W3.5, W9.5 x W4
and W12 x W5 is used for wall heights up to 51 ft (16 m).
vertical spacing between the reinforcing mesh is 18 in (46 cm) or
9 in (23 cm.).
Backing mats are installed behind the bent up
face portion of each reinforcing mat during construction.
The
backing mats which are made of 2 in x 6 in W1.7 x Wl.7 welded
wire mesh are oriented to reduce the openings between the mesh
wire to serve as additional support for the 1/4 in (6.35 mm) wire
screen or geotextile fabric, this serves as protection against

11

ROAQWAY
BACKING

MAT

FILL

FAa:: OF

STANDARD TRAYS

GR.:IVEL FILL

The Gridcote system is interlocking.

A backing mat is attached to the tray
facing for reinforcement. while a steel
screen prevents the loss of backfill
materials.

STANDI\RD

TIlAY
STANDARD

TRAy

Figure 3.

Hilfiker Welded Wire Wall system.

12

fallout of soil particles through the face.

estimated using sacrificial steel.

The durability is

Structure backfill for Welded Wire Wall is generally specified to

be free from stones or lumps exceeding 6 in in greatest
dimension, organic material, or other unsuitable material, as
determined by the Engineer.
In addition, backfill material shall
have a "Plasticity Index" not exceeding 10, as determined by
California Test 204.
This system is patented by Hilfiker Retaining Walls of Eureka,
California. ,Patent 14,117,686, 14,329,089 and 14,505,621.
Reinforced Soil Embankment (RSE):
The RSE system, also developed by Hilfiker Retaining Walls of
Eureka, California, uses either precast panels or a cast in place
concrete face.
For the precast panel system, the reinforcement is in the form of
heavy gauge welded wire mesh in a 6 in by 24 in (15 cm by 61 cm)
grid with the 6 in (15 cm) spacing oriented perpendicular to the
wall face.
Vertical spacing between the reinforcing mats is 24
in (61 cm).
Each reinforcing mat has a steel strip riveted to
its head which fits into slots in the top and bottom of the
prefabricated concrete facing panels to anchor the reinforcement
to both the top and the bottom facing panels (figure 4).
The RSE
has a prefabricated leveling pad placed directly on the
foundation to which the first layer of the wire mesh is attached.
The wire sizes range from W9.5 (0.348 inch, 9 mm, diameter) to
W20 (0.505 in, 13 mm, diameter).
Fill requirements are the same
as for the Welded Wire wall.
The RSE system with precast panels is patented by Hilfiker
Retaining Walls.
Patent #243,697, #243,613, .#4,260,296,
#4,324,508 and #4,343,572.
The soil reinforcements for the cast in place system are similar
to the RSE with panels.
The mat is bent up to provide the
vertical reinforcement for concrete face.
A temporary face and
geotextile fabric are used to hold the backfill until the wall
facing is cast in place.
Fill requirements are the same as for
the welded Wire wall.
c.

VSL RETAINED EARTH System

The RETAINED EARTH system is a composite soil reinforcement

system developed by VSL Corporation of Los Gatos, California.
employs welded wire mesh to reinforce soil placed as fill.

It

This system employs a hexagonal reinforced precast concrete

facing panel 59 1/4 in (1.5 m) high and 68 3/8 in (1.75 m) wide
between the apex points and 6.5 in (16.5 cm) thick (figure 5).
Half panels are used at the bottom and top.
Reinforcement of the
soil is by welded wire mesh (bar mats), which consist of either
13

Rei nfo<cem en1

Precosl C<lncrc1e
FocinQ Panel

Reinforcing
Welded Wire
Meso

EXPLODED VIEW OF RS.E COMPONENTS

Figure 4.

Hilfiker's Reinforced Soil Embankment Wall system.

14

Cop i0 9 ---1---

Reinforcinc
Mesh
-

Panel

\
Granular Backfill

leveling

SCHEMATIC

FRONT VIEW

68 3/8'
34 3/16'

7/32

CD

""-

Vl
~

"-

0
(T)

~.

;1;)
<D

)J

c:;'

Ii

""-~
L'l:

-:-1

Figur-e 5.

---- ----

_ _ ---J

;)

VSL RETAINED EARTH WALL system.

15

TYPICAL
PANEL

wll or w20 plain steel bars placed in a rectangular grid with

longitudinal and transverse bar spacings of 6 in (15 cm) and 24
in (61 cm), respectively.
Each mesh may have 4, 5 or 6
longitudinal bars, depending on the design requirements.
Epoxy
coated meshes are also available.
The overall length of the mats
depends on the geometry of the site, external loading and
physical properties of both the backfill and the earth to be
retained.
The longitudinal elements of the bar mats are looped at one end
which allows attachment to the concrete face panels by means of a
metal rod slipped through the loop and loops embedded in the
panel.
The VSL RETAINED EARTH system is licensed under a Reinforced
Earth Company patent, although it has a patent of its own on the
bottom head connection used to fasten the reinforcing mesh to the
precast facing panel.
The system is marketed in the united
States by the VSL Corporation of Los Gatos, California.
d.

Mechanically Stabilized Embankment (MSE)

This system was developed by the California Department of

Transportation based on a series of tests performed in 1973-75 to
determine pullout resistance of various arrangements of
reinforcements.
The first wall using this system was built near
Dunsmuir, California in 1975.
The facing panels are precast concrete, rectangular in shape,
12.5 ft (3.81 m) long, 2 ft (61 cm) high, 8 in (20 cm) thick to
the VSL system.
Four bar mats can be connected to each panel.
The mats have horizontal spacing of 75 in (1.91 m), center to
center, and vertical spacing of 12 in (30 cm).
Bar mats on this
system are normally not galvanized.
The bar mats are attached to
the facing elements by insertion of the two bar yoke through
precast holes in the facing panels.
The prethreaded bars are
bolted into positions and subsequently field epoxy coated to
eliminate corrosion at the critical threaded sections (figure 6).
e.

Georgia Stabilized Embankment (GASE)

The GASE system was developed by the Georgia Department of

Transportation in its search for a nonproprietary and more
competitive reinforced soil wall other than the ones offered by
vendors of the several proprietary systems.
This system consists of a precast concrete faced wall stabilized
with welded wire mesh.
The wire mesh used is similar to the one
described for the VSL RETAINED EARTH system.
The facing panels
are 6 ft (1.83 m) wide by 4 ft (1.22 m) high as shown in figure
7.
Four reinforcement mats are attached to each panel, at two
levels with horizontal center to center spacing of 2 ft 8 in (81
cm) and a vertical spacing of 24 in (61 cm).
A high quality granular backfill is used for the reinforced
structure.
Fill requirements conform to FHWA specifications.

16

Planter Blockout Top

Center of Each Panel

Half

6'-3"

1/2" Expansion Joint

Shear Key

1/2" Bolts

~j

: , 61/2" to 8"

Figure 6.

Facing panels for mechanically stabilized

embankment wall system.

17

6'-0"
8"

~t
~t

-,
0

'T

~1

~t

, ~t

I~

2'-8"

+
+

.,
I

I II

II

s.

6"

24"

-r

raar
I

-r

..............

l'

..............

~~

Reinf orcement

to Fe cing

o~ Panel Connection

-)
""'--Precast

Facing Pane1
(b) Plan

(a) Elevation

Figure 7.

Georgia stablizied embankment facing panel

and reinforcement attachment.

18

at

The GASE system is also licensed under a Reinforced Earth Company

patent.
f.

"Websol" System

The "Websol" system was developed in 1977 by Soil Structures

International Limited of the united Kingdom.
The system,
although superficially similar to other methods, has a number of
important distinguishing features, the most important of which is
that it uses a flexible synthetic corrosion resistant material.
The "Websol" system comprises a cladding of precast concrete
facing units with stability being achieved through the
interaction of the soil particles and the composite plastic
frictional anchors and an anchor bar (figures 8 and 9).
The lateral pressures in the soil are contained by friction along
the anchor ties and the bulkhead effect of the facing.
The
associated tension in the ties is anchored by friction and by the
tail anchor bars in the zone remote from the face.
The precast
concrete facing units are 6.28 ft (2 m) wide by 5.25 ft (1.6 m)
high and are normally 6.3 in (160 mm) thick.
They are T-shaped
in front elevation with a face area of 34.4 ft2 (3.2 m2 ).
They
are lightly reinforced with mesh.
The frictional anchors comprise a tendon made from high tenacity
polyester fibers concentrated in ten separated bundles encased in
a durable polyethylene sheath.
The tendon material, which is
called "Paraweb," was developed by Imperial Chemical Industries
(ICI) and has been in use in a variety of adverse environments
either in the flat form, or as a high strength rope under the
name of "Parafil," for more than 20 years.
The strength of the
material is in the fibers whereas the sheath forms the size and
shape of the anchor and provides protection to the polyester core
yarns.
The frictional anchors are laid continuously in a zig zag pattern
passing around the anchor bar at the rear of the structure
(figure 8) and are connected to the panels by a simple loop and
toggle bar arrangement.
The continuity of the anchor elements in
both plan directions helps to maintain the coherence of the
anchored mass, particularly when subjected to differential
settlement.
The loop and toggle attachments are normally round
reinforcing steel with their exposed parts coated in resilient
plastic or a fusion bonded powder epoxy coating.
The system is certified for use in the united Kingdom on
Government projects to have a service life of not less than 120
years.
Most on site fill materials may be used although fills with fines
contents of more than 20 percent passing a No. 200 U.S. sieve (75
micron) sieve require particular care.
There is no upper limit
to the size of rock which may be used except that imposed by
normal compaction requirements and by the depth between layers of
frictional anchors.
The resilient and flexible nature of the

19

"~"AW(S

ST~tP

( TO T Al OV[ 'U,...AF ..

';'.b#.)

ru~IPLAST

ALT~NATJVE

~OSITIOH

CL~P

1o.,..{:'"IN.
OVI{LAY

CLOSED CELL
rOLYfTHYLHE
FOU< 'ILLER

T[UAM FAa .. ,c

AHC_ lAo. FOf. THSIONIN(;

(OINTO. HOT LESS TtcAN

0."3 "".)

PLAN VIEW

STEL rlHS
(nt~rF.

HOT LSS THAN

TO IE LESS

O..a ... N<li'L(N(; , "

PReCAST C~C.E~E
,.ACIHC PANEL

(f"\AH

flA..AVE.1 VEaTICAt..

SPACI"';)

1'05 iliON

s-H. c

T<=R..RNJ\. FAe>Rlc..
O.le>i~.

\Ht(..K

t..

ELEVATION
LOOP

HfCAST ...... ELS

COO<PACT[O FILL

---------..:..

COOOCtT u..HAJo<O

TO

'~Tlo..

<"""I~)

P~'

Figure 8.

The Websol system.

20

~TIllPS

Figure 9.

Reinforcement and facing assembly for

the Websol system.

21

anchors permits them to follow the contours of most fills,

however coarse or angular the particle are.
In practice, fills
which have been used to date range from fine desert sands to very
coarse as blasted rock.
Several hundred Websol systems have been constructed to date.
Patents on the system are owned by its developers, Soil
Structures International Ltd., 58 Highgate High Street, London N6
5HX, England.
The system was judged to infringe Vidal's U.K.
patents nos. 1069361 and 1324686 in 1981 and cannot be used in
countries where Vidal's patents are registered until these
patents expire. Generally, this precludes use of the system in
the united States until 1990.
g.

The York Method

This method was developed at the Department of Environment,

united Kingdom in 1973, and so far its use has been limited to
only two or three small government financed projects in that
country.
It uses the Vidal patent in every detail except it
allows the inextensible reinforcing member a limited vertical
movement during construction.
It is really a connection method
rather than a system as such.
The York method has not been used
for some years.
The York method most commonly uses a lightweight glass fiber
reinforced cement facing unit weighing approximately 40 lb (14
kg).
The units take the form of hexagon based pyramid, 9 in (225
mm) deep and 24 in ( 600 mm) across the flat.
One pair of
diametrically opposite flanges on each unit is drilled with large
guide poles.
These guide poles which serve as facing
reinforcement are made up of short lengths of 1 3/8 in (35 mm)
diameter PVC tube with spigot and socket connections (figure 10).
In the finished wall, these pipes are reinforced with mild steel
bars grouted in-situ to render the vertical pipe rigid.
The reinforcement is in the form of strips with drilled holes at
one end which allows them to be threaded onto the vertical pole
at the required vertical spacing. When any settlement occurs in
the fill, the reinforcing strips slide on this vertical pole and
alleviate any settlement induced stresses in the connection.
The facing units interlock with one another both vertically and
horizontally.
There is a compressible gasket between the facing
units to accommodate irregularities in the units and to prevent
leaking out of filIon to the face of the wall.
The reinforcing strips used with the facing panels consist of
either galvanized mild steel, stainless steel or glass fiber
reinforced plastic.
However, the use of continuous lengths of
Terrylene reinforced plastic strips (Paraweb) or unidirectional
fabric reinforcement (Terram) can be accommodated easily.

22

I'T'Hn 0';1. pvc piCl"C' COO("'n'l"9

16...md.... MSbM

JS

too ~ OO""<XI. oC~ct"<t< ,,:p.c

~"'

t..~

co 1:2'SO f.aU

~'C{

'n(o

po<OVI<"Of"\CH'tc

_""'11.-"

. ..-

.3ff:."~i
~.

~
."

.:0

O.:;~-

~ ~,

.. .
.. ..;...

,~
~

,it-

~-.

==rr=:IJ==r.==:::::r,:===;~=7,===rr==~

F=n=
~rid r~inforcemet'\t

Ii

Ii

Ii

1I

11

II

pbn

foeing panel

reinforcement

ISOMETRIC

Figure 10.

VIEW

The York system.

23

h.

The Anda Augmented Soil System

The Anda system is a recent development of the principle of

reinforced soil and utilizes an unique combination of "Fibretain"
reinforcing elements and "Anda crib" type facing.
The facing consists of precast concrete header units 12 in (30
cm) high, spaced 47.2 in (1.2 m) apart.
In between the headers
are precast stretcher and infill panels which can have any
desired exposed face finish.
Each header unit has two cast in
place steel straps with holes to receive a vertical pin which
holds the reinforcing strap (see figure 11).
The "Fibretain"
reinforcing elements are a patentedpultruded Fiberglas
Reinforced plastic developed by Pilkingtons Ltd. of England and
manufactured from continuous 'E' glass voings, combined with
"Oere kane" thermosetting resin.
The straps are 1.6 in to 6.3 in
(40 mm to 160 mm) wide and 5/64 in to 5/32 in (2 to 4 mm) thick
and have ultimate strength of 14,400 to 72,000 lb (6,532 to
32,659 kg).
The Anda System is patented by the Anda Augmented Soils Ltd.,
Oaklands House, Solarton Road, Farnborough, Hants, GU147QL,
England (UK patent no. 1,443,167).
i.

Tensar Geogrid System

This type of soil reinforcement utilizes a nonmetallic tensile

resistant polymeric grid mat, produced and marketed under the
trade name Tensar by Netlon, Ltd. This type of reinforcement was
made available in 1980.
It is marketed in North America by the
Tensar Corporation of Atlanta, Georgia and by Contech
Construction Products, Inc.
The Tensar geogrid is a high strength polymeric grid reinforcing
mat made form high density polyethylene of polypropylene using a
stretching process (figure 12) that imparts a high tensile
strength to the polymer (in excess of 30,000 psi [2,100 kg/cm 2 l).
This results in reinforcements with strengths on the order 100 to
500 Ib/in (87.5 kN/m based on ASTM 0-4595.
The geogrid system of
reinforcement has the advantages of being high strength (close to
mild steel), ductile, durable, resistant to corrosion, ease of
handling (being relatively light), and ease of installation as no
special tools or equipment are required.
Some of the
disadvantages are prone to degradation due to aging, vandalism,
and fire, and creep under high stress.
Facing elements of different types can be used such as by
looping the reinforcement at the face with a wrap around, by
secondary grid reinforcement joined to the main reinforcement, or
by attaching to structural elements, for example, gabions,
concrete panels, etc. (see figure 12). The material is supplied
in rolls 3.3 ft (1 m) wide.
Connections may be made by using a
rod or by stitching with synthetic cord.
The reinforcing grid
can be turned up at the face of the slope and turned into the
embankment below the next reinforcing layer. A shotcrete layer
24

Header IInit
Stretcher IInit _
(nfill panel available
in ~xpo$ed -999 regate
fair (;ace or Coloured finish

Pre-can capping IIn;u

- - - - , . _ _- - - - - - - - - - - - - Road dunner
- - - - - - - - Sub-base

Fric!ion~(

fil( or

P.F.A Of Minestone

Figure 11.

The Anda augmented soil system.

2S

FACING
'----GEOGRIO

OF WRAP-AROUNO FACING

St'"""""9 ......,..,._te<l9__
=
- _ con:fCions. .......
normaI"' .... ..-OrieoWion(.).

S<uiaIr-gridso;g"", 3]_ produood I>y

longitudinal rib

~und4<

CIor"oQ>n

'"

TENSAR

Figure 12.

GRID

Tensar geogrid reinforced soil wall.

26

REJNFORCEMENT

can be applied on the geogrid for a different type of facing.

Several precast concrete facing systems have also been developed
by the Tensar Corporation.
Several hundred walls and slopes have been constructed with this
system.
There are no proprietary restrictions on the system
except that the reinforcing material is patented by Netlon, Ltd.
and is available only through the Tensar Corporation and Contech
in the united States.
j.

Other Polymeric Geogrid Systems

Several other polymeric geogrid reinforcement systems have

recently been developed. These include the Tenax system; the
Miragrid system by Mirafi, Inc., and the Matrex system by
Reinforced Earth Company.
k.

Geotextile Reinforced Systems

In this type of retaining wall, continuous sheets of geotextiles

are laid down alternately with horizontal layers of soil to form
a composite material.
Facing elements are formed by wrapping the
geotextile reinforcement around the exposed soil at the face
(figure 13) and covering the exposed fabric with gunite,asphalt
emulsion or shotcrete, or with soil and vegetation for long term
protection from exposure to ultraviolet light and vandalism.
Alternatively, structural elements can be used on the wall facing
such as precast concrete panels, steel soldier piles and wood
lagging, masonry, gabions, or even cast in place concrete walls
(figure 14).
Connection between the geotextile reinforcing sheet
and the structural wall elements can be provided by a number of
methods including casting the geotextile into the concrete
element, by friction, by nailing and overlapping or other bonding
methods.
Composite construction uses plastic reinforcing strips
along with geotextiles.
The face of a retained wall may be vertical or sloping.
Stress
transfer between the retained soil and the geotextile is by
friction.
A wide variety of geotextiles with a wide range of mechanical
properties and environmental resistances can be used, including
nonwoven, needle punched or heat bonded polyester and
polypropylene and woven polypropylene and polyester. A majority
of the geotextile fabrics used in earth reinforcement are made of
either polyester or polypropylene fibers.
Geotextile reinforced walls can be constructed at most sites even
with poor soil conditions and very steep slopes and in remote
areas, because of limited requirements for heavy construction
equipment.
The materials required are relatively light and
easily transportable. Geotextiles also permit a great
flexibility in the length of fabric and in the vertical spacing
of the reinforcing fabric.
There are no proprietary restrictions associated with geotextile
reinforced walls.
27

I'.

'/

..

\ '
-:-_---/,/"
' .
\ :.
,,"

-'

,/-'.
-

..

..

'-.' . ' /

'.

..
..

. .....

. -/-./ :- -:

- - - - - - -r--'----------p/ ,

., .

'"

7'

'

,-

/
/

'

-7'"
/-,
.' .,'

. . /-11

-/.

~----...,.

>-~--------------~-------/

,
,

' . ,/ . ~,
/'.

, /--='

';L..---_/

.. / ..

..

Figure 13,

Geotextile reinforced wall,

28

. ;'.. :'.
',

"

.
."

-77' -.",

.'

'/

- - -...,..,/

......:

'

'.

J11

(==-----

C:=

5h:J.p//'9 6'~/J/t-C' or SCru~Cqra/

~C/O/

%-rz/;;'o/h~~g$z' Co/7~.r.:."'.tC'

mm<!!,/JC h~/n~

.-

..,
I

r
,/

"j

~~------~-------

~rZ:/N/ C:O~L' ...:z4:pk~~

Caocr.:'o/.A1osa-w-y F~/~
Figure 14.

~rbc<7//%s.:?.I7.ry /Cc//y

Types of geotextile reinforced walls.

29

1.

Gabion Reinforced Fill Wall System

These walls are made of large rectangular wire baskets, wired

together and filled with rock.
Each basket or box gabion
consists of rectangular units fabricated from a double twist,
hexagonal mesh of soft annealed, heavily galvanized wire.
The
gabions are divided into cells by fitting diaphragms which have
the function of reinforcing the structure and making assembly and
erection easier.
They are flexible so they can undergo movements
without failing.
Once assembled and erected, the gabion structure acts as a
monolith rather than a system of separate cages placed side by
side.
The structure remains flexible to absorb localized
settlements, deflections and stiesses.
It has permeable
structure and so no excess hydrostatic pressures act on it.
The gabion structures are designed as a reinforced soil structure
or as a nonreinforced gravity retaining structure.
The
reinforced fill structure is designed similar to any other
mechanically stabilized extensible structure. Reinforcement
consists of double twisted wire mesh similar to that used in the
baskets or of a heavier gage, placed in between the baskets and
extending into the backfill. This reinforcing mesh may be of
uniform or of variable length in the various layers.
In the nonreinforced structure, the gabion walls are designed
similar to any gravity retaining structure. The wire mesh of the
baskets may be taken as an additional safety factor and is not
included in the design consideration. Various types of cross
sections can be considered, taking into account the soil
characteristics, the slope of the backfill, and any superimposed
loading.
Front face of the wall can be vertical, stepped or
sloping (figure 15).
This type of wall was originally developed by the Maccaferri Co.
of Bologna, Italy, who have their USA head office and factory at
Maccaferri Gabions, Inc., Governor Lake Blvd., R.R. 2, Box 43A,
Williamsport, MD 21795.
There is no patent on the system.
3.0

IN-SITU REINFORCED SYSTEMS

a.

Soil Nailing

Soil nailing is a technique for strengthening an in-situ soil

rather than an earth fill as in the case of reinforced soil.
Soil nailing consists of three elements, the in-situ soil, the
reinforcement and a facing (figure 16a). However, facings are
not always used.
The reinforcement generally consists of steel bars, metal tubes,
or other metal elements which resist not only tensile stresses,
but also shear stresses and bending moments.
The inclusions are
installed in the soil at relatively close spacings, one nail for
each 10 to 60 ft2 (1 to 6 m2 ) .
Nails may be prestressed to limit
30

II

< 15

1)

~-

2)

->

'

h
b

1,5

3)

4)

1.00 m
_I
r

Figure 15.

2.00 m'-l--

Maccaferri gabion reinforced fill and

gravity wall systems.

31

Ori<;)inal Grode

Nails

SOIL NAILING IN RETAINING STRUCTURE

Ori<;Jinol Grode

(a )

SOIL NAILING FOR SLOPE STABILIZATION

(b)

T--

-~---

Figure 16.

Uses of soil nailing and micropiles.

32

deflections.
The inclusions are either placed in drilled
boreholes and grouted along their total length or driven into the
ground.
The ground surface of the nailed soil usually is
stabilized by a surface skin which may consist of a thin layer, 4
to 6 in (10 to 15 cm) thick, of shotcrete reinforced with wire
mesh or by intermittent rigid elements which work like large
washers on a bolt or by using prefabricated metal panels (which
may later be covered by shotcrete).
There are no proprietary restrictions on the use of soil nailing.
However, some systems of facing and reinforcements (nails) are
patented (e.g. "TBHA" and "INTRAPAC" mails are patented by
Solrenfor and Intrafor-Cofor companies, respectively).
b.

Micropiles or Reticulated Root Piles (R.R.P.)

A similar method to soil nailing is by micropiles which are also

called reticulated root piles.
Micropiles and root piles have
been used extensively during the past 20 years for underpinning
and reinforcement of foundation soils.
The use of this technique
for slope stabilization and in retaining systems is a relatively
new application (figure 16b). Most of these projects were
executed in the least ten years or so.
The root piles are cast in place reinforced concrete piles with
diameters ranging from 3 to 12 in (7.5 to 30 cm).
In the smaller
diameter range, these piles are provided with a central
reinforcing rod or steel pipe, while those with larger diameters
may be provided with a reinforcing bar cage bound with spiral
reinforcement.
"Micropiles" consist of long and very strong (about 50 to 100
kips strength in tension) inclusions which are grouted in a
predrilled borehole at a wide spacing of about one bar per 30 to
60 ftl (3 to 6 m2 ) .
"Hurpinoise" is a term used to refer to
nailing with shorter, less resistant bars (10 to 35 kips in
tension) that are driven into the soil at a close spacing
(approximately one bar per 5 ft2 (0.46 m2 ).
The piles are arranged according to a three dimensional pattern
in order to form some sort of network in which the soil is
encompassed.
Therefore, whereas in the "soil nailing" the nails
behave as independent resisting elements, in a Reticulated Root
Pile structure the mutual action among the piles is essential and
it must be encouraged.
Therefore, the design of root piles has a
completely different approach from soil nailing.
The most
important design element in a Reticulated Root Pile structure is
the appropriate density of the piles.
Generally, there is no facing for the R.R.P., because the
majority of these structures are intended to be buried in the
subsoil, to behave like buried retaining wall, not to be exposed,
or to be only partially exposed.
This is the case of landslide
prevention, protection of buildings in presence of excavations of
tunnels for subways, etc.
33

One difference between micropiles or root piles and soil nailing

is that the reinforcing bar(s) in the micropiles and root piles
are grouted under pressure. The grout significantly increases
the adherence of the bar to the in-situ silo, and it enables the
bar to be installed at any orientation. Another difference
between this system and soil nailing is that the behavior of
micropiles or root piles is significantly influenced by a soil
pile interaction due to three dimensional arrangement of the pile
group.
The "Root Piles," (generally used for underpinning) and the
"Reticulated Root Piles," were invested in the early fifties by
F. Lizzi and patented by the Italian firm, Fondedile of Naples.
The same firm introduced and installed the system allover the
world. Although the original patents expired and therefore there
are no more proprietary restrictions (except for the trade marks
in the names of "Root Pile" and "Reticulated Root Piles") it must
be stresses that the design of a Reticulated Root Pile structure
requires specific experience.
c.

Composite In-situ Systems

Composite systems utilize combinations of reinforced soil and

other types of retaining systems (gravity, cantilever or anchor).
Similar combinations with in-situ reinforced systems can be
utilized, depending on the existing and proposed situations.
4.0

MULTIANCHORED SYSTEMS
a.

American Geo-Tech System

This systems of retaining wall has been developed by American

Geo-Tech, Inc. of Baltimore, Maryland.
The retaining wall
consists of precast reinforced concrete face panels, precast
reinforced concrete deadman and steel tendons connecting them
(figure 17). The system relies on passive pressures on the
deadman panels for stability.
Internal stability does not rely
on friction between the reinforcement and the soil, but requires
only adequacy of the face panels and the deadman to resist
lateral pressures and the tendons to resist the generated loads.
Standard face panels are 8 ft (2.44 m) wide and 5 ft 4 in (1.63
m) high.
Up to a lateral pressure of 1500 psf (71.9 kPa), each
panel has 8 tendons and for pressures of 1500 to 3000 psf (71.9
to 144 kPa) each panel requires 16 tendons. A transition row of
panels having 12 tendons is also provided. The tendons are epoxy
coated (8 mil thickness), 0.75 in (19 mm) diameter ASTM A36
steel, plain rods with ultimate yield capacity of 26.5 kips (118
kN) and ends of the tendons are threaded for connections to the
face panels and the deadman.
American Geo-Tech, Inc., P.O. Box 9696, Baltimore, Maryland,
21237 has applied for a patent for this type of retaining wall.
To date, this system has had very limited use.

34

Alignment Bars

Eight Inch Lifts

Of Soil Fill

Focin9 Panels

Leveling Footing -"

STANDARD 8" UFT LAYOUT

TENDON LOCATION

Figure 17.

American Geo-Tech system.

35

b.

Tension Retaining Earth System (TRES)

This is another method of retaining wall, developed by Columbia

Engineering Company of Silver Springs, Maryland. TRES uses
precast segmentally constructed face panels anchored by tendons
to earth anchors (figure 18). Wall facing is made of hexagonal
interlocking precast concrete panels in any desired size, from
man liftable to large rapid erection size. Tendons consist of
standard mild steel reinforcing bars, galvanized or epoxy coated
for corrosion protection. Most walls require only rocks or
concrete rubble for deadman anchors.
Precast concrete anchors
are also available.
Each face panel is anchored at four points
using tendons having attached anchor plates. The tendons have
loops formed (for deadman) and welded plates at each end for
connection to the face panel.
Patent on the TRES system is held by its developer, Walls
Systems, Inc., P.O. Box 2543, Columbia, Maryland 21045.
Only a
few structures have been constructed with this system.
c.

Anchored Earth

The Anchored Earth retention system was developed and patented by

the Transport and Road Research Laboratory (TRRL) of Crowthorne,
England.
This type of retaining wall is still in an
experimental stage and has not been used on any project in the
united States.
This system has precast concrete panels, rectangular in shape,
with adjacent panels having overlapping edges.
The panels are
typically 47 in (1.2 m) wide, 6 in (15 cm) thick and of varying
heights.
The reinforcement is by mild steel bars of 0.6 to 0.8
in (16 to 20 mm) diameter having a screw threaded portion at one
end.
The other end of the anchor is formed by bending the bar in
the form of a z or a triangular end in which the loop around is
welded to the main bar (figure 19).
For anchorage of the
reinforcement to the panel, the rods extend through the facing
panels in slots and are secured by nuts.
The reinforcing rod
extends through the overlapping portion of the two adjoining
panels. The protruding rods and the nuts can be housed in a
recess cast in the facing panels and then capped to provide a
flush appearance.
Anchored Earth is designed on the basis that passive resistance
is developed only at the deformed ends of the reinforcing
members.
It is likely to be more efficient in cohesive soils
than the other systems which rely predominately on friction.
The
pullout resistance is not sensitive to the surface
characteristics of the anchors because of the relatively large
area available for passive resistance at the deformed end.
The Anchored Earth system is patented in the united States and
elsewhere by the Transport and Road Research Laboratory,
Crowthorne, England.

36

CONCRETE CAP

,
\

ANCHOR
TENDON

FACE PANEL

Figure 18.

Tension Retaining Earth System.

37

FACE PANEL

TRRL .~ ur:ll rfS"""

FACING

UNIT

_ - - - ANCHOR FACING CONNECTION

eoq-T.uiblc

ttllioc

_i__1 . - _ - ' Figure 19.

TRRL anchored earth system.

38

"

'-

"./
.........~/./

..,~

precast concrete
facing units
(1.5 m X 0.8 m)

ties wi~h anchors

I
I

I
I

selected stone fil!

I

I
I
I

I
I

ordinary fill

I
I

I
I

___ LI ___________ _

Figure 20.

Ladder wall-sloped face.

Steel Rods Encased

in Concrete - - - - - - - - .

l----~----

Face Wall

Deadman ( e~ms of
Reinforced Concrete)
Typ.

Figure 21.

Ladder wall

39

with vertical face.

d.

Ladder Wall

The "Ladder Wall" (Mur Echelle) system invented by Coyne in

France in 1926 consists of a multitied back system associated
with thin facing made either of concrete panels or of a
continuous wall (figure 20).
The ties withstand tensile forces
which are constant along the sides.
The soil reinforcement
interaction is essentially realized by the passive lateral thrust
on the anchors.
5.0

ALTERNATIVE SYSTEMS
a.

Gravity walls

Traditional Gravity walls:

These walls are well known and need
no special description in this manual.
Crib Walls and Peller walls:
Crib type retaining walls are built
of interlocking prefabricated reinforced concrete units, which
are essentially stretchers at front and back and headers
connecting them, similar to "crib lock" system described below.
The spaces between these units are filled with a free draining
soil.
Crib walls permit some economy of concrete and can be
erected quickly.
They permit excellent drainage of the backfill.
With a sloping front and back faces, lateral pressures acting on
it are less than those for a vertical retaining wall.
Crib Lock:
This system was invented in New Zealand and has been used in
California since 1977.
The retaining wall is comprised of
precast reinforced concrete members which interlock to form
skeletal braces.
The various elements for front and back
(stretchers) and the headers connecting them are put together
like Lincoln logs.
The boxes formed thus are filled with a free
draining material and a gravity retaining wall is created (figure
22).
The system is particularly useful in remote and mountain
areas where heavy equipment cannot easily be used.
A novel feature of
this wall is the small dimensions of the elements.
Bin Walls:
Bin Wall is a gravity retaining wall in which
tenuously connected steel bins are filled with earth.
The earth
mass acts as the gravity wall with the steel members serving to
hold the earth mass intact.
Bin walls thus utilize the cellular
or crib wall concept but are not merely steel crib walls.
Bin
walls are constructed of lightweight, deep corrugated steel sides
with bolted corners.
Deflection is thus available in the sides
of the bins, permitting some stress relief from soil pressures
while the corners are positive connections able to distribute
shear forces.
Conventional crib walls with articulated corners
and rigid sides do not have these characteristics.

40

12" MOOLU:

FAi.St H::AOCR"\

]_

f"RONT STRlOR-.

FRONT STRETQ;R

oa.a..

(B>TH WAll..

I'I~

-J~_-i4n~g~~o~

__i.

~---JJ:::o.

....... /~

W!o1}f

OOTE' FALSE IAOCRS Ml.

NOT USED fj n rm:RICfl
sccn:JNS IF MULllPl
OCPTH WALLS.

MlU1PLE DEPTH WAlL (D'PICAL)

NO SCALE

:1
I.

.(
I

-0

~
~

Cl.DS8

..

3-~

-0

-9

[ ]-=S

11
1\
't'

- J'

]--!i

5QgR

5'-0

Figure 22.

Of!

. I''t

I.(};l

ST~

Crib lock retaining wall.

41

Doublewal:
"Doublewal" is a gravity retaining wall system which consists of
large precast, interlocking, reinforced concrete modules that are
placed like building blocks (without the use of fasteners) and
then backfilled with select material to form a gravity retaining
wall (figure 23).
Each module consists of two face panels held
rigid and apart by connecting beams. wall segments are available
in different sizes and like other precast systems, can be reused.
The wall can be built quickly.
The system is patented by the
Doublewal Corporation, 59 East Main street, Plainville,
Connecticut 06062.
Evergreen: The Evergreen wall is composed of precast elements
stacked on top of each other.
This precast concrete wall system has open spaces on its face
into which are planted shrubs, vines, etc. (figure 24).
It has
better acoustic characteristics, antigraffiti surface, and good
aesthetic appearance. This system was developed in Switzerland
and is patented.
The licensee in the united States was Evergreen
Systems, Inc., P.O. Box 345, Kings Park, New York 11754.
stresswall: The Stress wall system was developed by stresswall
International Inc., P.O. Box 10838, Denver, Colorado in
collaboration with International Engineering Company, Inc.
Patent on the system is pending.
This system has L shaped precast concrete elements called
counterfort or tieback units stretching into the backfill form
the wall face and hollow core wall panels completing the wall
space between the counterforts (figure 26).
The elements are
assembled in vertical tiers.
It is claimed by the patentee that
the key mechanism for stability is soil arching between the
counterforts. A counterfort unit is held from moving down or out
from the wall by the soil arching from neighboring counterforts.
Thus, the counterforts have far less depth into the backfill than
if it were purely a gravity system.
b.

Cantilever Walls

In certain favorable soil and rock conditions, cantilever walls

can be utilized for moderate heights of retained earth.
The
cantilever wall may consist of reinforced concrete with an
extended footing, steel sheet piles or sheet H piles with treated
wood lagging or precast concrete planks, or a row of large
cylindrical pipe piles bearing on rock and held with vertical
anchors installed eccentrically to resist the moments from earth
pressures.
In certain situations, reinforced concrete slurry
walls or precast concrete walls installed by the slurry method
can also be utilized to serve as cantilever retaining walls.

42

8EAM

Tor KEY

LIfTING HOLE

TYPICAL TOP MODUlE-A

TYPICAL TRANSITION MODULE -B

TYPICAL STANDARD MODULE - C

LEYEUNG PAD
TYPICAL BOTTOM MODULE - 0

Figure 23.

"Doublewal" retaining wall.

43

'\

'(

\ -_ _ _ _ _ _ _..!..L-'-!.-_ _ _ _ _ _ _~v?

Figure 24.

Typical element of Evergreen retaining wall.

Figure 25.

..

"Evergreen" retaining wall - front'view.

44

PARAPET

....

..

::!
><

HOl.l.O'W'
f"A.HELS

CO<I(
M<S1"RCSSt:O

..,'"
TI(R: . .(

""$(:
J

...::!

...

.X

TI(IACIC

0(

"'
TYPICALSE.CTION

TIEl!: 3
8.:SE
J

...::!><
u

""
C

....

Tlf~

..... S!:
.4 .... 0" TICI.lCX Jt(O'D.

,~ ~"~~Ko.~~1
..0'" -0

..J

...::!
'"

""

TKI.A.CX"

TyrtU.<..

<,

TIE_

c:
~

u. s:

TYPICAL WALL SECTION

----~
:v..~57-Ni\1..1..

f>SS,{.

'",I-f{:~'Oft "iY{

Figure 26.

stesswall details_

45

-,..

c.

Grouted Anchor walls

In this case a vertical or sloping retaining wall is supported by

anchors situated deep into or behind the backfill, beyond the
likely failure plane (figure 27). These systems are very useful
when it is not feasible to excavate behind a proposed wall.
In
many cases the anchored wall system is the retaining wall.
The
wall consists of sheetpiles or a slurry wall installed before the
start of excavation, or soldier piles installed before the start
of excavation and lagging installed as the excavation proceeds.
The excavation is done in stages 5 to 10 ft (1.8 to 3 m) depths
(the stages can be larger depending on the vertical spacing of
anchors) to a level about 1 to 2 ft (30 to 61 cm) below the
design location of the anchors.
Before proceeding further with
the excavation, a row of anchors is installed into the earth
behind the wall.
Thus, the wall is held in its original plane.
Excavation is then extended to the next anchor level and the
anchors installed at that level and so on until the final grade
is reached.
A cast in place or precast concrete retaining wall
is constructed on the face of the retained earth.
The anchors
can be designed for permanent support of the retaining wall.
The application and design of anchors is described in a Federal
Highway Administration report entitled "Permanent Ground Anchors,
Report No. FHWA-DP-68-1", dated March, 1984.
d.

Deadman Anchored Walls

In this design, deadman anchors (e.g. solid masses such as

concrete, rock, or plate elements such as sheet piles, etc.) are
placed in the embankment behind the wall (figure 28).
Deadmen
are connected to the wall by reinforcing steel bars which are
protected against corrosion by galvanizing, epoxy coating or by
wrapping with tar impregnated cloth or paper. The wall may
consist of sheet piles of soldier pile and wood lagging or
precast concrete.
e.

Composite System

In addition to the different proprietary systems described above,

it is possible to combine two systems for certain prevailing
conditions.
For example, a composite system may consist of a
gravity system of retaining wall at the base with a reinforced
soil system at the upper part.
The lower gravity system could be
anchored into the rock base (if one exists) or a permanent
grouted anchor system can be utilized in the lower gravity
system.
The upper system may consist of reinforced soil using
one of the proprietary systems described above to reduce the
overall cost of the system.
It is necessary to evaluate the stability and performance of such
composite systems carefully because of possible differential
lateral displacement at the junction between the two systems.

46

Permanent
Wall--....

Temporary
Sheeting or
Soldier Piles
and Logging

Figure 27.

Grouted anchor wall.

47

-.
;.a.

-"

"-

Permanent
Wall

:~

I
I

Deadman

- I
.:. I
.'.

\1

....

'"

.. ":.:

I
I

I
I

Temporary
Sheeting or
Soldier Piles
and Logging

I
I

Figure 28.

.~ ~ .

- ..:-.-

:'

... I
-:.:. I
7,

';)-'J

Tierod

Deadman anchored wall.

48

::

A combination of any two systems can be considered provided there

compatibility can be assured and the overall system can be

designed to perform satisfactorily.

49

CHAPTER 2
PULLOUT RESISTANCE EVALUATION
1.0

INTRODUCTION

This section provides supporting information for the evaluation

of pullout resistance using empirical relations and pullout test
results.
In addition, information on using the pullout data to
evaluate the deformation response is reviewed. Also, evaluation
of long term soil and reinforcement creep considerations are
discussed.
AS indicated in chapter 2 of volume I, the pullout resistance per
unit width p
for any reinforcement system can be estimated using
the following general equation:
pr

F* .

(X

a' v

. Le

. C

. C = the total surface area per unit width of the

reinforcement in the resistivity zone behind the
failure surface
Le
the embedment or adherence length in the
resisting zone behind the failure surface
C =
the reinforcement effective unit perimeter; e.g.,
C=2 for strips, girds, mesh and sheets; C=n for
nails, in which b = width of strip, grid, or
sheet, d = diameter of nail
the pullout resistance (or friction bearing
F* =
interaction) factor
(X =
a scale effect correction factor
a' v
the effective vertical stress at the soil
reinforcement interfaces.

where Le

This equation varies from the conventional equation in that it

has been normalized with respect to width to allow for a more
general evaluation of any reinforcement geometry. The actual
resistance available for any specific type of reinforcement can
then be obtained by multiplying P by the coverage ratio R or the
gross width of the reinforcement b divided by the center to
center horizontal spacing between the reinforcements s .
In
addition, to account for nonlinearity of stress distri~ution
along the length of the reinforcement, especially for extensible
reinforcements, the scale effect correction factor (X has been
added to the equation.
The pullout resistance factor F* can most accurately be obtained
from pullout tests performed in the specific backfill to be used
on the project. Alternatively F* can be derived from empirical
or theoretical relationships developed for each soil
reinforcement interaction mechanism.
The following subsections provide the necessary information to
determine F* and (X for the different types of reinforcement.
50

2.0

ESTIMATE OF FRICTIONAL PULLOUT RESISTANCE

As indicated in equation 1, the pullout resistance of the

reinforcement will be dependent on: the pullout resistance factor
F*i the scale effect correction factor a to account for
nonuniform load transfer along the length of the reinforcement;
the effective unit perimeter C to account for the effective cross
sectional pullout surface for the specific type of reinforcementi
and, the design conditions of length and overburden pressure.
This section provides guidance for evaluating the coefficients
F*, C, and a for the various reinforcement types.
a.

Inextensible Linear Strip Reinforcements

For inextensible linear strip reinforcement:

p*, C = 2, a = 1

F*

p*

is an apparent friction coefficient obtained from pullout

tests.
The apparent friction coefficient p* is mainly dependent upon the
surface characteristics of the reinforcements (e.g. rib effect)
soil properties (specifically, internal friction angle and
dilatancy) and the overburden confinement stress which restrains
the tendency of the soil to dilate.
Figure 29 shows typical
results of pullout tests on smooth and ribbed metallic strips
used in Reinforced Earth structures.(63}
Figure 30 shows the
recommended design p values.
For ribbed linear strips the maximum apparent friction
coefficient at the top of the wall can be estimated from the
following relationships:
p

* = 1.2

where:

+ log C

C = 060/DID is the uniformity coefficient of the

r~inforced backfill (060 and 010 are the grain
diameters at which 60 percent and 10 percent of the
soil sample weight is finer).
If C is not known,
assume p * = 1.5 for well graded fill.
If a very
uniform lill is to be used, assume p * = 1.
The
design p value decreases linearly toOtan at a
depth of 20 feet (where ~ is the internal friction
angle of the soil).

For smooth strips

p*

= tan

where:

0.4

p is the soil strip friction angle that can be

measured by a direct shear test.

51

~--------~---------.r---------~--------~

Gravel
'Y = 134 pcf
= 46 0 Internal Friction

.e = 27. 50

S0 i 1- s t rip Fric t ion

o ~--------~~----------~----------~--------~
10
o
15
5
Heigrlt of Fill Above Strip, H-ft

Figure 29.

Influence of reinforcement type and overburden stress

on apparent friction coefficient.

52

tone ton,

1//kY/k/'/ ",Y/""'),./~//~7/.k/IA'/1~ //;:;:'7//

I
1

I
I

(/)'

-a::
0..

t
I

l-

U)

I
I

l-

I
I

0
0
:E

I
I

Cl)1

20'

-----t--~

l(/)

a
w
Z

a:l
a:l

0::

Figure 30.

Reinforced earth design p* values for smooth

and ribbed strips.
53

"

Rod Reinforcements (e.g., rods, bars, wires)

F* = K . p*; p* = tan
C

= 0.4;

= 1

1 + Ko

C1 n '

where:

<J

Ko = 1 - sin cp'

<J v '

(3)

is the average effective normal stress acting on

'

t~e circumference of the reinforcement

Ko

is at the rest earth pressure coefficient

cp' is the internal friction angle of the backfill

material
K

value of 0.44 is commonly assumed yielding K

O~ 72.

b.

Extensible Geotextile Sheet Reinforcements

F* = tan

pea k '

C = 2

= 1

= should be determined from pullout tests on

instrumented geotextile specimens as explained later
in this section.

The peak soil fabric friction angle p

k
can be obtained from
direct shear tests carried out in accbr~ance withe the proposed
ASTM test method (modified ASTM 3080 using a large 12 in x 12 in
(300 mm x 300 mm) direct shear box with the geotextile clamped at
the shear interface (FHWA Geotextile Engineering Manual, Appendix
B). (25)
The interface shear stress shear displacement curve
obtained from the direct shear test will exhibit strain softening
which as shown in figure 31 will result in a residual direct
shear interface friction angle p res
The scale effect correction factor a indicates the nonlinearity
of the P - L relationship which is primarily dependent upon the
extensibility of the reinforcement.
Due to the extensibility of
the application of a pUllout force on the reinforcement results
in a nonuniform shear displacement distribution (figure 32). The
interface shear stress is therefore not uniformly mobilized along
the total length of the reinforcement.
The average shear stress
~a
mobilized at the peak pullout load depends upon the
reinforcement elongation during pullout which in turn depends
upon the extenSibility of the reinforcement materials and the
reinforcement length.
The scale effect correlation factor a can
be defined as:
~

a v
p

tan

54

(4)

t an~ peak

---

I ' = SHEAR STRESS

6= NORMAL STRESS

Ep

Ep

Figure 31.

is typically 2 to 5 percent

stress shear displacement curve from

direct shear test.
55

0.8

0.6

_0.4

-=
IZ
W
~
W

:50.2
a..
(j)

Ci

o
EMBEDED

Figure 32.

4
LENGTH

---

OF REINFORCEMENT L

10

12

(in)

variation of displacements along a woven

polyester strip during a pullout test.

56

where:

t
and t
are, respectively, the average and
uitimate ~nterface lateral shear stresses mobilized
along the reinforcement.

p
and P
k are, respectively, the average and
peak int~riace friction angle mobilized along the
reinforcement.

The correction factor depends therefore primarily upon the

strain softening of the compacted granular backfill material, the
extensibility and the length of the reinforcement.
Determination of :

Experimental Approach -

Pullout Tests

1.

conduct displacement rate controlled pullout tests using

proposed ASTM procedures (see Geotextile Engineering Manual,
Appendix B) on specimens with different embedment lengths
under a specified normal stress.(25)

2.

Establish the normalized pullout load (P /R.o ) versus

length (L) .curve as illustrated in figur~ 33a:
(R is the
coverage ratio used to equate the force per unit width of
discrete reinforcement to the force per unit width across
the entire structure.
R is equal to the gross width of the
reinforcement b divided by the center to center horizontal
spacing Sh between strips, sheets or grids).

3.

The P pullout load is the force applied to the confined

portion of the geotextile sample.

4.

The initial tangent at the origin of the (P /(R.o ) versus

L) curve corresponds to the peak interface friction angle
P
k.
The P
k angle derived from pullout tests using
t~ls procedur~ecan be different from that obtained from the
direct shear tests due to differences in the testing
procedure, sample preparation and restricted relative
soil-to-geotextile movement.
The pullout test provides a
more appropriate
P
k
value
for
the estimate of the pullout
.
pea
Capaclty.

5.

Establish the -L curve as illustrated in figure 33b.

value for a specific L value is defined by the ratio of the

secant tangent tan Pm for this L to the initial tangent than
Pp e a k

Experimental Procedure to Determine with Instrumented

Geotextile Reinforcements
1.

Conduct displacement rate controlled pullout tests on

instrumented reinforcements under a specified normal stress.

2.

Establish the normalized pullout load mobilized length

(P /R.o ) - L ] curve.
A section of the reinforcement is
cohsidered tornbe mobilized when the wire extensometer
indicates movement at its end.
The corresponding pullout
force is plotted versus the length of the section.
57

NOTE:

Pr PULLOUT LOAD APPLIED TO THE CON FINED PORTION

OF THE SAMPLE.

/
/

~taneres

/
/
/

/(tane

~./

It
/
/

./'

./'

//

tan\' peak
0)

0{

d:..=

tan (? m
tan

e peak

eres-RESIDUAL INTERFACE
FRICTION ANGLE

0<::= tanE'res ItanE'peak

b)

Figure 33.

Experimental procedure to determine n

for geotextile sheets.
58

Analytical/Numerical Procedure to Determine a

Modeling the load transfer mechanism in a pullout test on
extensible inclusions, requires appropriate constitutive
equations for the soils and the inclusions as well as a rational
interaction for the low, relating the shear stress mobilized at
any point of the interface to the soil reinforcement shear
displacement.
This interaction low can be obtained from direct
shear tests with soil geotextile interface following the proposed
ASTM testing method (modified ASTM 3080).(64)
The load transfer
model should allow for an estimate of the shear stress
distribution along the reinforcement and of the front edge
displacement caused by the applied pullout force.
The "t-z"
method which is commonly used in the design of friction piles can
be combined with the nonlinear displacement shear stress
relationship obtained from direct shear tests to derive a
rational load transfer model.
The load transfer model should be "calibrated" by numerical
simulations of pullout tests conducted on the geotextile
reinforcement specimens and will then provide an appropriate
engineering tool to predict the pullout capacity as a function of
the reinforcement length.
The use of such a model allows to
parametrically evaluate the effect of material properties and
interface shear behavior (i.e., strains softening) on the pullout
capacity.
Figure 34 illustrates the use of such a model for the
evaluation of the average limit shear stress at the peak pullout
load for different reinforcement lengths and stiffnesses.
3.0

ESTIMATE OF PASSIVE PULLOUT RESISTANCE

In anchored systems that rely on multideadman elements or in

reinforced soil systems that use resisting elements oriented
transverse to the pullout force direction (e.g., bar mats, wire
meshes or geogrids) the pullout resistance is derived entirely or
partially from the passive soil pressure on the transverse
elements (or deadman).
The maximum passive pullout resistance on deadman or grid
transverse elements can be estimated from bearing capacity
formula for deeply embedded strip footings (i.e., embedment depth
- z significantly greater than the thickness - t of the bearing
number).
A generic equation for the pullout passive capacity per
unit width of reinforcement can be derived:
Pr
where:

[c.t.F c

+ y' Z.t.F q

+ y'

Z t

Fy 1

(5)

P = the pullout resistance per unit width of

reinforcement.
c' is the effective soil cohesion, y'
effective unit weight of the soil

59

is the

0.30
k-""-':--.L-_ _=---------------~".--

CI)

a.

E Co :3x 107psi

1--0.25
~..J

-a.

-I,

WlC

- - - - -____--:::--:;:: E.

~~0.20

E ..

O::w
Wo:
>1<l(/)

Uy

1.5 x 104psi
1.5 x I03psi

tan c./Jr .. 0.18 psi

~o::O.15

Ww

REINFORCE MENT:
p:: 4 em (2 .. 2em)
S= 0.4 em 2

o..:r:

(/)

SOIL: FOUNTAINBLEAU SAND

'" :: 42 0
cJI~ = 32 0

G/ayd:: 6 ( I/mm)
a y J:2kPO

0.70

REINFORCEMENT

Figure 34.

8
LENGTH

10
1

12

14

L (ft)

Numerical procedure simulating pUllout tests to

establish a-L curve for specific soil
type and reinforcement properties.

60

fp is the fraction of the transverse member on

which bearing can be fully developed
N number of transverse bearing members
F , F

and F

(~r e~bedment)

are, respectively cohesion, surcharge

and friction bearing capacity

factors.
As soil cohesion c' of the backfill material is usually very
small the cohesion bearing term can be neglected.
In addition,
for the deeply embedded transverse members (i.e., t z) the
frictional bearing term can also be neglected as it too is
relatively small at that depth.
Hence, eq 5 becomes:
Pr

= Fq .Y'z.fb.t.N

(6 )

substituting: N = L /S i where S is the longitudinal

spacing between transverse elements ana L equals the length of
the reinforcement.

c = 2; a v

y'z; and,
fb

0:13

where
resistance.

--=S--x

is a structural geometry factor for pullout

0:13

Eq. 5 becomes:
p

Fq

(7 )

0:13 a v ' L e C

Note that eq. 7 extends the pullout equation to account for the
effect of the structural geometry of the reinforcement on the
passive pullout resistance per unit width of reinforce.
Introducing the scale effect correction factor 0: into eq. 8
yields the generic equation for passive pullout resistance:
p

F*.o:.a v '.L e

wi th F*

.c

(8 )

Fq

The maximum passive pullout resistance is obtained as the grid

and the soil contained within the grid act as a rough sheet of
thickness t being pulled through the soil.
For this case the
optimal structural geometry is:
S opt

2 tan 4>

61

(9 )

Grid Bearin9 Members

--+-+-~~--t+-~----+--l~

- - - -- p

To be Fully Rough:
25 max

ttttttt

Figure 35.

It

Definition of bearing stresses on

transverse elements.

62

CTV

tan ct> :: r

(j

b'

There would be no increase in the passive pullout resistance for

sit values smaller than those calculated from eq. 9.
Greater Sit
values will result in a system that does not fully mobilize the
maximum passive resistance within the area of reinforcement.
a.

Determination of

(XI'!

For anchored systems

fb = 1 ;

S x = L',

(XI'!

2' L

For bar mats and wire mesh systems, the transverse and
longitudinal members are in different horizontal planes; hence:

2 " ' Sx
For geogrid systems, all members are in the same horizontal
place; hence:
t

Sx
b.

Determination of F
----------------------q

The bearing capacity factor F , defined as the ratio of the

effective bearing resistance ~Q' developed on the transverse
elements to the effective vertlcal stress 0 "
(i.e., F =
a p '/0 " see figure 35) is primarily dependent upon soil strength
characteristics and dilatancy properties.
The effect of
reinforcement roughness and initial stress state in the soil on
the bearing capacity of these deeply embedded members can be
neglected.
F

values obtained as a function of soil friction angle using

36.
Rowe and
Davis's curves are derived from finite element analysis of
horizontally loaded vertical surfaces considering respectively
(1) soil dilation, and (2) constant volume (no dilatancy).(65
The upper and lower bounds are derived from Prandtl's bearing
capacity solutions for the two failure mechanisms, frictional and
bearing, illustrated in figure 35.(34)
s~veral analysis procedures are shown in figure

The upper bound is given by:

Fq

= tan 2

(~
4

+1.2

exp [rr.tanq,]

63

( 10 )

The lower bound is given by:

F

(.!!
4

+.!2

( 11 )

exp [ 2 ' tan.p]

The results of several pullout tests (and large direct shear

tests reported in figure 36 show that Rowe and Davis's curves
define reasonably well the range of experimental values.(25)
F design values have been specified for different reinforcement
sJstems and are summarized in the NCHRP report No. 290 as
follows:(2)
Bar mats:
Pullout test results on VSL bar mats and bearing
capacity factor design values (F ) are shown in figure 37.(66)
In spite of the spread and varia~ility in the test results
obtained in different types of soils, the F values obtained for
the range of backfill materials commonly us~d are within the
predicted range of F values indicated in figure 36.
Most of the
experimental values approach the upper bound defined by eq. 10 or
by Rowe and Davis's solution.
welded Wire Meshes:
For commonly used backfill materials
(i.e., pea ~ravel, silty sand, clean sand) pullout tests
yielded:(67

Fq = 36.5 to 38
These values approach the upper bound solution defined by eg. 10.
The pullout test results are reported in figure 38 as passive
pullout resistance component versus the overburden pressure.
Linear regression of the test results yielded a "cohesion
intercept" term F for the noncohesive soils (i.e., for clean
sand F = 633 lb/ft, for pea gravel (F = 712 lb/ft) that has
been incorporated in empirical correlations developed for design
purpose.(2, 67)
Because of the uncertainties inherent to these
correlations a more conservative approach is retained in this
manual neglecting the Fe values in noncohesive soils.
Geogrids:
Recommended F design values for the estimate of
the passive pullout resistanc~ of geogrids are indicated by the
dashed line in figure 36.(34)
Anchored Systems
(Anchored Earth, Geo-Tech Systems, Tension
Retaining Earth System:
F values for both triangular and h Z
type" anchors can be calcu1ated from the equation developed by
Mu r ray: ( 6 8 )
For individual triangular anchor (Q = 70)
F q = tan 2
where:

12

exp [2(n - Q)
cos Q

tan .p']

is the angle defining the bearing wedge

64

(12)
(~

70)

1000

Anchor test date

1. Oversen b St r cx;;a I,
2. Oversen A Stroman
Neely et al.
Das to Seely

100

-7
/

./

(J.'

/y

.
@Jl
/

cr'
v

6c:

, "
6'"

)'

/
1/(4)

.j.~~3

~ ~,~
V
Sa

:,0/
V

V ~
V

10
./

......-

---

1/

~~

./

./

./'

Grid test data

(17.)

3. Huedel 1. Kwa 5 ni ewsk.i(l7)

4a,c. Chang t. a1.
(9)
Sa,b. Peterson
(39)
(20)
Je
....
ell
6.

1
20

,/

30

I
40

Soil Friction Angle

50

W
( 34 )

(1)

(2)
(3)
(4)

Prandtl (Adapted by Jewell et al.)

Rowe and Davis ( 651 - Dilation
Rowe and Davis(65) - No Dilation
Jewell et al. (25)
- Punching Failure

Figure 36.
Theoretical relationship and experimental
results bearing stress versus soil friction angle.

65

70r---~~~~,---,-------~------.--------.-------.~

~--~

Assumed Design Limit

60

I.

em;

..M

<D
<D
rodIJ~

.0

C;

.,. x

ox 0

10

o
OL-----__L -______L -______~--~~~------~------~~"------~

F~, (Bearing Capacity)

Explanation
Soil
Soi 1
<I> Soil
* Soil
+ Soil
x Soil
Soil
0
Soil
c;

Type
Type
Type
Type
Type
Type
Type
Type

Figure 37.

1
2
3
4
5
6
7
8

Silty Sand and Gravel

Crushed Rod
Sand and Gravel
Sand and Gravel
Si lty Sand with Rock
Sand and Grave 1
Silty Gravel
fine-Medium Sand

Soil
Soil
Soil
Soil
6
Soil
'V Soil
&l Soil

Type
Type
Type
Type
Type
Type
Type

9
10
11
12
13
14
15

Surge Stone
Crushed Stone
Sand with Gravel
fine-Medium Sand
Crushed Limestone
Sand and Gravel
Sandy Silty Gravel

Pullout test results - VSL bar mats and bearing

capacity factor design values.

66

For closely spaced anchors

IT

Fq

exp {2 3n
cos Q

Q)

tan </l']

( 13 )

Figure 39 shows a comparison between field pullout tests and eg.

12.
F values for the passive pullout resistance of deadmen can
be estimated using eg. 11.
c.

Determination of the scale effect correction factor

At present, there exists no theoretical formula to account for

the scale effect in the estimate of passive pullout resistance.
The following are preliminary guidelines but further research on
the load transfer along reinforcements that rely on passive earth
resistance is required to provide the necessary data basis for
more reliable methodologies:
A:1chor deadman
Inextensible metallic grids
Geosynthetic geogrids

a
0'.

1.

1.
< 1.

For geogrids the 0'. value should be obtained from pullout tests on
instrumented samples following the testing and interpretation
procedures outlined in section 2.b.
In the absence of pullout
test data use ~ = 0.6 for geosynthetics.
4.0

ESTIMATE OF THE PULLOUT CAPACITY OF COMPOSITE REINFORCEMENTS

COMBINING FRICTIONAL AND PASSIVE SOIL RESISTANCES

Several composite soil reinforcement systems (e.g., bar mats,

wire meshes, geogrids) mobilize their pullout capacity through
combination of interface sliding friction and passive soil
resistance on the transverse elements.
The relative proportions
of the pullout capacity mobilized through each mechanism depends
primarily upon the ratio of grid aperture size to soil particle
size.
The relative soil to reinforcement displacement required
to fully mobilize these two interaction mechanisms is
significantly different (see table 7, volume I) and therefore
their relative contribution to the pullout capacity depends upon
the displacement.
The ultimate pullout capacity of these
composite soil reinforcement systems is estimated per unit width
of reinforcement by superposing the interface sliding frictional
resistance P f onto the passive pullout resistance per unit width
of reinforcement p p
p

( 14 )

where:
~ =

As/AR is the fraction of the reinforcement

surface area that resists direct shear with soil.

67

140000

,-----------------------------------------~

o Pea Gravel
a Si lty Sand
.A Washed Sand

120000 -

Sand
100000

L...

<.n
0...

80000

60000

8
40000

20000

/~

//

,...,..{

"," ,,,
,I

,"

o-~--------~--------~----------~------~
o
1000
2000
3000
4000

Overburden Pressure (PSf)

Pp=Pullout capacity developed by passive resistance
n =Number of Transverse Wires
d =-Wire diameter
w =Width of mat
Figure 38.

Pullout test results - welded wire meshes.

68

Verticle Stress (psf)

SOD

250

750

1000

1250

1500
20,000

---_.---

70

17,500

Triangular --.
Elements

60

15,000

...

so

..

12,500

..a

<J

<ll
U

l.O

L...

--;
c..

lJ.

10,000
"'-

30

0
~

7500
20

5000

10

o
o

10

Vertical

Figure 39.

20

30

60

40

70

eo

Stress (kN/m 2 )

Relation between pullout force and vertical stress.

69

CL

where:

is the surface area of the elements of the

reinforcement and AR is the gross area of the
reinforcement.

Pp

F q IX.

IX" (J
t-'

'.

L e. C

Hence:
( 15 )
Note that eq. 15 extends the pullout equation to account for the
effect of the structural geometry on the pullout capacity of
composite reinforcement with:
( 16 )

The maximum pullout resistance of a planar composite

reinforcement is obtained fro the optimal structural geometry
( s x = S op t) de fined by eq. 9.
F*
a.

~;

tan

IX f

1; K

1;

O.

IXI3

Determination of F* for composite reinforcements

Welded wire mesh and bar mat systems
D

IT

s;-

-2-

where:

S is the lateral spacing between longitudinal

elements.
D is the diameter of the longitudinal
elements.
/1*

tan 0' K

2"

-S-i

Fq

18.5

D
S

+ 0.475
x

Sy

Welded wire meshes frequently used:

wI.7
w2.5
W3.4

0.148 in diameter D, 6 in x 6 in or 6 in x 9 in
0.178 in diameter D, 6 in x 6 in or 6 in x 9 in
0.207 in diameter D, 6 in x 6 in or 6 in x 9 in

Bar mats frequently used:

W11
w20

0.375 in diameter D, 6 in x 24 in
0.5 in diameter D,
6 in x 24 in

70

37;

Geogrids
The pullout resistance and direct sliding resistance of geogrids
depend upon the Grid Aperture/Average Grain Size (D 0) ratio.
Eq. 16 provides a reasonably conservative design value for Grid
Aperture/Grain Size ratio> 3.(69)
Hence:
For S x

< S opt

(Eq. 8)

F* = tan <fi ; a f
For S x

1;

> S opt
tan (, ; where
fri ction angle

(,

a f = a. s . where a
area In the gridS

is the fraction of solid surface

11* =

is the interface soil plastic

fb
af3

where:

0;

af3

s;
x

fb =

(1

S)
y

r is the width of the longitudinal element in the

grid
F*

= Fq

af3

+ tan 0 .

a5

For large soil particles that cannot penetrate the grid

aperture (Grid aperture/Grain Size ratio < 1):
F*

11

tan (,;

71

<X f

1; aJ3

8
E= 1.45 KSI
c

z
w
L

<..)

:5a...

""-

E= 14.5 KSI

(f)

2
, / E= 145 KSI

<t
W

, - E ::: 30500 KSI

a..

-5
REINFORCEMENT

Figure 40.

10

15

LENGTH, L(ft)

Peak pullout displacement for geosynthetic

reinforcements vs. reinforcement length.

72

5.0

SUMMARY OF PULLOUT DESIGN PARAMETERS

The pullout capacity of all reinforcement systems is estimated

using the general equation:
Pr

F'

(X

(X

(J

13

( 16 )

p'

( 17 )

The pullout capacity design parameters for the generic soil

reinforcement systems considered in this manual are summarized in
table 7 of volume I.
6.0

EVALUATION OF ALLOWABLE DISPLACEMENTS FROM PULLOUT TESTS

The soil to reinforcement relative movement required to mobilize

the design tensile force depends upon the load transfer mechanism,
the extensibility of the reinforcement material, its creep
characteristics and soil type.
As illustrated in figure 40, for
extensible geotextile sheet reinforcements, geogrids or plastic
parallel strips, the peak pullout displacement is primarily a
function of the reinforcement length.
Extrapolation of pullout
test results, to reinforcement of different dimensions requires a
careful evaluation of the scale effect as it relates to the load
transfer along the length of the reinforcement.
Estimate of the
scale effect requires an adequate estimate of the confined
stress-strain properties of the reinforcement and the appropriate
soil to geotextile (or geogrid) interaction mechanism.
An analytical solution for the front displacement-pullout force
relationshir for extensible reinforcements was derived
assuming: (7 )
(a)

Reinforcement is linearly elastic.

(b)

The interface layer is elastic perfectly plastic.

The front edge displacement Yo

A

Yo
where:

"2 yc
To

[ 1 + (_)2

yc

is given by:
T
EA

(_0 )2

( 18 )

is the pullout force

A = .rEA/kCb is a reference "transfer length"

E is the elastic modulus of the reinforcement, as measured in
a tension test.

73

A is the area of the solid portion in the cross section

of the reinforcement,

c is the effective unit perimeter of the reinforcement,

b is the gross width of the reinforcement,
k is the shear modulus of the interface,
is the shear displacement required to mobilize the
limit interface shear stress:
l
ky c .
max

The interaction parameter k can be derived either from direct

shear tests with soil-geotextile (or geogrid) interface or from
pullout tests.
Figure 41 illustrates the interpretation
procedure for pullout tests on extensible reinforcements to
obtain the interaction parameters k and A.
This procedure consists of drawing the pullout curve in the plane
of (T/EA)2 vs. Yo'
For Yo > y the experimental curve can be
assimilated to a straight line~
The linear regression of the
experimental results will provide:
An initial Yo

intercept equals yc/2.

A slope equals A2 /(2y c ) =

n.

The soil-reinforcement interaction parameters can then be

calculated as:
A

7.0

= [2y c

EA

n]lj2

= Cb . >::i tan

SOIL NAILING - ESTIMATE OF PULLOUT CAPACITY

The load transfer mechanism between the nail and the subsurface
soil (or rock) and the ultimate pullout capacity depend upon
several parameters including:
installation technique, drilling
and grouting method, grout pressure, size and shape of the
grouted inclusion, engineering properties of the in-situ soil and
specifically its relative density (or overconsolidation ratio),
hydraulic conductivity, and shear strength characteristics.
a.

Grouted nails

Grouted nails are generally gravity grouted.

Their pullout
resistance is therefore expected to be approximately the same as
that of an equivalent straight-shafted anchor, tremie-grouted
under low (or no) grout pressure.
The pullout capacity of these
anchors if often estimated by:

74

01

PULL OUT CURVE

<t
t-w

(/)
(f)

a::

a:
t-

(/)

0::

4
W

:r:

(/)

<...)

0
LL
f-

::>

""

\.

"

"\
()

Tmol: 0+

..-J
..-J

Y = Yc

:::>
CL

FRONT DISPLACEMENT
I Yc
I
I FRONT DISPLACEMENT Yo
I

of <t
'-=w

=
Yo

I
I

~
2

Yo

L+ (~)2

fl

. (

Yc

To
EA

A :: ['A I (K'Cb)] 112

Tmax

Tmox. :: ton

'It. Yh

FRONT DISPLACEMENT Yo

Figure 41.

Interpretation procedure for pullout tests on

extensible inclusion.

75

= n

where:

D L

(19)

Tu 1 t

Tult
is the ultimate lateral shear stress at the
ground - grout interface (also called shaft
friction) ,

D and L are, respectively, the effective diameter

and length of the grouted anchor.
It is commonly assumed that:
In competent rocks:

= 10% * Sa for S a < 600 psi

where:

Sa

( 20)

is the uniaxial compressive strength.(71)

In cohesive soils:

=o::s u
where:

(0::)

(21)

is an adhesion factor,
is the average undrained shear strength of the

s~il.

The adhesion factor (0::) generally varies within the range of 0.3
to 0.75 with the lower values obtained for stiffer and harder
clays.(45, 72, 73)
In granular soils an "apparent friction coefficient" is generally
used to take into account the restrained dilatancy effect on the
soil-nail frictional resistance:
Tult

where:

y'

y'

( 22 )

h 11

is the effective unit weight of the soil,

h is the overburden height above the nail at the

mid point of its adherence length,
11

is the apparent friction coefficient.

It should be indicated that the effective diameter D of the

grouted nail is difficult to estimate since it is highly
dependent upon the installation process, ground porosity and
grout conductivity.
The drilling of the borehole for the grouted nail produces an
unloading of the disturbed surrounding soil that can
significantly affect its mechanical properties.
The soil-nail
interaction is primarily dependent upon soil recompaction due to
grouting.
In cohesionless soils, grouting pressures of 50 to 100
psi are commonly used to prevent caving as the casing is

76

withdrawn.
This pressure grouting will induce ground
recompaction associated with grout penetration into permeable
gr~velly seams, thereby increasing substantially the pUllout
resistance of the nail.
In a fine grained cohesive soil the
tremie-grouting results in a rather smooth soil - inclusion
interface.
The presence of water at the interfaces, specifically
in plastic soils, will generate a lubrification effect decreasing
substantially the pullout resistance of the nail.
Figure 42
shows a summary of pullout test results obtained with low
pressure grouted nails in different types of soils.(74)
In the
majority of these projects, the nails were installed in
preaugered boreholes.
The results illustrate the variability of
pullout resistance and the difficulty in extrapolating pullout
information from one site to another.
The results support the
need for pullout testing, even for preliminary design evaluation.
b.

Driven nails

Pullout tests on driven nails in a granular nailed wall reported

by Cartier an~ Gigan have shown that the apparent friction
coefficient ~ values correspond to the design values generally
used for the ribbed metallic strips in Reinforced Earth walls
(figure 43).(37)
At relatively
depth, due to the restrained
dilatancy effect, the value of ~ is significantly greater than
tan ~ and it decreased with depth to tan~.
However, the
construction process will significantly affect pullout capacity.
Laboratory scale pullout tests in a medium dense sand illustrated
that Reinforced Earth (i.e., placing the inclusion during the
construction and compacting the sand around it) produces a
substantially higher apparent friction coefficient than nailin~
by driving the inclusion into the compacted sand embankment.(4 )
In the latter case, nail driving will significantly reduce the
retained dilatancy effect on the pullout resistance.
Therefore,
design guidelines for Reinforced Earth walls cannot be safely
extrapolated to soil nailed structures.

tOW

c.

Jet grouted nails

These nails are installed under a grout pressure than can exceed
20 MPa and is sufficientl high to cause hydraulic fracturing of
the surrounding ground.(7)
Similar to high pressure grouted
anchors, the jet grouting installation technique produces a
mechanical interlocking between the penetrating grout and the
surrounding ground which results in a substantial increase of the
effective nail diameter.
It also provides recompaction of the
surrounding ground that significantly improves the pullout
resistance of the composite nailed soil inclusion.
Field pullout
tests on jet grouted nails yielded ultimate lateral shear stress
values as high as 400 kpa in sands and 1000 kpa in sandy
gravels.(74)
d.

Estimate of the pullout resistance

To date, estimate of the pullout resistance of nails (or ground

anchors) is mainly based upon empirical formulas (or ultimate
77

T(KIPS)

T _,

. 100 KN/M
40

30

20 .

fO

T CLAY

o
L(ft)

Figure 42.

Pullout test results.

78

500

1500

1000 1500

500

2000

.fA

TL (lb/ft)

10
A

20

?fJ

~r
g

1-

~VERAGE

10

1165 Ib/ft

wr
~~ -

20.

~-

30

SPECIFtCA TlONS fOR

REINFORCE.D EARTH

.-TAN

q,

40'

40

Z (fO

Z (ft)

SOl L: SAND 4=33 c= 17psi

NAI LS: DRIVEN f'ROFllE

Figure 43.

2000

Pullout test results on driven nails

in granular soils.

79

lateral interface shear stress values) derived from field

experience.
These formulas are useful for feasibility evaluation
and preliminary design.
Table 2 provides a summary of ultimate
interface lateral shear stress values as a function of soil (or
rock) type and installation technique.(76)
Recently an increasing attempt has been made to develop field
correlations between the ultimate lateral shear stress ~ I t
mobilized along anchors and nails and the engineering properties
of soils obtained from commonly used in-situ tests, specifically
the standard Penetration Test and the self boring pressuremeter
test.
The available field data pertaining to the pullout capacity
of nails is presently still too limited to substantiate
development of reliable correlations. An attempt has been made to
predict the pullout capacity of both driven and grouted nails
using the French recommendations for the determination of lateral
shaft friction on bored and driven concrete piles from
pressuremeter test results.(77)
Figure 44 shows that in fine
grain soils (i.e. fine sands, silts, non plastic clays) predicted
c ~
values correlate reasonably well with pullout test results
wh1ie in dilatant gravely soils, compacted moraine or fissured
rocks they generally underestimate the measured ultimate lateral
shear stress.
It appears that further research and particularly field testing
could significantly improve the data base for estimating the
pullout capacity of soil nails.
However, in light of the large
variability of parameters affecting the load transfer mechanism,
specifically in heterogeneous soils, empirical correlations can
only be used for preliminary design.
Pullout tests are required
to provide reliable data for final design and to verify the design
during construction by testing non-service witness nails.
8.0

LONG TERM PERFORMANCE AND CREEP CONSIDERATIONS

Long term performance of the reinforced soil system depends

primarily upon the creep potential of the reinforcement and the
soil.
In order to assess long term pullout capacity of soil nails
in the fine grain soils and geosynthetic reinforcements and to
evaluate the potential of the ground inclusion system to creep,
the pullout tests should be load controlled.
a.

Soil Nailing

The field testing procedure currently used for soil nailing is

similar to anchor loading test.
It usually consists of 10 to 20
min. sustained load increments of 0.15 f
(where f
is the yield
strength of the steel nail).
Each load Increment Is maintained
until measured deflection is negligible (i.e., displacement rate
smaller than a specified displacement increment per log cycle of
time). The incremental loading is applied until pullout failure
is generated.
The tested nail, minimum length 8 ft., should be
installed using the same installation process of production nails.

80

Table 2.

Ultimate Lateral Shear Stress Data for Preliminary

Design of Soil Nailing.(74, 76)

Soil (or Rock) Type

Construction
Method

ultimate
Lateral Shear
Stress (psf)

Silty sand
Silt
Piedmont residual

Rotary drilling

2,000 1,200 1,500 -

Driven casing

6,000
8,000

Sand
Sand/gravel
Granular Soils
Dense moraine
Colluvium

4,000
1,600
2,500

8,000 - 12,000
2,000 4,000

Sand
Sand/gravel

Jet grouted

8,000
20,000

Silty sand fill

Augered

400 - 600

Augered

400 - 600
800 - 1,200
1,000 - 2,000

Soft clay
Still clay
Clayey silt
Cohesive Soils
Calcareous sandy
clay
Clayey colluvium

4,000 -

6,000

Driven

1,000 - 2,000

Marl/limestone
phillite
Chalk
Soft dolomite
Fissured dolomite
Weathered sandstone
Weathered shale
Weathered shist

Rotary drilling
(dry)

6,000 2,000 10,000 8,000 12,000 4,000 2,000 -

Marl
Stily marl

Augered

4,000
6,000

81

8,000
6,000
12,000
12,000
20,000
6,000
3,000

20

Ronoe of Measu(d Values

f<onQt of Calculated Volyc:s

02

11

6q::,J

'+IJl
.:>C.

2
9

I-

<::(

-I

..J
:::::>

.....J

-I

<::(

U
::J

-I

t,o

2
L.ult

20

MEASURED (ksf)

lEGEND:

1.
2.
3.
4.
5.
6.
7.

Driven bars in fine graired soil

Grouted bars in fire graired soil
Drive.n bars in granular soil
Grouted bars in weathered rod<.
Grouted bars in soft clay
Grouted bars in stiff clay
Grouted bars in rrarl
a. Grouted bars in still rrarl
9 . Grcuted bars in clayey silt

Figure 44.

10.
11.
12.
13.
14.
15.
16.
17.

Drilled "..Gr.cluted bars in silt

Drilled " Grouted bars in silty san:]
Dri"\len casing grouted bars in s.:md
DriVCJl casing grouted bars in M;)raj r
Driven cas:in:l grouted bars in Co1-1uviun clayb
Drilled" Grouted bars in rrarl-1 in~stcne
Drilled" Grouted bars in soft rod<
Drilled" Grouted ba..rs in fissured rock

Comparison between measured and estimated values for

ultimate lateral shear stress.

82

In soils susceptible to creep, the critical creep load should be

established, following a procedure similar to that used for ground
anchors which is illustrated in figure 45.(33)
For each load
increment the measured nail displacement (s) is plotted versus
Log. Time (T). An upward concavity of the creep curve indicates
an accelerated creep inducing failure.
The slope of the (s) vs.
(Log. T) line is plotted against the applied pullout load to
determine the critical creep load Tc .
The time dependent nail displacement under a constant load can be
consistently described using Singh and Mitchell's creep model.(78)
For ground anchors, the creep displacement of the nail under a
sustained load can be estimated using Singh and Mitchell's type
equation:
( 23)

61
61

where:

+ Ae aT

In

for m = 1

(24 )

T and 61

are respectively the applied pullout

force an~ the initial displacement prior to creep,

A, a and m are interface creep parameters that are

obtained from the experimental "log 61 - log t"
(i.e. displacement versus time under constant
load) and "log 61 - T" (i.e. displacement versus
applied load) curves, (see Figure 46).
61 is the

displacement rate.

Figure 46 illustrates the creep behavior of an anchor in a plastic

clay and the determination of the relevant interface creep
parameters.(79)
The "m" parameter which is the slope of the "log
61 - log t" linear relationship indicates the creep potential of
the soil.
"m" values smaller than one indicate a relatively high
potential for accelerated (or tertiary) creep inducing a creep
rupture.
Geosynthetic reinforcements
Creep considerations witn regard to the long term performance of
geosynthetic reinforcements have been outlined in NCHRP Report
290 and are briefly summarized below.(2)
The time dependent stress-deformation behavior of unconfined
polymeric materials, schematically illustrated in figure 47,
involves instantaneous recoverable primary creep, long term non
recoverable secondary creep and tertiary creep to rupture similar
to that observed in soils by Singh and Mitchell.(78)
The creep
potential depends upon the basic polymer properties, structural
aspect (woven vs. nonwoven), manufacturing process (e.g., heat or
resin bonding vs. needle punched) and environmental factors,
specifically temperature.

83

b) Critical Creep Tension

.5 .6

TC

.} Creep Curves

Figure 45.

Anchor tension test for determination

of critical creep load.

84

.7.8.9

'Ig

-3
~ ~------------------~~------~

10

Time (min)

Rate of displacement-time relationship

~.

m>l

~~~~~-----11 A e~D

... - -- -
50

------------------------- m::;
.-...;..;.~_ (,I-m_ 11
-m
100

150

200 KIPS

log

Rate of displacement-load relationship

Figure 46.

Typical creep curve&

Modeling creep of anchors in clays.

85

Rupture

c:

Pr im3ry

Creep
Secondary
Creep

Tertiary
Creep

Log Time

Figure 47.

phases of creep for a typical geotextile tested

without soil confinement at constant
load and temperature.

86

The main design concerns are (1) to evaluate the critical creep
load or creep strain below which creep rupture is unlikely to
occur, and (2) to predict the long-term creep displacement under a
constant tensile load.
For design purpose, creep test data may be plotted in the form of
isochronous curves to allow extrapolation of the test results to
the specified service life of the structure under a given
performance limit strain.
Figure 48 shows creep tests data
obtained for Tensar SR-2 at 20C. The data may also be plotted as
shown in figure 49 is recommended for the determination of the
critical creep strain data.(SO)
The data reported in figure 48
suggest that for Tensar SR2, creep rupture is unlikely to occur
for overall strains smaller than 10%. The critical creep strain
translates to a critical creep load through the use of stiffness
isochrones shown in figure 49 that provide the deformation modulus
for the specific service life of the structure.
The creep response of geotextiles confined in the soil may be
substantially different from their in-isolation performance.
The
confinement effect is highly dependent on the structural aspect.
It is generally significantly greater with nonwoven geotextile
where the combined effect of grain interlocking and soil
confinement substantially reduces the short and long term creep
deformation.
Figure 50 shows the confinement effect on the
load-deformation time behavior of nonwoven needle punched
polypropylene Terram 1000 and nonwoven polyester Bidim U24.(Sl)
Apparent unconfined creep testing of geotexti1es can significantly
overestimate the in-soil long term deformation.
Load controlled pullout tests should be conducted to evaluate the
in-soil confined creep behavior of geosynthetics. The creep
displacement is function of material properties and reinforcement
length.
Similarly to the evaluation of the scale effect
correction factors for extensible reinforcements it is necessary
to evaluate the scale effect in extrapolating the results of
laboratory creep tests to reinforcement of actual length.
This
scale effect can be derived from the results of pullout tests on
reinforcements with different lengths or assessed using analytical
or numerical load transfer models which have to be "calibrated"
through numerical test simulations.
The Singh and Mitchell creep
model can be used in the interpretation of the pullout tests to
derive relevant creep parameters for the extrapolation of the
laboratory test results to actual reinforcement size.(78)

87

6O~-----4------+-

4000
10 hrs.

3000

....,

40

4...........

--

..0

~30

10,000 hrs

"

~
0

-'
20

2000

Extrapolated
for 120 years
1987 1b/ft at
10~~ strain

10

Sample Size: 5 bars x 15 ribs

I
0

.c

12

I
16

Total Creep Strain (%)

Figure 48.

Load versus total creep strain for Tensar SR2.

88

-0
to
-1

lOOO

Toa! creep strain (')'.)

1..0

10..0

20.0

15.0

-1

Units on curves
are lb/ft
I
I
-- _____ -- 2820
2810

2640

I
~~~~-r_+_~~I~~--4-------~----~--~

~~2040
00

r-

-1-----+----

Figure 49.

ur'SR2
rnple size: 8 bArs

J:

3 ribs - - - - -

5 bars

J:

1Sribs ---...:----

p:20"C

Creep strain rate against total creep strain.

89

30

UnconfonC'd

In-lsot::lt 101'1--;'

I
~

40

12

In-lsolotlOn " ~ _

30

--

-=<
c:

+-

ell

16

, ,

., '"

.-

Confining

~12

Con11nlng p r e-5Su r C'

100 kNIm2

pr~ssurC'

100 HU";

~20
K

<

10

A)

Figure 50.

or-

-----------

20

UncO'">flnC'd

Terram 1000

B)

Bidim U24

Confinement effect on creep behavior

of non-woven geotextiles.
90

CHAPTER 3
REINFORCED FILL WALLS
1.0

INTRODUCTION

AS previously indicated in the Introduction section, all the

equations and design methods presented in chapter 3 of volume I
were based and verified by recent research carried out in this
country and abroad. A part of that research was performed in the
frame of the present study in order to prepare this manual. This
section provides the main results of all that research, as it
relates to the design methods in volume I.

2.0

LIST OF RECENT RESEARCHES, INCLUDING FHWA PROGRAM

a.

General reports on soil reinforcements

One of the principal variations in reinforced fill wall design is

the extensibility of the reinforcement.
The first indication of
the influence of reinforcement extensibility on reinforced soil
wall behavior has been presented by Schlosser, et al., who
reported the results of a full scale experimental wall constructed
with Paraweb (i.e., extensible) reinforcements and compared it
with the behavior of Reinforced Earth walls (inextensible
reinforcements).(12, 82)
Since that report, several general
reports partly or totally devoted to soil reinforcement, have
mentioned this aspect of the behavior, most of which are reviewed
in NCHRP 290.(2)
b.

FHWA research program on reinforced soil wall

The FHWA research program performed for preparing this manual

included many types of tests, including:
Reduced scale three dimensional models of reinforced
soil walls with different types of reinforcement:
aluminum foil, plastic strips, plastic grids, non-woven
and woven geotextiles and anchors.
Small scale centrifuge testing of reinforced soil walls
using the same reinforcements used for the reduced scale
model tests.
Full scale experiments of eight 20-foot high reinforced
soil walls, with different types of reinforcements,
facings and fill materials (table 3).
Large scale centrifuge tests on models of the full-scale
walls.
Parametric study of reinforced soil walls using a
calibrated finite element computer program.

91

Table 3.

List of full scale reinforced soil walls constructed

for the FHWA program.

Wall

Reinforcement

Facing

1.

Metallic strips
( 40 x 5 mm 2 )

Concrete panels

Sandy gravel

Rectangular

2.

Geogrid

3.

Bar mats

Concrete panels
concrete panels

Sandy gravel
Gravelly sand

Rectangular
Rectangular

4.

Bar mats

Concrete panels

Cobbles

Rectangular

5.
6.

Concrete panels

Silt

Rectangular

Geotextile

Gravelly sand

Rectangular

7.

Ba r ma ts
Non-woven
Geotextile
Wire mesh

Gabions

Gravelly sand

Rectangular

8.

Wire mesh

Gabions

Gravelly sand

Trapezoidal

c.

Fill Material

Geometry

Reinforced Earth Company's Research

The Reinforced Earth Company has performed significant research in

the last ten years, the results of which are contained in internal
reports. At the onset of the FHWA study, the Reinforced Earth
Company agreed to provide results from this research.
The following results were used in the development of this manual:
Inclination of the thrust at the back of the wall,
resulting from a parametric study using a finite element
method (F.E.M.) program.
Full-scale experiments on Reinforced Earth walls with
short reinforcement length (L/H = 0.4).
Behavior of Reinforced Earth walls under seismic
loading, resulting from a F.E.M. study using a
calibrated computer program.
d.

Research on polymeric reinforcement

In 1987, NATO sponsored an Advanced Research Workshop on the

"Application ( of
Polymeric Reinforcement
in Soil Retaining
8 3 )

Structures".
The workshop cons1sted of reports on case
histories from Europe and the united States, evaluation of
material properties, analytical techniques and design methods,
construction methods and economics, and a research needs summary.
A major focus of the workshop was a prediction exercise in which
participants were asked to predict the performance of two full
scale walls that were constructed, externally loaded and
monitored, as part of the program.
The information from this
workshop provided significant information concerning the
predictive capabilities of existing design methods for extensible
reinforcement.
92

3.0

SUMMARY OF FHWA RESEARCH RESULTS

a.

Small scale models

Results of all model tests performed in the FHWA project clearly

indicate that reinforcement extensibility and density (amount of
reinforcement) significantly influence the design model.
The
small scale model results found:

For the less extensible reinforcements (red plastic

strips, plastic grids), distribution of the tensile
force is close to the K line at the top of the wall and
less than the K line at the bottom.
This is in good
agreement with all observations made on metal strip
model walls.
For woven geotextile polyester strips, K line would be
still valid at the top but the distribution at the
bottom is closer to the Ka line.
For nonwoven geotextile strips, tensile forces would
agree with the Rankine's distribution (K a ) .
Maximum tension in the more extensible plastic ~trips
appears to correspond to the Rankine's distribution on
the whole height, which is in agreement with
observations on actual structures built with extensible
reinforcements (paraweb strips).
b.

Full Scale Field Wall

Similar results in the small scale models have been obtained from
the full scale experiments performed for the FHWA project and by
F.E.M. calculations.
For instance, figure 51 shows the comparison
of the K lines for full scale model walls 1 and 2 in which the
only difference is the type of reinforcements.
For wall 1 with
metallic strips, the distribution of maximum tension with depth is
close to K i for wall 2 with geogrid reinforcements, the
distribution is close to Ka' except in the upper 2 meters.
Likewise, figure 52 shows a comparison of the K line for wall 3
constructed with inextensible bar mats and wall 7 constructed with
extensible woven wire mesh.
Again, the distribution of the
maximum tension with depth is close to Ko for wall 3 and close to
Ka for wall 7.
Another interesting result is given by the comparison between
walls 4 and 5. They are both constructed with VSL bar mat
reinforcements with the only difference being the fill material:
cobbles and gravelly sand in wall 4, clayey silt in wall 5.
Distribution of maximum tension with depth in wall 4 was found to
decrease from Ko condition at the top to a Ka condition at the

93

"t]
.....
I..Q
c:

'"1

ro

U1
I-'

HI
HI

~~

ro
rt

(')

OHI
;:l

rt
rt::r
::rro
ro
ro
~X
rt
(') ro
o ;:l
ro UI

I.D

Hl .....

.t::>.

HIti'

..........
(')1-'
..........

rort
;:l~
rt

6f-

3
0

f-

10

CD
J:.

a.

'7

-rt
::r
ro

;:l
HI

0
'"1
(')

ro
!3
ro

;:l

rt
UI

12 I-

o.

-g.

.c

D\

1
1

,
\

Ka
0 1

20 L.

Ko
,

a' \
\

10

\
\

12

~,

\
\

Distribution of Hi1~ffT1ur.r Ten"lon With Oeoth 1n \,'ttl1 1.

(reinforced ~art.h IItrlps)

,,~\
\
\
\

14

,a

,,
\

\
1

\1Ko

, \

.j

1 6 "1
\

(J)60 \

I-

IE

\,.

ell

I-

.2

141-

\"
\
\

"

0.

\'

16

'0

\
\

00

.....

o,,~

'"1

ro

-0

",-t\
.co

\
\

-;;

~:,

\\

I-

, I

0.

:I

'0

Maximum Ten,lon (kilt)

\ \

a~

Fini!e Elemen!

"""""'" I "

'5

Mea5ured

I'

\ \
\ \
I- ('0\

o Fini!e Elemen!

Mo,imurn Tension (kilt)

,f "

ttl

o Measured (11/19/87)
o Measured (12/24/87)

,
' Ko
\

20

b)

DI.trlbut Ion of li"lmu," Tons[on !:Ith Dopth In ~'.l J 2.

(Ten .. r g,ogrlda)

"fl
~.

LO

c:

t"'\

en

U1

tv

Maximum Tension (kilt)

2
1

::l

\ \

()

'\

1.0

art"
c: ro
a ::l
III
rt" ~.
roO"

U1

t-'.

::l

Ill ......
t-'. t-'.

art"
::l~

:0

\
\

X ro

~.><

\:

.-

, ...
,,
,, <>\
,,
,,
,
'00
,

'0
0-

....0

"
.c

12

a.u

14

c:
rt"
...,a

::l

o'iJ

,,\0
I
I

I Iii

I
I

,,
,,
,
\
\

18

f-

co

Distribution of Ha:dmllm tl"n51C'1n lath OC!pth in \\'.111 l

(VSL bar mats)

,
\
\

' \ Ko

,
\

\
\

\ Ko
10

Cl

,,
,,

12

g. 15

20
.)

:5

\ 0

18

Qj

\
\

0
l-

\
16

,,

a.

~,

0"

"0

0,\

(fl

t"'\

:::

t-'.

rt"

IIi

I -cq
I
\
I
\
I
\
I
"

10

OJ

f-o

'-'

".::l

0-

'5
3

Iii I I I I I

1\
1\

1\
I \

::l

I i I

.1\

,'\

c:

ro

PI

o ~I

......

Maximum Tension (kill)

I
2

o ro
m
a
am

o Measured (12/04/87)
a Measured (I 2/23/87
o Finite Element

Measured f I I! I 9 !87~
Measured 12/24/87
Finite Element

0
0
0

I
\

,,
,

21 '
b)

Distribution

or

Hodmurn T.n.lon Il1th D.pth in 11 1,

(Macc.terri gablon woven ..,1re ml!sh)

bottom in good agreement with other results on inextensible

reinforcements. On the contrary, in wall 5 the distribution is
closer to K line, indicating that the greater lateral
displacements observed in this wall during and following
construction directly influenced the K value.
An indepth evaluation of the field results was performed using a
finite element computer program.
c.

Summary of Parametric Study on Reinforced Soil walls

A parametric study of reinforced soil walls was made using a

validated, discrete finite element computer program, SSCOMP, to
study the effects of wide variations in structure geometry,
loading, foundation soil type, wall facing type, and soil
compaction on internal stresses and deformations. The computer
program was validated during an earlier phase of this research
project by showing that it predicted actual stresses in 15
full scale structures with good accuracy. The finite element
analysis results were also compared with conventional
design/analysis assumptions in an attempt to gain insight into the
strengths and weaknesses of the conventional design procedures.
Scope of Study - The study included a range of wall conditions
which are variations on a reference wall called the baseline case.
All cases that were analyzed are illustrated in table 4.
Variations in backfill and reinforcement type were not included in
this study.
The performance variables evaluated were maximum
reinforcement tension, tension distribution along the
reinforcement length, and horizontal displacement of the wall
facing.
Since the baseline case is representative of a typical highway
wall, as described in the following section, the results of this
study are expected to be applicable to many common FHWA design
situations. As with any analytical study it is uncertain whether
the results obtained are applicable to wall conditions outside the
range studied.
computer Program SSCOMP - The program is a plane-strain finite
element code which can simulate incremental soil placement,
incremental soil compaction, and soil structure interaction. The
analyses for this study were performed using the three major types
of element models available in the program, namely soil,
structure, and interface elements.
Soil elements were organized
into layers corresponding to construction lift intervals.
Structure beam elements (including bending stiffness) were used to
model the concrete wall facing and structural bar elements (axial
stiffness only) modeled the reinforcement layers.
Interface
elements with essentially zero thickness were used along the
soil/reinforcement and soil/wall facing contacts to allow relative
movement between the soil and reinforcement or wall facing.

96

Table 4.

fJi
Wt

14ft

b.selln~

Short Rein(orcement
at Bottom of Wall

Compare FEM results

to FEK Baseline Case:

'--1-5.5

A.

Rcinforcem~nt

tension
1.

2.

ft

1. Increase of 0-70\
in upper 4 reinf.

levels. Decrease of

Max. RainL
Tension

30-50t in bottom 4

Distribution
along roin!.
length

2. Location of max.

J. Sum of max.
tension for

all levi!ls

levels.
tension shifts
distinctly toward
wall fac& for bottom
4 levels. Pattern in
top 4 levels becomes
irregular.
J. Value 15\ lower

than Baseline case.

Comparison of FEM analysis cases. (61)

Short Reinforcemant

':r~
~.'
1-14

Variable Vertical
neinf. Spacing

fto!

1. Change of 70t in
top 3 reint. levels -

1. Values are similar

but direct level-bygreatest increase in

level comparison not
2~ level from top.
possible since reinf.
Less than tlOt change elevations differ
in bottom 5 levels.
from Baseline case.
Max. tensions 1n
2. Location of max.
bottom half of wall
tension shifts
are more uniform than
distinctly toward
in Baseline case
wall face for top
(So .. Figure 41 (bll
reinf. levels.
2. Location of max.
3. Value 5t less than tension same as
Baseline case.
Baseline.

Sloping Backfill
(modeled as sloping
!lurch.ar~.
2H,lV
lHll

.....
o
N

~o

Full-Height Braced
Wall Facing

im
.. 14

..
'"

... ,

-11\ ' -

1. Increase of

approx. 25t at all

rain!. levels.
2. Location of max.
tension is further
away frorn wall face
for most reinf.
levels.

J. ValUe 20t greater

than baseline.

tt ..

,,""

-0

C4SC6

Results
are
after

4'~~~~~al

1. Decrease of 0-350\

In lowQr 7 rein!.
levels. Increase of
about loot in top
level.
2. Location of max.

tension unchanged in
lower 6 levels;
shifts to wall face
for top 2 levels.
3. Value 15 \ less

than Baseline case.

J. ValUe 3t smaller
than Baseline case.

\.0
-..J

n. Lateral Wall racQ

Max. deformation 150\

greater than Baseline
with most of the
increase occurring in
the shortened layers.
Deformation profile
has a distinct bulge
12 ft. below the top
of the wall.

Deformation sarna as
Max. deformation 60t
Baseline in lower
greater than
.
halt of wall, but
Baseline. Location of
greater in upper balf max. deformation
with location of max. shifted downward to
shifted downward.
about 7 ft. below the
Max. deformation 60t
top of the wall.
greater than
Baseline

Similar to Baseline
but max. deformation
70t greater.

Location of max.
deformation shifted
down to near wall
mid-height. Max.
deformation JO\ less
than Baseline casco

Compare FEM Reinf.

tension to value
predtct"d by
conventional design

tlo conventional
calculations
performed.

No conventional
calculations
performed.

No conventional
calculations
performed.

Used Conv. Method

Deformation

me thods

1. 1lax. tension for

individual levels
2. Sum of max tension
for all levels

Used Conv. Method A

FEM values
generally 110-170t of
conVentional, though
eot at bottom" level~
Conventional values
almost uniform at all
rein!. levels.
1.

2. FEM value 120t of

conventional.

1. FEM values for

middle 6 levels

85-120 \ of

conventional; for top

level about 500 \ of
conventional.
2. rEM value same as
conventional.

Table 4.

~r~
~""""'

Comparison of FEM analysis cases.(61)

~ ini,' ,',' .. ,

h. Reinforcement
tension
1. Max. Rein!.
Tension
2. Distribution

along rein!.
length
). SUIU of max.
tension fer
all levels

~m.

,2

Increased Foundation
Compressibility
--14 '-...

~ ~

"'000.

case 10

Compare FEM results

to f'EH Oasolino Caso:

Strip Loading
(Crane Load)
1 ftTr2 ft

strip Loading '1

(Bridge Seat)

lOa

'"
-l~'-

1. Large increase in
upper 2 levels about
equal to max. value
in 2~ level frOm
bottom in Baseline.
Fairly uniform
increase in lower 6
levels of magnitude
of about )0\ of the
above max. value.

_-,,:'-'!

s~il

stiffness
20-30\ of baseline

-14'-

1. Large increase at
all levels equal to
100-250\ of max.
value in 2~ level
from bottom in
Baseline. Max.
tensions are fairly
uniform in middle 2/)
of the wall - near
yield for 60 ksi
steel bar mat.

case_

1. No change, except
slight increase in
tension in the lowest
layer.

2. Essentially same
as Baseline case.
). Values 0-5\
greater than Baseline
case.

2. Locations of max.
2. Locations of max.
tension near the
center of the strip
load.

tension shift toward

the center of the
strip load.

3. Value 300\ greater

3. Value 80\ greater

than Baseline case.

than Baseline case.

B. Lateral Wall Face

Deformation

Compare FEM Relnf.

tension to value
predicted by
conventional design
methods

Similar to Baseline
but deformation
increasing all the
way to top of wall.
Max. deformation 40\
greater.

Deformation same as
Baseline in lower 13'
of wall height, but
greater in upper 7'.
Max. deformation 25\
greater than
Baseline.

Deformation at all
points on wall face
(including bottom of
wall) greater than
Baseline case by a
constant increment.
See Table 4 for
values.

Used Conv. Methods B,

C, and D.

Used Conv. methods B,

C, and D.

Used Conv. Method A

Comparisons given in
Figure 40 and in
Table 4.

Comparisons given in
Figure 40 and in
Table 4.

1. Max. tension for

individual levels
2. Sum of max tension
for all levels

1. FEM values for

bottom 7 levels
100-135 \ of
conventional: for top
level 250 \ of
conventional.

2. FEM values about

125\ of conventional
values.

Definition of Conventional Hethods:

A. Conventional Meyerhof procedure for bar ~at reinforced soil
valls, as in Mitchell and Villet (1987), chapter 5.
B. Increase in vertical stress due to strip loading calculated by
1H:2V load dispersion vithin reinforced zone, as in Mitchell and
Villet (1987), appendix A, chapter 1.
C. Increase in vertical stress due to strip loading calculated by
Boussinesq elastic load dispersion vithin reinforced zone.

D. Laba and Kennedy (1986) load-sharing procedure for strip loading

effects.

98

Table 4.
~t!~~~

...
'"o

~basellne

Comparison of FEM analysis cases. (61)

Flex ibl e Facing

(Geotextile wrap)

l~
~ou.

Compara FEM results

to rEM Baseline Case:

No Compaction
Stressell
r

1=14

Increalled Compaction
Stresses

tH

l-----.c e

'"
-'-

A. Reinforcement
tengion
1. Hax. Reine.

Tension
2. Distribution

along reine.
length

). Sum of max.
tension for
all levels

'"01"

rH

(t.-l

~-

___ eA8e Be

'"

....LJ---

1. Decrease of 10-l0\

2. Locations of max.
tension are 2-l ft.
closer to the wall
face for all but tho
top rein!. level.
Reduction in tension
at the wall face is
more pronounced.

2. Location of max.
tension same as
Baseline case.

2. Location of max.
tension same as
Baseline case.

3. Value 15\ less

3. Value 10\ greater

than Baseline case.

in all levols.

than Baseline case.

rF==

...

Increase of 5-35\
in all levolll.
1.

Uniform Surcharge
(400-1200 psf)
q I I I I I

1. Increase of 0-20\

in all levels.

3. Value 20\ greater

than Baseline caso.

'"
o

J.--

I---

case
98, 9b,
9c

-'--'__ 14 ' -

1. Relatively uniform

increase at all
reinforcement levels.
Magnitude of increase
close to qSyK
2. Distribution shape
similar to Baseline
but location of max.
tension further from
wall face in upper 3
reinf. levels.
3. Value increases by
3l\ of Baseline value

I.Q
I.Q

for each 400 psf

increase in
surcharge.
B. Lateral Wall face
Deformation

Similar to Baseline
case but max.
deformation 20\
greater.

Similar to Baseline
but location of max.
deformation shifted
downward to about 6
ft. below tho top of
the wall, and max.
deformation 25%
smaller.

Similar to Baseline
but max. deformation
10\ greater.

Deformed shape
similar to Baseline
but max. deformation
greater by about 25\
per 400 pst
surcharge.

Compare fEM Rein!.

tension to value
predicted by
conventional design
methods

Used Conv. Method A

Used COny. Method A

Used Conv. Method A

Used Conv. Method

FEM values
125-250 \ of
conventional.

1. FEM values for

1. Max. tennion for

2. FEM value 145% of

conventional.

middle 4 levels
100-10S\ of
conventlona l;
B5-125\ ot cohv. at
other levels.

1. FEM values
100-140 \ of

individual levols
2.

Sum of max tension

for all levelS

2. FEM value same as

conventional.

l\

1. FEM values
100-140 \

conventional.

of conventional.

2. FEM value 130\ of

conventional.

2. FEM values about

120\ of conventional

values.

Value of
is
within 3\ of FEM sum.
(0.5(., )HJ+qH]K.

Table 4.
~14

fW

Comparison of FEM analysis cases. (61)

Reinforcement Length,
L- 5.5 ft

~""' lL.
lli

Compara rEM results

to rEH Baselina Case:
A. Reinforcement

tension
1. Max. Rein!.

tension
2. Distribution

along reinf.
length
J.
>-'

o
o

Sum of max.
tension for
all levels

0 ...

Reinforcement length,
r,.. 0.5 it
1------+ 8

...
"o

"

1. Increase of JO\ in
upper two rainf.
levels. No change in
next two layers.
Decrease of 20-50\ in
bottom 4 levels to
values near
z( "r ) s,K,.

2. Location of max.

tension shifts to
wall face at all
levels.

ft

Reinforcement length,
L- 11.5/ 16, 20 ft.

case 19-21

r
case 10

1. Change of -40\ in

top reinf. level,

+50\ in next level,
and +30\ to -15\ in
bottom 6 levels.

2. Locations of max.
tension are 2-5 ft
closer to the wall
face.

J. Value 5\ greater

than Baseline case.

"o"

'11."

.t'

ft-

--.---,.-.,

1. Change of tIO\ at
all reinforcement
levels (or all three
lengths.
2. For La 16' and 20'
distribution similar
to Baseline case.
ror La 11.5' the
locations of max.
tension are 1-2 ft
closer to wall face.
3. Value within 5\ of

J. Value 25\ smaller

Baseline case for all

three lengths.

than Baseline case.

Wall Height

U" , -

40 ft.

w . '"

1. Values similar to

Baseline case (see

Figur 39 And

2S~1

Uniform surCharge, q.
raises max. tensions
by a bout qs,K
2. Location of max.
tension 0.25H-0.3H

behind wall face

except at extreme top
and bottom of wall.
3. Value 320\ greater
than Baseline case
Ox .... 11 height

qive. 4x reinf.
tension sum.)

D. Lateral Wall Face

Deforr.la tion-

Compare FEM Rein!.

tension to value
predicted by
conventional design
methods
1.

Max. tenslon for

individual levels

2.

Sum of max tension

for all levels

Max. deformation 250%

greater than
Daseline. Wall is on
tho verge of overall
stability failure.
Several reinforcement
levels on the verge
of pullout failure.

Location of max.
deformation shifted
downward to about 6'
below top of wall.
Max. deformation 100\
greater than
Baseline.

Deformed shape
similar to Baseline.
Max. deformntion for
~11.5' 25\ greater
than Baseline, for
~16' 10\ less, and
for ~20' 25\ less.

Similar to Baseline
but ratio of max.
deformation to wall
height greater by

Used Conv. Method A

Used Conv. Method A

Used Conv. Method A

Used Conv. Method A

1. FEM value for top

level is 330\ of

1. FEM values
95-150 \ of
conventional.

1. rEM values for

bottom 7 levels 90140\ of conventional I
for top lovel about
200\ of conventional.

1. rEH values for

upper' levels
135-170 \ of
conventional: for
bottom level 70\.

2. FEM values about

120\ of conventional

2. FEM value 150\ of

conventional. Value
of O.5.(~)'II"K. is
within 2\ of fEH sum.

conventional I for
next 5 levels 60-140\
and for lower 2
l.wels 15-25 \.
2. rEM value 40\ of

conventional.

2. FEM value 110\ of

conventional.

values.

10\.

The program employs different stress-strain models for each of the

three major element types.
Soil elements are modeled with a
hyperbolic stress-strain path for primary loading and a linear
path for unload/reload response.
In addition, a hysteretic
unload/reload model is used to model compaction stress effects.
Soil elements become essentially plastic upon failure.
The
structure elements are modeled as completely linear elastic
materials.
Interface elements are formulated with a linear
elastic stress-strain path in the direction normal to the physical
boundary being modeled, and with a hyperbolic path in the shear
direction.
The hyperbolic path is used during primary loading and
reloading, and a linear path is substituted during unloading.
Baseline Case - The structure chosen as the reference, or
baseline, case for this study is shown in figure 53a.
It is a
20 ft high bar mat (VSL) wall with a segmental concrete facing,
constructed with eight layers of reinforcement and compacted
gravelly sand backfill. Moderate compaction stresses, believed to
be representative of typical construction practice, were included.
This wall was selected as a baseline for the following three
primary reasons:
It is very similar to wall #3 of the "FHWA Field Walls,"
a wall that had been successfully modeled during an
earlier phase of the project.
It has a reinforcement system of medium stiffness, i.e.,
between that of the stiffest (welded wire) and most
flexible (geotextile) wall systems.
Its height, reinforcement length and construction
sequence are typical of a large number of actual
reinforced soil highway walls.
Some of the F.E.M. results for the baseline case are presented in
figure 53a and b and briefly summarized below. The results for
the baseline case are compared with the results for other analysis
cases in table 4.
The maximum reinforcement tensions versus depth predicted by the
F.E.M. analysis (figure 53b) generally fall near the K line shown
on the plot.
The tension in the bottom level, however~ is between
the Ko and Ka lines.
If the baseline case is analyzed without soil compaction stresses,
then the predicted maximum reinforcement tensions fall between the
K and K lines at almost all reinforcement levels.
This sort of
r~inforcement tension pattern is very similar to that assumed in
conventional design procedures for steel strip walls which use a
lateral stress coefficient that varies linearly between K and K
from the top of the wall to a depth of 20 ft.
It is possible that
the relatively slow time required for construction of field wall
may have influenced the results.
Of course, another possible
explanation for the discrepancy is that the F.E.M. model did not
accurately model the compaction conditions used for construction
of the baseline case.
101

8 Reinforcement Layer:s
Reinforcement Type :
VSl - 0.S'x2'
4 longitudinal Bor:s (Wl1)
Reinfarcement Spacing3
Sh - 4.92 ft.
s.,. - 2.5 ft.
S<!gmental Concrete Facing

Zone

of---------3=
c;f----------

8 Compacted Soil Layer.s

Backfill Uatenal : Cravelly Sand

-.130 pet
- 40 degree3
c "" 0.0 p:s1

:c 1----------

-.;
'"

:z:1----------

.,.---/.,-"'..,.,--/,-:..<.--.-------'- - - - - - - - - - - - - - - - - - - - - -

foundation Soil :
Uedium Oel1:!e Sand With Grovel Trace:!
(SP-S\.IJ)

Physical Conditions

Displacement at Wall Face

Baseline Case

3:
o

a..
10 ~

~.

o
CD

..c

15

a.
CD

1.0

Lateral Displacement
at Wall Facing

o.Q2o

Toto! Lateral Displacement (in)

Figure 53a.

Geometry and F.E.M. r~$ults for the

baseline case. ISS )
102

Di"tribution of Tension In Reinforcemenb

1.0

'I.

Loyer No.8

0.0
1.0

,.......
Layer No.7

0.0
2.0

~
0~

1.0

3:
Loyer No.6

0.0
2.0

co

1.0

0.0
2.0

.,E
.2.
c

.E
c

c()

>-

....0

';...

f-

Loyer No.5

;;:
0
<D

CD

1.0

Loyer No.4

0.0
2.0

....0..

'.\ ' ~

" ~'-"
\\

,\
'\

~\

\
12

,
\

'\

\'
'.\
\

'-

\
\

.r:

14

Loyer No. :3

.'\

",
\

"

18

'"

~"

1.0

Loyer No.2

"\ '~,\\\.,
\

16

0.0
2.0

lO

0.0
2.0

'

, '\"
\

<D

1.0

Conventional Design
No Compaction
Medium Camp. (Baseline)
Heavy Compaction

\'\ \
\

0..
0

0::

....

'-"

"
"1~\

'

\D"
/
.L-r..~'
o

',to \ \

/"

20 -rrrTT". ."rrrr~rrrrTT~"rrrrn
0.0
1.0
2.0
3.0

Maximum Tension in Reinforcement (kips/ft)

1.0

Loyer No.1

0.0

II

10

12

Maximum Reinforcement
Tensions

Oi3tonce from Wall fa<::e (ft)

Reinforcement Tension
Distribution

Figure 53b.

F.E.M. results continued for the baseline case.(SS)

103

The predicted tension distribution along the reinforcement levels

generally has an intermediate value at the wall face, a peak
behind the wall face, and a value of zero at the free end (figure
53b).
This is a characteristic pattern found by many
investigators.
The predicted pattern of wall face deformation (taking into
account the construction sequence) increases curvilinearly from
the wall base up to the top of the wall (figure 53a).
Conventional Design/Analysis - The conventional design methods
used for comparison in table 4 are taken from the sources listed
at the end of the table.
The conventional method "A," used for
comparison for all but the strip loading cases, features the
Meyerhof vertical stress distribution procedure, a bilinear
assumed failure plane, a lateral soil stress coefficient varying
linearly from K at the wall top to K at the wall base, and an
empirical pullott capacity based on VSL bar mat behavior. This
method does not take into account facing type, compaction
stresses, or foundation soil type and therefore gives the same
results for cases where these variables are altered.
The method
differs from the method proposed in volume I of this report.
Conventional Methods "B," "e" and "D" are used only in the strip
loading cases and are all superposition methods that add
incremental stresses due to the strip loading onto stresses
calculated by method "A."
Results and Implications of study - The predictions, by F.E.M. as
well as by conventional methods, of reinforcement tension and wall
face deformation for the cases studied are compared and summarized
in table 4.
Detailed results are included in the FHWA project
final report along with F.E.M. results for the magnitude and
distribution of vertical soil pressures.
If the F.E.M. results are assumed to be accurate then the results
in table 4 have at least three major implications for conventional
design practice.
The first is that the conventional design
methods may underpredict reinforcement tensions when there is
significant compaction of the backfill during construction. A
possible explanation for this is that the conventional methods may
have been calibrated against instrumented walls that were
meticulously constructed, and therefore may not represent a
typical highway wall.
The second implication is that in a few cases the predictions by
conventional and F.E.M. methods differ by magnitudes too great to
be explained by the compaction effects discussed above.
These
differences are often unconservative in nature, i.e., the
conventional analysis predicts lower tensions than the F.E.M. A
possible explanation for these observations is that the
conventional method may not be reasonable for non-standard wall
conditions.
Results indicating these unconservative conditions
are marked by stars in Table 4. One example is that the
prediction of reinforcement force increases due to strip loading
appear quite unconservative in the lower half of the wall (figure
54).
On the other hand, overconservative conventional design is
also undesirable.
104

Parametric Study -

Strip Loading Case

Maximum Ten~ion in Reinforcement (kips/It)

1.0

2.0

J.O

~.O

4.0

6.0

7.0

11.0

.. -.;.
2.

...

,/~

(;
~

"0

\
\

II

,f

,:

t:

II

f
i
,

0.

0
t-

::-..:>

~A

10

Gi

CD 12

..

,
,

.s:::

0.
.,\4

Bo~etin..

f[}.f R.~ult

o Lebo .. Kennedy method

a 1H:2V locd Oi3:pen.ion
... 8ou~:s~ Load Oi3penion

16

III

20

Bridge Seat

.a)

Parametric Study -

Strip Loading Case #2

Maximum Tension in Reinfa~cement (kips/ft)

1.0

--

....

\
2

2.0

,......
~

(;

"

\,..
\"
\

"0

II

,,
,

a.

1
\

I- 10

Gi

CD 12

.s::

1., \

,
"

\,

Ci.

"'14

~,

Cl

16

18

\.

8c,eGne

fl).A Result

~2V 't.!:;~!Od

.. Boo,s.. Lood o;sp."ion

,,

\
\

20~----------------------~

____ ______ __________

~

bl Crane Load

Figure 54.

Maximum reinforcement tensions for strip loaQin?

cases by F.E.M. and conventional methods.(85

105

The third implication is that wall face deformations, an important

performance criterion, may change significantly due to factors
that conventional design practice does not consider.
For example,
the baseline wall deformation can be decreased by 25 percent by
increasing the reinforcement length to 20 ft, or can be increased
by 60 percent using a variable reinforcement spacing.
4.0

INCLINATION OF THE THRUST AT THE BACK OF THE WALL

A large F.E.M. parametric study using a calibrated computer

program was performed from 1982 to 1984 by the Reinforced Earth
Company to specifically evaluate the inclination of the thrust at
the back of the wall.
More recently, the French Administration and the Reinforced Earth
Company conducted a large research program including full scale
experiments in order to have a better knowledge on the critical
value of the L/H ratio.
The results of both studies showed that the thrust at the back of
inextensibly reinforced soil walls is inclined downwards and that
the inclination angle X agrees fairly well with the empirical
relation:

X = (1.2 -

( 25)

L/H)~b

where L is the length of the reinforcements, H is the height of

the wall, and ~b the friction angle of the retained backfill.(86)
There are several reasons for the inclination of the thrust:
1)

The reinforced fill section is stiffer than the

nonreinforced backfill.

2)

The relative settlement of the backfill has been found to be

greater than the reinforced section creating downdrag at the
back of the stiffer reinforced section.

3)

The smaller the L/H ratio, the greater the tilting of the
reinforced soil wall and the relative settlement of the
retained fill.

However, these findings have not been validated for reinforced

soil walls with extensible reinforcements.
Furthermore, since the
stiffness of the reinforced section of walls constructed with
extensible reinforcement is not that dissimilar to the backfill,
the relative movement of the reinforced fill and the unreinforced
backfill are approximately the same.
Therefore, inclination of
the thrust is not assumed to occur and a value of X = 0 has been
conservatively recommended for walls constructed with extensible
reinforcement.
Figure 55 illustrates this new inclination concept for reinforced
fill walls.

106

I---L----I

ACTUAL
RESERVATION

a)

Inextensible reinforcement.

~----- L------~

I
H

~--------------__--~F=~----._Pa

Figure 55.

Thrust at the back and vertical stress distribution

at the base of reinforced soil walls.
107

5.0

DISTRIBUTION OF THE VERTICAL

cr v AT THE BASE
____
__ STRESS
_________

~~~~~~~~~~~~~

Since the early development of the Reinforced Earth technique, a

Meyerhof distribution has been used to analyze the vertical stress
G
at the base of a reinforced soil wall.
Field instrumentation,
as well as F.E.M. studies, have shown that this distribution is a
fairly good and simple approximation of the actual value of
maximum pressure (figure 55).
No difference has to be made
between walls with inextensible and extensible reinforcements.
6.0

INFLUENCE OF SYSTEM STIFFNESS ON THE HORIZONTAL STRESS IN THE

REINFORCED SOIL SECTION.

In reviewing the small scale model and finite element results, it

became apparent that the relative stiffness of the reinforced soil
system has a direct influence on the tension developed in the
reinforcement.
To further evaluate this relationship, the lateral
stress ratio K was back calculated from available field data and
compared to the active earth pressure coefficient K as shown in
figure 56.
K values were calculated from:
a
T max

where:

Gv

( 26 )

yZ

Tmi\X is the measured maximum tension in the

re1nforcement at depth z.
Sv' SH are the vertical and horizontal spacing of
reinforcements.
The data was then grouped according to the relative stiffness of
the reinforcement system Sr' as shown in figures 57, 58, and 59,
where
Sr

~~~ as defined in figure 26, of volume I.

( 27)

The K/K relationships shown by the solid line in each plot

correspond fairly well to the K values used in conventional
analysis of the specific reinforcement syst~ms represented by each
set of data.
The results show clearly that for most of the
reinforced soil structures for which data was available, in which
S is between 600 and 2,000 k/ft2, the K/K value is close to 2 at
the top of the wall and becomes 1 at depthaof 20 ft (figure 58).
The other set of field data (figure 59) clearly indicates that for
S r > 2,500 k/ft2, the K/K a value is close to 3 at the wall top and
1.5 at a depth of 20 ft.
The relationships developed from the analysis were subsequently
used to develop figure 26 1n volume I.
Analysis of the predicted
horizontal stress in a 20 ft high wall using figure 26, volume I

108

""l

f-'.

u:)
~

'"1
It)

KIKa

U1

0'1

8. 0

1.0

2.5

2.0

1.5

3.0

.3.5

Variation of K with depth

For all value of Sr

<:

,.,

01

f-'.

.,(1)

01
rt

I!l

A+

f-'.

0
~

<!

.......
0

to

;t
0

0.
'tj

rt

::r

A.

10

I!lftft ff
C!>

I!l

I!l

.c.

00

0.

A~

<1l

15

,.,

tl

t:l.

It)

.....
0.
~
01

.....
.....

(;)

...1

C!>
C!>

(fJ

H1

Gl
ftff

f-'.

C!>

v
ro

20

x
I A. ...

~I!l

e
I!l

I!l

o"'ocPAi!l

H1

It)

ftff

I!l

0..

(fJ

It)

I!l

"

ft

c:

....

,.,

I!l

:s:

It)

(fJ

-.....;

Backfill: Sand or Sandy Gravel

al!l

I!lI!l
,,-..

H1

"

DI

ff

A.

;:3

4.0

/tNt

(fl)

C!>
(!I!)C!>

UCLA. Wall
A.A.A.A.J. Dunsmuir
STS Wall 1
..... WES Wall
00000 Baxter
'" At. At. STS Wall 3
H+++ STS Wall 4
""""" CalTran
H H f f Hayward Wall
1!l1!l(!)01!l Seattle VNJ
tlt:lt:ltltl STS Wall 2
00000 STS Wall 7

20
16
20
12

16
20
20
20
20

47
20
21

Sf

System

2160
1390
1500

RE

(k/n2)

622

1858
1037
1037
997
2342
3010
56
20

RE
RE
RE

MSE
VSL
VSL
MSE
VSL
Welded Wire
Geo~rid

Gablon

'"'1
1-1'

..0

c:

t1
rtl

U1

-...J

8. 0

0.5

1.0

KIKa
1.5 2.0

3.0

3.5

4.0
Variation of K with d~pth
20
100 k/ft
Sr

<:

OJ
t1

1-1'

OJ

rt

1-1'

Il.

:='
r;-

0
Hl

Baokfill: Sand or Sandy Gravel

5f-

....
~

"-'
~

"!Io

<:

I-'
I-'

rtl
t1

en
en

c:

0
;;:

....0

0.
0

I-

10

10

0-

0.

ro

'0
rt

:50.

rtl

:::r
Hl

(ft)
Il.

Q)

Cl

15

1l.t!.1l1l1l

0
Il

.,(/J
0
N

LIl

I
0
0
:><;"

"Hl
rt

IV

20

STS Wall 2
7

00000 STS Wall

t1

I-'

System

Sr

(k/ft 2)

20

56

21

20

- - Proposed FHWA Line

Geo~rid

Goblon

'"'l
f-'.

t.O

..,

K/Ka

ro

g.O

(J1

0:>

2.0

1.5

1.0

0.5

<:

..,

PI
1-"

PI

...

C/l

rt
f-'.

::l
0

(J1 Hl
0
0::>::

ro

I-

;0;-..,
"Ul
t-t,c

rtUl

..,

10

4.0

Variation of 1< with depth

Sr :::: 500 - 2000 k/ft

I~

Backfill: Sand or Sandy Gravel

0
j

'0

t-t,

3.5

0-

en

!J

J4I

3.0

.c
-<J

rt

3:

'" , .
.0
ro

'0

.....

CI"
0
0<

"-'"

I"

N::<:

.-.
.-.
.-.

,-...
..J
.....

2.5

0.
Q)

15

00

20

At

)(

IA+

(it)
UCLA Wal!
..........u. Dunsmuir
I I I I STS Wall
1
WES Wall
00000 Baxter
A Ab. A A STS Wa II 3
+++++ STS Wall 4
)( ~ )( ~ ~ CalTran

1--

20
16
20
12
16
20
20
20

Sr

System

2160
1390
1500
622
1858
1037
1037
997

RE
RE
RE
RE
MSE

(k/ ft2)

ProposedFHWA K/Ka-Z Line

VSL
VSL

MSE

I'I:j

.....
I.Q

t::

1"1

C1l
U'1

\0

K/Ka

8. 0

c::::

01

r-t

..,

.....

..... 00

iB

<:

H1 p,

'1

:;.:: .....

"" ~

:;.:: ~

III

;;l

<: ~
I-'
I-'

tv

:>'1

I\)

1"1

-..

U'I

p,'r.

t:: <

U'I

(I)

0.(1.
C1l

'O~

rt" :T
;:::r...,
0
H1 '1

C/)

1"1

'1

...C/l

..-..
..;

""

H1

rt"

'"

3.0

3.5

4.0

Variation of K with dep~h

Sr = 500 - 2000 k/ft
Backfill: Sand or Sandy Grovel

'+-

'-"

'0
3=
'+-

0.
0

f-

u7

lO

Qi

(ft)
I I I I I

J.

ro
.c
.....

a.

of.

(l)

A A.& A

.....

00

00000

"'" h.i.. '"

++ ++ +

15

)()()()()(

'"
Vl
0
0

UCLA. Wall
Dunsmuir
STS Wall
WES Wall
Baxter
STS Wall 3
STS Wall 4
ColTron

20
16
20
12
16
20
20
20

Sr

System

2160
1390
1500
622
1858
1037
1037
997

RE
RE
RE
RE
MSE

(k/ft 2)

- - Proposed FHWA K/Ko-Z Line

...
"

tv""

;;.;-

2.5

v -..
""
U'1
0
0

2.0

1.5

./

o .

;:l

1.0

".....

01

0.5

20

)(

.01.+
I.

VSL
VSL

MSE

is compared to the conventional Reinforced Earth method and the

New French Recommendations in figure 60.
The K/K - Z
relationship corresponding to S = 500-2000 k/ft 2A was used.
Similar comparison for a high wall (40 ft) is shown in figure 61.
From these plots, it appears that the horizontal stress is
comparable for the FHWA Manual method and the New French
Recommendations.
The FHWA Manual method may give slightly more
conservative values near the wall top, but gives a less
conservative value near the bottom of the wall.
Notice, however,
that the field data in figure 58 show measured K/K data in the
upper wall levels exceed the proposed values in several cases and
at depths greater than 15 ft, all data is less than the proposed
value.
Thus, the proposed FHWA Manual method is still on the safe
side.
7.0

INFLUENCE OF THE LENGTH OF THE REINFORCEMENTS

Initially, for Reinforced Earth walls, the ratio of the

reinforcement length to the wall height L/H was taken equal to 1
for preliminary design. At the end of the 70's, based on account
the results of about 15 instrumented Reinforced Earth walls and a
number of model tests, the recommended L/H ratio value was reduced
to 0.7 (French Specifications on Reinforced Earth, 1979).
In
volume I, an L/H ratio value of 0.5 is considered for preliminary
evaluation of reinforced fill walls.
The L/H value of 0.5 results from the model tests, parametric
F.E.M. studies and full scale experiments discussed in section 2.
As the similitude is not respected in classical reduced scale
models, centrifuge model results were given more consideration
than the reduced scale model results.
The F.E.M. results in section 2.c., Table 4, corresponding to the
extreme case of L/H equal to 0.275 clearly shows the influence of
a reduced L/H on the wall system.
The main conclusions concerning
the behavior are:
F.E.M. calculated tensions are less than the ones
calculated using the design method proposed in this
manual, indicating that the tension analysis method is
conservative for a decreased L/H ratio.
The horizontal displacements were found to be 250
percent greater than with a L/H ratio equal to 0.7
(Baseline) and the wall appeared to be at the verge of
the overall stability failure.
At an L/H ratio of
0.425, the wall was found to be stable, but had a
corresponding horizontal displacement of 100 percent
greater than the baseline case.
These conclusions indicate the importance of evaluating lateral
deformation response for reduced L/H ratios.

113

'"'l

f-'.
\.Q

Typical Reinforced Earth Wall, H=20 ft.

t1
C1)

0'1

- - OrigInal ReInforced Earth Method: Meyerhof

New French Recommendation: Meyerhof +
- - FHWA Manual Method: ')'Z + K.. K,Sr)

I'r"" 129 pcf, cjJr=39

I'b",,129 pd, b=39

1-3

"<

Horizontal Stress ah (psI)

Horizontai" Stress ah (psf)

'0

f-'.

oo

()

200

400

600

800

1000

r ,~~-.--r-~~-r-T-'~

OJ

+ KC2K o to K.

" +

K""Ko to K,

Horizontal Stress ah (pst)

o "-.--r-.--r-.---r-.--.-.--

200
400
600
800 1000
o o.-,.'-r--r-,.---,.---,----,--,--.,.-.....-

200

400

600

800

1000

I-'

!t!

C1)

f-'.

:::l
t-h

o,\
\

()

('[)

0.
f--'
f--'
.t

t1

-:;-

('t

4)

'-'

~ :5

10

10

\
\

\~
12

10
I

\'

12

12

14

14

16

16

\
\

....

'"

0.

I-' 4)
I-'Q

\
\

::TJ!
OJ

\
\

tTl
OJ

o\
\

t1

14

\\

16

I \I
I \

18

L/H

:::I

0.5

18

\
\

20

..

(a)

,,

\\
\\

L/H

18
:::I

., ,

20
(b)

.I

L/H = 1.0

20 ' - - - - - - - - ' - - - - - - . . !
(c)

"l'J
~.

to

High Wall H=40 ft.

11
C1>

0"1
f-'

- - Original ReInforced Earth Method: Meyerhof + K=Ko to K.

.--~.-- New French Recommendation: Meyerhof + A + X .. Ko to K.
- - FHWA. Manual Method: yZ + K",KtSr)

/'r=129 pcf, tP,=39

/'b= 129 pef, tP b=39

::r:

Horizontal Stress 0\ (psi)

Horizontal Stress 0\ (psf)

o
11
,....
N
o

400

800

400

1600 2000

1200

800

Horizontal Stress CF h (psI)

1200 1600 2000

400

800

1200 1600 2000

::l

("t

OJ

12

12

12

16

16

16

20

20

20

24

24

28

28

32

.32

I-'

en
("t

11
C1>

en
en

f-'
f-'
U1

'Z'

'0
OJ

....S

11

'-'

en

:5Q.

::l

Cl

,....
o

OJ

24

\1'I'

28

\\
\ '\
I
\ ..I
\ '\

.32

H\

11

::r
~.

to

::r

\
\

OJ
I-'
I-'

.36
40 (

L/H

....

= 0.5 \ \,
\

ob
IV

..

(a)

.36

\ \

L/H

36

40

"
(b)

L/H =

40

\\
(c)

Full scale experiments performed recently by the French

Administration and by the Reinforced Earth Company have found that
carefully constructed reinforced soil walls with inextensible
reinforcement and a L/H value equal to O.S behave
satisfactorily.(86)
Until the L/H ratio was reduced to less than
O.S, the location and shape of the maximum tensile force line was
not affected.
8.0

LATERAL DEFORMATION AT WALL FACE

As indicated in chapter 3 of volume It the current design method

for evaluating the lateral displacements at wall face is handled
empirically based on the results of the F.E.M. study (section
2.c.) and the results of the eight instrumented full-scale walls
(section 2.b.).
The L/H ratio of the field walls was 0.7, so that
tilting and sliding movements were very small. As indicated in
the previous section and as illustrated in figure 62, the F.E.M.
results showed that the L/H ratio has a significant influence on
the lateral deformation at the wall face.
Measurements of the lateral displacements on wall 2 of the FHWA
study as compared to the other walls found that the rigidity of
the reinforcement connections at the face has a large influence on
tilting deformations.
It should be realized that all lateral movement due to design
stress conditions should take place during construction.
Postconstruction movement should only occur due to settlement of
the reinforced soil section.
Therefore, lateral deformation is
mainly a construction problem that can be handled by proper
batter.
Excessive postconstruction movement may be an indication
of foundation problems or an overstressing of the reinforcement
due to an inadequate design.
9.0

RECENT RESULTS ON SEISMIC BEHAVIOR OF REINFORCED SOIL WALLS

The basis for seismic design in volume I is an internal report to

the reinforced earth company by seed and mitchell in 1981 as
updated, based on an extensive f.e.m. study and half-scale shaking
table model tests performed by the Reinforced Earth
Company. ( 19, 87)
the results of recent research has found that the maximum tensile
force line in a reinforced soil wall under seismic loading is
essentially the same as under static loading, even for strong
accelerations (~ ~ 0.4).
in addition, the lower strips are the
most affected (i.e., required to withstand the greatest dynamic
increment) under a seismic loading.
the use of a pseudo-static
dynamic thrust Pae was proposed by seed and whitman.(21)
the design principle considers that the total tensile forces in
the reinforcements results from the vertical overstresses due to
the seismic thrust p
and from the inertia force p. acting on the
active zone.
the fo~ces are then distributed among 1 the different
layers of strips based on the area of the resistant zone and the
distribution of reinforcement in the resistant zone.
116

""":'8.0
c

VSl Wall with Gravelly Sand

H-20 it

<ll7.0

()

u...
_6.0
o
:3
...... 5.0
a

4.0

:z

a
E3.0
La

<ll

02.0

E
.::! 1.0

'x

00.0

Figure 62.

10.
20
()
Length of Reinforcement ft

30

Influence of change in reinforcement length on lateral

deformation at the face of the wall anticipated
during construction.
117

Results of finite element calculations have shown that this simple

design method represents very well the total tensile forces
developed in the strips under a seismic loading.
The safe
behavior of numerous reinforced fill walls designed using this
approach which have been subjected to seismic loading,
particularly in Japan, is additional proof of the validity of this
method.

118

CHAPTER 4
REINFORCED ENGINEERED SLOPES
1.0

INTRODUCTION

The methods for internal stability analysis at working stresses

described in chapter 3 of volume I have been developed for
vertically and quasi-vertically faced walls.
They are empirical
methods and they cannot be easily generalized to sloped
structures.
Consequently, the internal stability of reinforced
engineered slopes is better analyzed using the limit equilibrium
analysis.
Although several limit equilibrium approaches have been formulated
(see reference 25 for summary of approaches) and successfully
utilized to design numerous reinforced engineered slopes, very
little research has been performed to substantiate these methods.
A parametric study using a computer program was performed by
Jewell and used to develop design charts for the Tensar
Corporation (as modified b Schmertmann, et all in figure 43,
chapter 4, volume 1.(26, 6 )
Other substantive numerical,
laboratory, or field research does not appear to be available.
In
a study of polymeric reinforced soil structures constructed in
North America, no instrumented field projects were identified.(48)
Therefore, an important aspect of the FHWA Behavior of Reinforced
Soil Study was to construct and instrument field structures to
verify the limit iquilibrium design approach.
For the FHWA project, two 25 ft high, 1 horizontal to 1 vertical
slopes and two 25 ft high 0.5 horizontal to 1 vertical reinforced
soil slopes were constructed as summarized in Table 5.
Table 5.

Full scale reinforced slope experiments constructed for

the FHWA program.

Slope

Fill Material

Reinforcements

0.5 H:1V

Polypropylene
Geogrid (T u1t

2.

0.5 H:1V

Polypropylene
Woven Geotextile
(T u1t = 214 Ib/in)

Silt (4)

3.

1H:1V

polypropylene
Geogrid (T u1t = 145 Ib/in)

Silt (4)

1H:1V

Polypropylene
Woven Geotextile
(T u1t = 214 Ib/in)

1.

4.

= 145

Ib/in)

Silt (4)

All slopes were designed using the rotational failure limit

equilibrium analysis approach in chapter 4, volume I and checked
using Jewell's design charts.
For all structures, eight layers of
reinforcement with a uniform spacing of 2.5 feet (0.76 m) and a
119

total length of 14 ft (4.3 m) were

tension T of 60 lb/in (10.5 kN/m)
this arrangement provided a design
1.0 for slopes 1 and 2 and a FS of

used.
Using an allowable
assumed for both materials,
factor of safety FS approaching
1.3 for slopes 3 and 4.

The complete results of the field program are contained in a

separate report to the FHWA.
Figures 63 through 66 show the
tension measured in each reinforcing layer as compared to the
computed maximum tension from the limit equilibrium method
(chapter 4, volume I) and from finite element analysis. As can be
seen from the figures, the limit equilibrium methods are somewhat
conservative for estimating the maximum tension.
Figure 67 shows a representative plot of the distribution of
tension measured in the reinforcement as compared to that
predicted by the finite element method. Also shown is the
predicted location of maximum tension in the reinforcement from
the limit equilibrium method which appears to be in close
agreement with the measured values.
Figure 68 provides an example of the measured and predicted
deformation response.
For all slopes, the measured deformation
was much less than the estimated deformation.
Most likely, this
was due to the inherent additional factor of safety for the
material tension (i.e. Ta versus T u1t ) used for design.
As a result of the field study, a rotational limit equilibrium
analysis method is proposed for the design of reinforced
engineered slopes.

120

Embankment 1
Slope = .5 H/1 V
Geogrid (EA=90 kips/ft)
Embankment Height

20 ft.

Measured (12/4/87)
finite Element (c=50 psf, 'fl=35 deg.)
----- ~imit EquiL
(based on ~=35 deg.)

o
A

20~-------------------------------.
\
\
\
A\
\
\
\

15

l!J\

\
\
.t. \

\
\
\
\
\

\
\
\
\
\
\

\
\
\

\
\

at.

\
\
\

0.00

Figure 63.

0.25
0.50
0.75
1.00
1.25
Maximum Tension (kips/ft)

Distribution of maximum tension with

depth in embankment 1.
121

1.50

Embankment 2
Slope = .5 H/1 V
Geotextile (EA = 96 kips/H)
Embankment Height

20 ft.

Measured
Finite Element (c=50 psf, !j=.35 deg.)
- Lir.lH EQui 1.
(based on j=.35 deg.)

20-r--~----------------------------~

0\

15

\
o

'\

\
h\
\
o

\
\
\
\

0.00

0.25

0.50

1.00

Maximum Tension (kips/ft)

Figure 64.

Distribution of maximum tension with

depth in embankment 2.
122

1.25

1.50

Embankment 3
Slope =
1 H/1 V
Geogrid (EA = 90 kips/ft.)
Embankment Height

25. ft.

Measured (maximum of recovered data)

Finite Element (c=50 psf, ~=35 deg.)
Limit Equi 1.
(based on ~=35 deg.)
25~------------------~------------~

\
\
~

\
\

20

\ .t::.

\
\
~

.t::.

15

.t::.

\
\
\

.t::.

\
\
\0
\
\
'h
\

A \

\
\
\

\
O-+~~~'~I~~~'I~'~j-j~,~~-r~rT~~"j~l~j~ji7IJj~1

0.00

Figure 65.

0.25
0.50
0.75
1.00
1.25
Maximum Tension (kips/ft)

Distribution of maximum tension with

depth in embankment 3.
123

1.50

Embankment 4
Slope = 1 H/1 V
Geotextile (EA = 24 kips/ft)

= 25. ft.

Embankment Height
o

Measured (maximum of recovered data)

Finite Element (c=50 psf, q,=35 deg.)
Limit Equi 1.
(based on =35 deg.)

Ll.

25
\

"
\

20

\
\
~

15

.,.--.,.

......

' ---'

A\

..c

01
Q)

:r: 10

A \

\
A

Ll.

\
\
\
\

O-+~~~~~~~~~~~~~~~~~~~~
I

0.00

Figure 66.

I'

I '

0.25
0.50
0 ..75
1.00
1.25
Maximum" Tension (kips/ft)

Distribution of maximum tension with

depth in embankment 4.
124

1.50

Embankment
Slope = .5 HI
Geogrid

1
V

o Measured (12/4/87)
Computed

r---------------------

25

I
/

Embankment Height

20. it.

20

"'-

r--

0.<

Range of Tmax from Limit

Equilibrium Analysis

~00

if o.J
......
III

0.0

~O

--y.-_ _ _--L_ _~_ _ __

.<J

15 ..c
en

..s- 0.::P
Q:

.<J

'+........,

Q)

:r:

-y----+----'------

-y--~-~--

__- - - - - - -

10

t::: 00

J.!!! 0'1
0. 0

----;-----,<--7"-------

0. 0
0.0

Figure 67.

12
18
24
30
Distance From Embankment Toe (ft.)

36

Distribution of tension along different levels

of reinforcement in embankment 1.
125

Embankment 1
Geogrid and Clayey Silt
Deformation at 12.5 ft. From Toe
Height of Fill = 16.25 ft.
Gl
Field Data
o Finite Element

20

12.5

18

16

- -- - -- - - - - - - - ---1

,
I

14

I
I
I
I
I
I!J

-- - - - - - - - - -- - - - -1
12 .

,,
,,
\

~,

-1-0--- ----- - - - ---1

\
I
I

I
2.0

1.0

Deformation (in.)

Figure 68.

Deformation at 12.5 feet from the toe of embankment 1.

126

CHAPTER 5
SOIL NAILING

1.0

INTRODUCTION

Several approaches have been developed to estimate the resisting

forces required to restrain ground displacements in cut slope
retaining structures such as bracing supports, tiedback walls or
soil nailed systems.
They can be broadly classified into three
main categories:

2.0

(1)

Empirical design earth pressure diagrams.

(2)

Finite element analyses.

(3)

Kinematical limit analysis method.

EMPIRICAL DESIGN EARTH PRESSURE DIAGRAMS

Selection of an appropriate earth pressure diagram for the

determination of nail forces should be consistent with the
anticipated level of the structure and ground movements.
Measurements of facing displacements in nailed soil cut slopes
illustrate (figure 69) that in non plastic soils these
displacements are comparable to those measured in braced
excavations.
Therefore, design diagrams proposed by Terzaghi and
Peck and Tschebotarioff for the design of braced excavations,
provide a rational estimate of working tensile forces generated in
the nails. ( 39, 44, 88, 89 J
These diagrams are schematically
illustrated in figure 70.
Note that Terzaghi and Peck's design
diagram for sands has been slightly modified in order to calculate
nail forces.
The maximum tension force mobilized in the nail is
expressed as a normalized, non-dimensional parameter:
T max

. Sv

at the relative depth of z/H

( 28 )

TN is the working tensile force generated in the

reinforcement.
Tmax is the maximum tension force mobilized in the nail.
where:

H is the total structure height (or excavation

depth)
Sh and S
are, respectively, the horizontal and
verticalVspacings between the nails

For sands (c/yH < 0.05, where c is an apparent soil cohesion):

( 29)

where the active lateral earth pressure coefficient

K

( 30 )

127

-~

LL t-

t-

l1-

Z...J
W...J

::::!c::r
W3:

<.0

<IW

ci

..J:x:
o..t--

en
-

lJ...

00
-10..
<IO

t--t-Z

Ow
N:x:
OCt--

0
I

10

20
30
40
WALL HEIGHT (FT)

SOIL

NAIL

REFERENCE

l1edium sand

driven

38

Silty sand (SM)

grouted

II

Fine sand (SP) to

clayey sand (SC)

driven

37

Residual clayey silt

weathered shale. sandstone

grouted

39

Fontainbleau Sand (SP)

grouted

40

.,
0

Figure 69.

50

Horizontal displacement of nailed soil walls.

128

Nons:
-

Vertical cut slope

yh

- Horizontal upper surface.

lateral earth pressure

= overburden pressure

O.2SH

O.5H

TERZAGHl
I 8 PECK (1967) O.25H

.i.
TN= 0.65 Ka

SAND .

CLAYEY SA.'W

.-S-'
yH ~/ 0.05

= tan2

I..

.!

(nIl, -

Figure 70.

~/2

CLAY

TN = Ka ( 1 -

T = 0.2 yH

4c

0.4 yH

Empirical earth pressure design diagram.

129

.-L

).0.65 K

yH.fKa'

'"

and for a cohesive soil with both cohesion (c) and friction angle:

K [1 - ~
J < 0 . 65 Ka
yH (1-)0.5
K
a

( 31 )

Figure 71 shows nail forces and structure displacements measured

in four instrumented soil nailed structure. (1,37,39,40)
The
measured nail forces, specifically in grouted nails, were found to
agree fairly well with the assumed earth pressure design diagrams.
These results illustrate that the observed behavior of nailed cut
slopes is similar to that of braced excavations.
The use of the empirical earth pressure diagrams in the design of
soil nailed retaining structures presents some severe limitations.
In particular, these diagrams correspond to conventional cases of
bracing supports with simple geometry of a vertical wall,
horizontal ground surface and lateral braces.
Therefore, they
cannot be used to assess the effect of design parameters such as
inclination of the facing, inclination and rigidity of the
inclusions, surcharge, etc. on the working forces in the
inclusions and structure displacements.
They do not provide any
data with regard to the shear forces and bending moments that can
develop in the nails.
In addition, as shown in figure 72, in
cohesive soils the empirical earth pressure diagram is highly
sensitive to small variations in soil properties.

3.0

FINITE ELEMENT ANALYSES

The finite element method has been used by several investigators

to analyze the behavior of soil nailed retaining
structures.(ll, 90, 91)
These analyses involve different
constitutive equations for the soil and interface elements to
simulate soil-facing and soil-inclusion interaction. Attempts
have been made to compare finite element predictions with observed
behavior of instrumented structures.(ll,
1)
However, the use of
finite element method in design is currently limited by the
relatively high costs and raises significant difficulties with
regard to:
The actual construction stages and installation process of
the inclusions are difficult, if not practically impossible,
to simulate.
The complex soil-inclusion and soil-wall interaction is
difficult to model.
Several interface models have been
developed but their implementation in design requires
relevant interface properties which are difficult to properly
determine.(92, 93)

130

Tmo ..

HORIZONTAL DISPLACEMENT

{mml

10

0.01

0.02

0.03

0.04

r-------r-----~----~~--~--_.----_.----r_--------

0.25

yH =0.114

0.50

0.75

o
-INCLINOMETER AT
1.5 m FROM THE
FACING

q, =33

INCLINOMETER AT 4.5m
fROM THE fAC I NG

EMPIRICAL DESIGN ME:THOD

1.00
-

END Of CONSTRUCTION
EXPERIMENTAL RESULTS
H = 9.2m

Z/H

a.

DAVIS \lALL ( SheD et 81.. 1981 )

HORIZONTAL DISPLACEMENT
(mm)

Tmox
yHS ti Sv

0r-____,20
______~I0r-----,0~--0-.Or5----0~rI0---0-TI5----0-r20--------__~
END OF CONSTRUCTION
EXPERIMENTAL RESULTS
o H = 3m

0.25

0.50

= 7m

5m

KINEMATICAL
APPROACH

(.;, = 35, c = 0)

0.75
- FACING
DISPLACEMENT

MPI RICAL DESIGN

M::THOD

(H " 7m)

1.00

fACING DISPLACEMENT
(H = 3m)
Z/H

CEBTP FUll SCALE EXPER1J1D."'T ( Plumelle, 1986 )

Figure 71.

Experimental data and theoretical

predictions of tension forces.
131

HOR IZONTAL DISPLACEMENT

(mml

0.05

0.1

0.15

0.20

r------r_.--,--.--r_~c-----r_----_r------~----_.------

o1

0.25

I
T'7'7':A-~-1--

b
\

KIN E M AT I CAL
APPROACH

0.50

~
INCLINOMETER AT
:
0.75
1.5m FROM THE TOP
OF THE WALL

o INCLINOMETER AT
3.9m FROM THE
TOP OF THE WALL

1.00

END OF CONSTRUCTION
EXPERIMENTAL RESULTS
H"' 12..0m
H = 9.0m

Z/H

c.

CUMBERLM"D CAP \lALL ( FH\IA, 1985 )

HORIZONTAL DISPLACEMENT
(mm)

0.1

0.4

'\
\

0.5

DESIGN METHOD

o
\

0.25

Q
\

0.50

~"30"

<{

c=O

0.75 INCLINOMETER AT
3M FROM THE
FACING
o INCLINOMETER AT
1.00
7M FROM THE
FACING

END OF CONSTRUCTION
EXPERIMENTAL RESULTS
I<c:" 0.65 K"

H "' 5.6m

Kc = 0.8 K"

Z/H

d.

Figure 71.

PARISIAN \lALL ( u.rtier and Gigan, 1983 )

Experimental data and theoretical predictions

of tension forces (continued).

132

"4.7

12

Bending stiffness of reinforcement

........
E

u.

0.0
e!:

.,...c:

10

()

<:l

,oj

<:l

...,
CJ

....
c:

10

p 0

100

3.9

I
I
I

I
I

inclinatiCJtl 30e

<f)

CD
J:::

I
I
I
I

(.')

c:

3.1

u
r.:

CD

....;

,...

CD

(.)

.,...
I/.

"0

......

2.4

...,c::

a.
<f)

:a
c;
.....

c
0

to;:

..-!

.....
c

<..>

inclination 0'":

0I

c:

0
I

.!:::!
t-

1.6

:r:

I
I
I
I

I
I

I
I

I
I
I

Figure 72.

NOTE: FOR

0.8

I
I
I

I
I
I

CONVENTIONAL
Eolo= 2x107 1b' in 2

0
NAILS

Effect of the bending stiffness and the inclination

of reinforcement on the facing displacements.(91)

133

TENSILE

10

FDRCE IN KN

20

~o

30

__

~--

__

~_,~~4-

~
i i
0--0

-.

"""~
t~

"

f3:=

~o~o~~7

Rigidity

10Eo I 0

30 0

I!

~---r--~~--~;~'~-+----+----+--~13

W.

II

JW

z
::r:

Rigidity E I

____-+__________________

,.~:iI I!Clination~

c:t:

7D

~~~~~\.~~1~~:~~.~~111~~0

---

(j)

6~

58

~1\t,. . . ",',.j"\t-"'-" '. II 1I. -,~j~t_~~ ~!'~"~__~_~__'-_"~"~'______~:____

All --~- II .....
.. I ""

~____~,______

Q..

\,

J
1

Jqf 20

""

I ". ,1-r'cI

I"
1
,
1

tL

w
o

<

6_

"C:+-

"::5

i"_~1,

z
I

j'---

LL

__

'.;

I-

-",

I-

B ~----~----~----~----~----------~~--J26
o
2
4
7
9
II
13
TENS I LE

- 0,1

FORCE

IN

KIPS

OA

r---~-----+----~----~

Tc

- i

-c

TmQt

.I

0,2

I
\

0.1

Tmo.x
I.H \,Sv

OA

nail

I~
~l

O.

OA

0,6

J:t-

0.8

shear force

1.0

Z~

Figure 73,

Tmax tensile force

f3

Of

1.0

ZIH

Effect of bending stiffness of the inclusions

on nail forces.

134

Various elasto-plastic soil models can presently be used to

predict soil behavior during excavation. However,
determination of soil model parameters generally requires
specific and rather elaborated testing procedures limiting
the practical use of these models.
The finite element method has therefore been used mainly as a
research tool to evaluate the effect of the main design parameters
on the behavior of the structure, ground movement, and working
forces in the inclusions.
Figure 72 shows the results of a parametric study using the finite
element method to evaluate the effect of bending stiffness and
nail inclination on facing displacement in vertical nailed cut
slopes.(40, 91)
These results illustrate that, for nail
inclinations used in practice (10 to 15 deg.), the greater the
nail bending stiffness is, the smaller is the facing displacement.
As shown is figure 73a, for inclined nails an increase in the
bending stiffness results in a decrease of the maximum tensile
forces.
The behavior of inclined nails is substantially different
from that of horizontally placed nails.
During construction, inclined nails tend to undergo a local
deformation approaching the horizontal direction of maxi~um
extension strain in the soil.
This local deformation which is
controlled by the bending stiffness of the nails, results in an
increase of the structure/facing-displacements.
For horizontal
nails, as illustrated by both reduced scale model tests and
numerical test simulations (figures 72 and 73), the bending
stiffness has practically no effect on the mobilized nail forces
and structure displacements. Although the finite element results
are rather qualitative, they provide a significant insight into
the fundamental understanding of the system behavior and relevant
input into the selection of the main design parameters.
4.0

KINEMATICAL LIMIT ANALYSIS DESIGN METHOD

This limit analysis approach was developed for the design of

nailed soil retaining structures.(41, 4 )
It permits an
evaluation of the effect of the main design parameters (i.e.,
structure geometry, inclination, spacing, and bending stiffness of
nails) on the tension and shear forces generated in the nails
during construction.
The main design assumptions, shown in figure
74, as reviewed in volume I, chapter 6 are:
Failure occurs by a quasi-rigid body rotation of the
active zone which is limited by either a circular or a
log-spiral failure surface.
The locus of the maximum tension and shear forces at
failure coincides with the failure surface developed in
the soil.

135

The shearing resistance of the soil, defined by

Coulomb's failure criterion, is entirely mobilized along
the sliding surface.
The shearing resistance of stiff inclusions is mobilized
in the direction of the sliding surface in the soil.
The horizontal components of the interslice forces Eh
(figure 74) are equal.
The effect of a slope (or horizontal surcharge F h ) , at
the upper surface of the nailed soil mass, on the
tension forces in the inclusions in linearly decreasing
along the failure surface.
The effect of the bending stiffness is analyzed using a
conventional "p - y" analysis procedure, assimilating the
relatively flexible nail to a laterally loaded infinitely long
pile.
The solution involves a non dimensional bending stiffness
parameter, defined as:
Kh
N

where:

L0

D
Sh

.
.

L0

( 32)

Sv

[.LQ]1/4
Kh

is the transfer length which characterizes the

relative stiffness of the inclusion to the soil;
note that the length of the inclusion L is
substantially greater than three times the transfer
length L and it can therefore be considered as
infinitely long,
D is diameter of the nail,

E and I are the elastic modulus and the moment of

intertia of the nail, respectively,
Kh

is the modulus of lateral soil reaction.

As provided in volume I, the charts shown in figure 75 can be used

to obtain Kh values as a function of soil shear strength
parameters.

136

j4-~'"

"'-

to< (a ). f

Loodin9 Effec1
on Noil Forces

MECHANICS OF fAILURE AND DESIGN ASSUMPTIONS

T ... oc - O"tf.As; TcsTf(.As

As: SECilON ARtA OF HAIL

b) STATE Of STRESS IN THE INCLUSION

f( :.. Jo(. ~ x

\-...........>......

d~
I

c)

Figure 74.

to( -

0.32 Ttl tl

0 ...........

THEORETICAL SOLUTION FOR INFINETELY

LONG BAR ADOPTED FOR DESIGN PURPOSES

Kinematical limit analysis approach.

137

30

20

ftH'f=O
10

Note:

t/m3 = ~etric ton/~

metric ton = 1000 kg
I Ib = 0.454 kg

5
Figure 75.

C -tIm (cohe!>ion)

Horizontal subgrade reaction as a

function of the soil shear parameters.

138

A failure surface which verifies all the equilibrium conditions of

the active zone can be defined.
In order to establish the
geometry of this failure surface it is necessary to determine its
inclination a with respect to the upper ground surface.
Observations on both full scale structures and laboratory model
walls show that for the relatively flexible nails the failure
surface is oractically vertical at the upper part of the structure
(a

4.<U'

36)

The normal soil stress along this failure surface is calculated

using Kotter's equation. The maximum tension force (T
) in each
inclusion is calculated from the horizontal force equiilbrium of,
the slice comprising the inclusion.
Following the mobilization of
shearing stress assumption, analysis of the state of stress in the
inclusion yields the ratio of the mobilized shear T to tension
T
forces as a function of the inclination of theCinclusion with
respect to the failure surface.
Figure 71 shows a comparison between predicted and measured values
of maximum tension forces in soil nailed retaining structures.
It
illustrates that the kinematical design approach provides a
reasonable estimate of tension forces mobilized in the inclusions.
Specifically, the results of the full scale experiment conducted
in France on a 7m deep granular soil nailed wall (field data
reported by reference 40) which are reported in figure 71c for
several excavation depths, illustrate that the total excavation
depth has only a negligible effect on both the normalized tension
forces in the nails and the geometry of the active zone.
Therefore, at any relative depth (Z/H) the maximum nail tension
forces are approximately proportional to the total excavation
depth.
The predicted distribution of the maximum tension forces
agrees fairly well with the earth pressure design diagrams
proposed for braced excavations.
5.0

STABILITY ANALYSIS OF SOIL NAILED RETAINING STRUCTURES

The design of soil nailed retaining structures should verify:

The local stability at the level of each inclusion.
The global stability of the structure and the
surrounding ground with respect to a rotational or
translational failure along potential sliding surfaces.
a.

Local stability analysis

At the level of each inclusion the design should satisfy the

following internal failure criteria:

139

Pullout failure of the inclusion:

T max
IT

(33)

1 a <

Tmax is the maximum tensile force in the nail,

where:

1 is the adherence length, l It is the maximum

stear resistance at the soil-nall interface, and
is the safety factor with respect to pullout.

FI

This design criteria implies that for a soil nailed cut slope, the
structure geometry defined by the L/H ratio (where L is the total
inclusion length) should verify at each reinforcement level:
T
N

Jl

where:

TN
l

IJ

Tmax

Sh

u I t

( 34 )

IJ

Sv

Sh

Sv

S is the nail length in the active zone

Breakage failure of the inclusion:

For flexible nails which withstand only tension forces:
. A

where:

Sv

( 35)

Fall and A are the allowable tension stress and

cross-sectIonal area of the inclusion,
respectively.
For rigid nails which can withstand
both tension and shear forces, considering Tresca's
failure criterion:
Fall
y

where:

> TN

Sh

Ke q

TS

Sv

[ TN

Sh

>

+ 4

Tc
Sh

( 36 )

K eq

2 ]1/2
S

Tc is the maximum shear force in the inclusion,

140

Failure by excessive bending of a stiff inclusion:

M

>

is a factor of safety with respect to plastic

Fm = 1.5),

where:

(37)

max

b~nding (usually,

M is the plastic bending moment of the nail; for a

gtouted nail, an equivalent plastic bending moment
is calculated considering that the ~rout has a
compressive strength f of 210 KN/m (3,000 psi),
and zero tension stren~th.
The bending moment Mmax is derived from the "p - y" analysis:

0.32 Tc

Mmax

Mp

Lo

Sv

Sh

Lo' hence:

> 0.32 FS m Ts

( 38 )

Normal interaction between the soil and stiff inclusions:

The
normal interaction between the soil and relatively stiff inclusion
results in a progressive mobilization of the passive lateral soil
pressure on the inclusion, as illustrated schematically in figure
76.
This soil-inclusion interaction is analyzed using the "p - y"
analysis procedure outlined above.
In order to prevent plastic
flow (or creep) of the soil between the inclusions the maximum
lateral soil pressure Pel"
should not exceed half of the
ultimate lateral pressure1~~ the characteristic "p - y" curve.
In
french practice, this lateral soil pressure is limited to the
creep pressure obtained from a pressuremeter test.
The shear
force in the inclusion should therefore not exceed:
Tc
b.

P(lim)

Lo

D/2

(39 )

Global stability analysis

This analysis consists cf evaluating a global safety factor of the

soil nailed retaining structure and the surrounding ground with
respect to a rotational or translational failure along potential
sliding surfaces.
It requires determination of the critical
sliding surface which may be dictated by the stratification of the
subsurface soil or, in rock, by an existing system of joints and
discontinuities.
The potential sliding surface can be located
inside or outside the soil nailed retaining structure.
Evaluation of the global safety factor is generally based on the
rather conventional approach of limit equilibrium methods.
Slope
stability analysis procedures have been developed to account for
the available limit pullout, tension, and shearing resistance of
the inclusions crossing the potential sliding surfaces.
The limit

141

." -POTENTIAL
FAILURE SURFACE

BENDING OF A RIGID

Figure 76.

Bending of a rigid inclusion.

142

INCLUSION

equilibrium methods commonly used involve different definitions of

the safety factors, and a variety of assumptions with regard to
the shape of the failure surface, the type of soil-inclusion
interaction and the resisting forces in the inclusions.(ll, 12,

4 6 )

The German Method:

Stocker and coworkers proposed a limit force
equilibrium method (figure 77) considering a bilinear sliding
surface and assuming that the nails withstand only tension
forces.(46)
The shearing resistance of the soil, as defined by
Mohr - Coulomb's failure criterion, is assumed to be entirely
mobilized along the potential failure surface. The global safety
factor is defined as the ratio of the sum of the available
resisting limit nail forces ET to the total force ET required to
maintain limit equilibrium, th~t is:
FS = ET p /ET

( 40)

As shown in figure 77, the total force ET required to maintain

limit equilibrium is readily obtained considering the polygon of
forces acting on the rigid soil wedge limited by the potential
failure surface.
The resisting forces T are provided by the
pull-out capacity of the nails (i.e., th~ pullout capacity of the
portion of the nail located beyond the potential failure surface).
The inclination e of the failure surface is iteratively
determined to yie~d the minimum factor of safety. GassIer and
Gudehus have shown through stability analyses that the minimum
factor of safety is usually obtained for e = (n/4 - ~/2] assuming
a vertical line at wed~e A to be limited by the back of the
reinforced soil mass.( 8)
The assumed bilinear failure surface is mainly based on a limited
number of model tests where failure was caused by substantial
surcharge loading.
However, it does not appear to be consistent
with the observed behavior of nailed soil retaining structures
which are subjected mainly to their self-weight.
In particular,
stability analyses show that this bilinear failure surface is
generally not contained in the nailed soil mass and therefore
yields an active zone (or potential failure wedge) which is
substantially larger than that observed on actual structures.
The Davis Method:
Shen et al developed a similar force
equilibrium method (commonly called the Davis method).(ll)
They
consider a parabolic failure surface passing either entirely or
partially within the nailed soil mass and assume that the nails
withstand only tension forces.
Failure of the nailed soil system
can be generated by either pullout or breakage of the nails or
sliding of the soil along the failure surface.
The assumed
failure surface (figure 78b) is based on the contours of factor of
safety derived from finite element simulations, as shown in figure
78b.

143

F.S.

= ~Tp
'i:.T

Tp = T u1t .

Figure 77.

iTO. La

Force equilibrium method for global stability analysis

of nailed soil retaining structure.(46)

144

Potrnlul f~jlurf
!.urt3Ct

H " 25 It.
L=151t.

Contours of factor of safety derived from finite

element analysis

,H

Element 2

Sl

Nj

WI

E1ement 1

W
_

rZ
Ct

SJ

N(

Element 2

W: Body Weicht

s:

Element 1
F

Figure 78.

=~
c

I/}

",i
<t>

Ii>m

F= ~
L

.Tp

, ul t 1T. D . La

F -= F
L
S

Location of critical failure surface.(lll

145

TaoCem-ia\ force

In this analysis it is implicitly assumed that the safety factors

with respect to the shear strength of the soil,

cc

i.e., Fe

and F;

= tan

tan Pm

( 41 )

(where c and P are respectively the soil cohesion and internal

frictionmangle actually mobilized along the critical potential
failure surface) and the safety factor with respect to the
ultimate interface lateral shear stress,

i . e ., F 1

'Tu 1 t

( 42 )

/1: m

(where 'T is the lateral shear stress actually mobilized at the

soil-nail interfaces) are equal, and the global safety factor is
defined as:
FS

Fe

F+

( 43 )

F1

A minimum safety factor of 1.5 is generally required.

The lateral shear stress at the interfaces is calculated according
to Mohr-Coulomb failure criterion:
( 44 )

where N is the average normal stress along the adherence length

L.
This Driginal formulation has been extended under FHWA soil
nailing study to input interface limit lateral shear force per
unit length of nail as obtained from pullout tests.(76)
A slope stability analysis procedure, using the method-of-slices,
has been implemented to iteratively determine the critical sliding
surface and the minimum factor of safety.
To calculate the
interslice forces a stress ratio parameter K (i.e., ratio of the
lateral to the vertical stresses at the interslice) is input with
K values of 0.4 for frictional soils and 0.5 for cohesive soils.
Parametric sensitivity analyses have shown, however( that the
safety factor is fairly insensitive to the K value. 76)
Shen et al have evaluated their design procedure through analysis
of observed failure surfaces and failure heights of centrifugal
soil nailed model walls.
The method predictions were found to
agree fairly well with the experimental results.(11)
The French Method:
Common to all the limit equilibrium methods
specified above is the assumption that the inclusions withstand
only tension forces.
A more general method for the stability
analysis of nailed soil retaining structures, considering the two
fundamental mechanisms of soil-inclusion interaction (i.e.,
lateral friction and passive normal soil reaction), has been
developed by Schlosser.(12)

146

1is method (commonly called the French method) takes into account
)th the tension and shearing resistance of the inclusions as well
; the effect of their bending stiffness.
For an inclusion that
Lthstands both tension (T
) and shear (T ) forces, the
)bilized limit forces aremcalculated according to the principle
: maximum plastic work considering Tresca's failure criterium.
le T /T
ratio is a function of the inclination, a, of the
lclusiona~ith respect to the potential failure surface.
multi-criteria analysis, illustrated in figure 79, is conducted
) evaluate the global stability of the nailed soil system with
~spect to the four potential failure modes:
shear failure of the
)il along the critical sliding surface, pullout failure of the
lil, nail breakage by either excessive bending or combined effect
: tension and shear forces, and creep or plastic flow of the soil
!tween the nails. The global factor of safety is defined by
[uation:
Fe = f t = F 1 "= FS, and a minimum safety factor of 1.5
generally requ1red.
,r design purpose, input data of interface limit lateral shear
'rce per unit length of nail are obtained from pull-out tests.
Ie tensile strength of the inclusion is defined as the elastic
mit f
and the shear resistance as f /2.
The maximum shear
,rce ahd bending moments that can be tleveloped in the nail are
,verned by the soil-nail normal interaction.
They are calculated
,ing a conventional "p - y" analysis procedure, simulating the
'latively flexible nail by a laterally loaded, infinitely long
Ie, and are given, respectively, by equations 39 and 38.

ce

,is multi-criteria analysis procedure uses a classical slices

thod (e.g. Bishop's modified method or Fellinius's method) which
modified to take into account the effect of resisting nail
rces on the equilibrium of each slice. This analysis procedure
,ich is significantly more elaborated and comprehensive than
,ose outlined above, permits an evaluation of the effect of soil
ratification, ground water flow, and seismic loading on the
obal structure stability.
It can also be effectively used for
e design of mixed structures combining ground anchor and soil
il systems.
Post failure analyses of several nailed soil
taining structures have illustrated that with an appropriate
put design value of the ultimate lateral shear stress this
sign procedure could predict fairly well the pullout failure of
e structures.(391
c.

Evaluation of global stability analysis procedures

e Davis and the Fr'ench design procedures have been evaluated

rough the analysis of the field data reported in figure 71 which
re obtained on four full scale structures.(39)
Figure 80 shows
at the observed locus of the maximum tension forces in these
ructures agree fairly well with predicted locations of the
tential failure surface.

147

FORCES IN THE

BAR

----'-.. . . . ~/=:

f max

/
/

-~~--------~h~<----~~

/
"-SLIP

SURFACE

FAILURE CRITERIA
Sheer resistance of the bar
Scil bar friction

<..

max"""

Tmax"
:?..:r D Tult

A. f
s

< Rc =

A f
s

/2

NOTWel Lateral Earth Thrust on the bar

Shear resistance of the soil

Figure 79.

< c + a tan <t

Multicriteria slope stability analysis method.

148

tl21

J
I

0----<>--.1

DAVIS METHOD
FRENCH METHOD
KIN["MATICAL APFRQACH

o. FULL SCALE EXPERIt./,ENI CEBTP

(Experimentol Resul1s. Plumelle, 1986)

Figure 80.

b. PARISIAf\: WALL
(Experimenlo I Result 5, Cort ier
ond GiQon, 1983)

Predicted and observed locus of maximum

tension forces in nails.

149

The stability analyses of these structures were conducted using

the measured tension forces as yield forces of the inclusions and
assuming a pullout resistance large enough to prevent any sliding
of the inclusion (with F > 1.5).
This procedure yields, for the
mobilized nail forces, the actual value of the safety factor with
respect to the shear strength of the soil (F = F,).
The results
of these analyses illustrate that in soil nailed cut slopes the
factors of safety with respect to soil strength are generally
close to one.
Specifically, the safety factors obtained using the
Davis method for most of these structures were within the range of
1 10 percent.
The Davis method generally yielded safety factor
values which were about 15% lower than those predicted using the
French method.
It can be concluded, that in soil nailed retaining structures, as
well as in braced excavations, due to the staged construction
process, the soil resistance to shearing along the potential
failure surface is practically mobilized at the early stages of
excavation.
As the excavation proceeds, the load increments are
being entirely transferred to the inclusions.
Therefore, for the
design of these structures it appears more consistent to assign a
safety factor of one with respect to the shear resistance of the
soil along potential failure surfaces passing inside the soil
nailed system.
The global factor of safety of the system, as
defined by Stocker et ale (eq. 40), is evaluated with respect to
the pullout resistance of the inclusions.
Considering this
definition of the global safety factor a minimum safety factor of
2 is recommended.
GassIer and Gudehus recommend for nailed soil
structures the use of residual soil strength parameters factored
by 1.25 to comply with statistical evaluation criteria concerning
the probability of failure.(38)
The major limitation of the slope stability analysis procedures
currently used in design of soil nailed retaining structures lies
in the basic definition of a global factor of safety.
Observations on both full scale structures and reduced scale
laboratory models have illustrated that pullout failure is a
progressive phenomenon which is generally induced by the sliding
of the upper inclusions.
Therefore, this internal failure
mechanism cannot be adequately defined using a "global" value of a
unique safety factor for all the inclusions.
The local stability
at the level of the sliding inclusion can be significantly more
critical than the estimated "global" stability with respect to
general sliding in the retaining system or the surrounding ground.
The main conclusion that can be drawn for a reliable design of
these composite nailed - soil systems is that, consistently with
most reinforced soil structures, the design engineer should
attempt to evaluate both the local stability at the level of each
inclusion and the "global" stability of the structure.

150

REFERENCES
1.

AASHTO, "Standard specifications for Highway Bridges",

13th ed., American Association of state Highway and
Transportation Officials, Washington, D.C., 1983.

2.

Mitchell, J.K. and Villet, W.C.B., "Reinforcement of

Earth Slopes and Embankments", NCHRP Report No. 290,
Transportation Research Board, Washington, D.C., 1987.

3.

Devata, M.S., "Geogrid Reinforced Embankments with

Steep Slopes," Proc. of Conf. on Polymeric Grid
Reinforcement, Sponsored by the Science and Engineering
Research Council, p82, Telford, London, March, 1984.

4.

The Reinforced Earth Company, Arlington, Virginia

promotional publication, 1988.

S.

Enka Systems, promotional publication.

6.

Chang, J.C., Beaton, J.L., and Forsyth, R.A. "Design

and Field Behavior of the Reinforced Earth Embankment
California Highway 39" presented at the Jan. 21-2S,
1974 ASCE National water Resources Engineering Meeting,
Los Angeles, California, 1974.

7.

Jewell, R.A., "Material Requirements for Geotextiles

and Geogrids in Reinforced Slope Applications", Proc.
23rd Int. Man-Made Fibres Congress, Dornbirn, Austria,
Sept., 1984.

8.

Hueckel, S.M. and Kwasniewski, J., "Scale Model Tests

on the Anchorage Values of various Elements Buried in
Sand," Proc. Sth International Conference on Soil
Mechanics and Foundation Engineering, Paris, 1961.

9.

Schlosser, F., Juran, I. and Jacobsen, H.M., "Soil

Reinforcement" General Report, 8th European Conf. on
Soil Mechanics and Foundation Engineering, Helsinki,
1983.

10.

Phan, T.L., Segrestine, P., Schlosser, F., and Long,

N.T., "Stability Analysis of Reinforced Earth Walls by
Two Slip Circle Methods", Proc. Int. Conf. on Soil
Reinforcement, pp 119-123, paris, 1979.

11.

Shen, C.K., Bang, S., Romstad, J.M., Kulchin, L. and

Denatale, J.S., "Field Measurements of an Earth Support
System", J. Geot. Eng. Div., Vol. 107, GT12, ASCE,
1981.

12.

Schlosser, F., "Analogies et Differences Dans Ie

Comportement et Ie Calcul des Ouvrages de Soutennement
en Terre Armee et par Clouge du Sol", Annals de
L'Institut Technique du Batiment et des Travaux
Publics, No. 418, 1983.
151

13.

DiMaggio, J., "Mechanically stabilized Earth: walls and

Slopes", Report to Federal Highway Administration,
Washington, D.C., 1988.

14.

Weatherby, D.E., "Tiebacks", Report No. FHWA/RD-82/047,

Federal Highway Administration, Washington, D.C., 1982

15.

Elias, V., and Juran, I., "Soil Nailing", Report to

Federal Highway Administration, washington, D.C., 1988.

16.

Bastick, M., schlosser, F., Amar, S., and Canepa, Y.,

"Monitoring of a Prototype Reinforced Earth Wall with
Short strips," to be published in the Proc. of the 12th
ICSMFE, Rio de Janeiro, 1989.

17.

Terre Armee Internationale, Finite Element Study of

Reinforced Earth Structures, Rosalie, Internal Report,
1983.

18.

Meyerhof, G.G., "The Bearing Capacity of Foundations

Under Eccentric and Inclined Loads", Proc. 3rd Int.
Conf. on Soil Mechanics and Foundation Engineering,
pp 440-445, Zurich, Switzerland, 1953.

19.

Segrestin, P., Bastic, M.J., Seismic Design of

Reinforced Earth Retaining Walls: The Contribution of
Finite Element Analysis," Proc. of the Int. Symp. on
Theory and Practice of Earth Reinforcement, Kyushu,
Japan, Oct, 1988.

20.

Seed, H.B., and Mitchell, J.K., "Earthquake Resistant

Design of Reinforced Earth Walls", Internal Study for
the Reinforced Earth Company, Progess Report, Berkeley,
California, 1981.

21.

Seed, H.B. and Whitman, R.V., "Design of Earth

Retaining Structures for Dynamic Loads", Proc. of the
ASCE Specialty Conf., Lateral Stresses and Earth
Retaining Structures, Cornell University, pp 103-147,
Ithaca, NY, June, 1970.

22.

Christopher, B.R., Unpublished, STS Consultants, Ltd.

Report on Task C Field Test Wall Construction Prepared for FHWA, Washington, D.C., Contract No.
DTFH61-84-C-00073, "Behavior of Reinforced Soil", 1987.

23.

Wichter, L., Risseeuw, P., and Gay, G., "Large Scale

Tests on the Bearing Behavior of Woven Reinforced Earth
Wall", Proc. of the 3rd Int. Conf. on Geotextiles, Vol.
IV, pp 1073-1078, Vienna, Austria, 1986.

24.

Christopher, B.R., and Holtz, R.D., "Geotextile Design

and Construction Guidelines Manual," prepared for FHWA
National Highway Institute, Under Contract to
GeoServices, Inc. Contract No. DTFH61-86-R-00102, 1988.
152

25.

Christopher, B.R. and Holtz, R.D., Geotextile

Engineering Manual, STS Consultants, Ltd., Report to
Federal Highway Administration, No. FHWA-TS-86/203, pp
1044, Northbrook, Illinois, 1985.

26.

Schmertmann, G.R., Chouery-Curtis, V.E., Johnson, R.D.,

and Bonaparte, R., "Design Charts for Geogrid
-Reinforced Soil Slopes", Proc. Geosynthetics, Vol. 1,
pp 108-120, New Orleans, 1987.

27.

Cheney, R.S., and Chassie, R.G., "Soils and Foundation

Workshop Manual", FHWA, Washington, D.C., 1982.

28.

Fukuoka, M., "Fabric Retaining Walls", Proc. 2nd Int.

Conf. on Geotextiles, pp 575-580, Las Vegas, 1982.

29.

Juran, I., Schlosser, F., Legeay, G., and Long, N.T.,

"Experimentation en vraid Grandeur sur un mur Soumis a
des Surcharges en Tete a Dunkerque", Proc. Int. Conf.
on Soil Reinforcement: Reinforced Earth and Other
Techniques, Vol II, paris, 1979.

30.

Murray, R.T., "studies of the Behavior of Reinforced

and Anchored Earth", ph.D. Thesis, Hariot-Watt
University, Edinburgh, 1983.

31.

Juran, I., "Behavior of Reinforced Soil Structures",

Partial Report No.2, FHWA Contract No. 61-84-C-00078,
1985.

32.

Winkler, E., "Die Lehre von Elastizitat und

Festigkeit:, Prague, 1867.

33.

Cheney, R.S., "Permanent Ground Anchors," Federal

Highway Administration Report No. FHWA-DP-68-1R,
November, 1984.

34.

Jewell, R.A., Milligan, G.W., Sarsby, R.W., and Dubois,

D., "Interaction Between Soil and Geogrids", Proc.
Symp. on Polymer Grid Reinforcement in Civil
Engineering, Science and Engineering Research Council
and Netlon, Ltd., March, 1984.

35.

Gouvenot, D., and Bustamante, M.G., "Mesures In-situ

Sur Les Ouvrages Maritimes de soutenement" , Report,
Annales de L'Institut Technique du Batiment et des
Travaux Publique, Sept, 1979.

36.

Juran, I., Beech, J., and Delaure, E., "Experimental

Study of the Behavior of Nailed Soil Retaining
Structures on Reduced Scale Models," Proc., Int. Symp.
on In-situ Soil and Rock Reinforcements, paris, 1984.

37.

Cartier, G., and Gigan, J.P., "Experiments and

Observations on Soil Nailing Structures," Proc.,
Seventh Conf. of the ECSMFE, Helsinki, Finald, 1983.
153

38.

GassIer, G., and Gudehus, G., "Soil Nailing: Some

Mechanical Aspects of In-situ Reinforced Earth," Proc.,
10th ICSMFE, Vol. 3, pp 665-670, Stockholm, Sweden,
1981.

39.

Juran, I., and Elias, V., "Soil Nailed Retaining

structures: Analysis of Case Histories". ASCE, Special
Geotechnical Publication No. 12, pp 232-245, 1987.

40.

plumelle, C., "Experimentation en Vraie Grandeur d'une

Paroi Clouee," Revue Francaise de Geotechnique, No. 40,
pp 45-50, 1987.

41.

Juran, I., Beech, J., "Analyze Theorique du

Comportement d'un Soutenement en Sol Cloue", Proc. Int.
Conf. on In-situ Reinforcement of Soils and Rocks, pp
301-307, Paris, 1984.

42.

Juran, I., Baudrand, G., Farrag, F, and Elias, V.,

"Kinematical Limit Analysis Approach for the Design of
Nailed Soil Retaining Structures," Proc. Int.
Geotechnical Symp. on Theory and Practice of Earth
Reinforcement, Fukuoka Kyushu, Japan, 1988.

43.

Terzaghi, K., "Theoretical Soil Mechanics", John Wiley

& Sons, Inc., New York, 1943.

44.

Terzaghi, K., and Peck, R.B., "Soil Mechanics in

Engineering Practice," John Wiley & Sons, Inc., New
York, 1948 and 1967.

45.

Tschebotarioff, G.P., "Foundations, Retaining and Earth

Structures", McGraw-Hill Book Company, 1951.

46.

Stocker, M.F., Korber, G.W., GassIer, G., and Gudehus,

G., "Soil Nailing", Proc. Int. Conf. on Soil
Reinforcement, Vol. 2, pp 469-474, Paris, 1979.

47.

Steward, J.E. and Mohney, J., "Trial Use, Results and

Experience Using Geotextiles for Low-volume Forest
Roads," Proc., 2nd Int. Conf. on Geotextiles, Vol. 2 pp
335-340, Las Vegas, Nevada, August 1982.

48.

Yako, M.A., and Christopher, B.R., "Polymerically

Reinforced Retaining Walls and Slopes in North
America", The Application of Polymeric Reinforcement in
Soil Retaining Structures, P.M. Jarrett and A. McGown,
eds. pp 239-282, 1987.

49.

Dunnicliff, J., and Green, G.E., "Geotechnical

Instrumentation for Monitoring Field Performance," J.
Wiley & Sons, New York, 1988.

50.

Bell, J.R., and Steward, J.E., "Construction and

Observations of Fabri~ Retained Soil Walls," Prac. Int.
Conf. on the Use of Fabrics in Geotechnics, vol. I, pp
123-128, Paris, France, April, 1977.
154

51.

Mohney, J., "Fabric Retaining Wall, Olympic National

Forest," Highway Focus, Federal Highway Administration,
Vol. 9. No.1, pp 88-103, May, 1977.

52.

Chassie, R.G., "Geotextile Retaining Walls, Some Case

History Examples," prepared for presentation at the NW
Roads and Streets Conf., Corvallis, Oregon, February,
1984.

53.

AI-Hussaini, M.M., "Field Experiment of Fabric

Reinforced Earth wall," Proc. Int. Conf. on the Use of
Fabrics in Geotechnics, Vol. I, pp 119-121, April,
1977.

54.

AI-Hussaini, M.M., and Perry, E.B., "Effect of

Horizontal Reinforcement on stability of Earth Masses,
"Technical Report No. 5-76-11, U.s. Army Engineer
waterways Experiment Station, Vicksburg, Mississippi,
1976.

55.

Douglas, E.G., "Design and Construction of Fabric

Reinforced Retaining Walls by New York State,"
Transporation Research Board, Transportation Research
Record 872, pp 32-37, Washington, D.C., 1982.

56.

Bell, J.R., Barrett, R.K. and Ruckman, A.C.,

"GeotextileEarth-Reinforced Retaining Wall Tests:
Glenwood Canyon, Colorado," Transportation Research
Board, Transportation Research Record 916, pp 59-69,
Washington, D.C., 1983.

57.

Barrett, R.K., "Geotextiles in Earth Reinforcement,"

Geotechnical Fabrics Report, pp 15-19, March/April,
1985.

58.

Forsyth, R.A. and Bieber, D.A., "La Honda Slope Repair

with Geogrid Reinforcement," Proc. of the Symp. on Grid
Reinforcement in Civil Eng., Paper No. 2.2, London,
U.K., 1984.

59.

Berg, R.R., LaRochelle, P., Bonapart, R., and Tanguay,

L., "Gaspe' Peninsula Reinforced Soil Seawall Case
History," Proc. ASCE Symp. on Soil Improvement,
Atlantic City, New Jersey, April, 1987.

60.

Anderson, R.P., "Soil Reinforcement Objective: Polymer

Geogrid Replaces Galvanized Metal Strips in
Concrete-Faced Retaining Walls," Geotechnical Fabrics
Report, January/February, 1986.

61.

Scott, J.D., et aI, "Design of the Devon Geogrid Test

Fill," Proc. Geosynthetics '87 Conf. New Orleans,
Louisiana, February, 1987.

155

62.

Jaber, M., "Behavior of Reinforced Soil Walls in

Centrifuge Model Tests", Dissertation submitted in
partial satisfaction of the requirements for the degree
of Doctor of Philosophy, University of California,
Berkeley, Department of Civil Engineering, March, 1989.

63.

Alimi, Bacot, J., Lareal, P., Long, N.T., and

Schlosser, F., "Etude de l'Adherence Sols-armatures,"
Proc. 9th Int. Conf. on Soil Mechanics and Foundation
Engineering, Vol. I, pp 11-14, Tokyo, 1977.

64.

ASTM, "Soil and Rock; Building stones," Annual Book of

ASTM Standards, Section 4, Vol. 04.08, American Society
for Testing and Materials, Philadelphia, PA, 1988.

65.

Rowe, R.K., and Davis, E.H., "The Behavior of Anchor

plates in Sand," Geotechnigue 32, No.1, pp 25-41,
1982.

66.

Bloomfield, R.A. "Proposed Design Modifications - VSL

,Retaining Earth," VSL Corporation, January, 1984.

67.

Nielsen, M.R., "Pullout Resistance of welded Wire Mats

Embedded in Soil", Masters Thesis, utah State
University, Logan, UT, 1983.

68.

Murray, R.T., "Studies of the Behavior of Reinforced

and Anchored Earth," ph.D. Thesis, Heriot-Watt
University, Edinburgh, 1983.

69.

Jewell, R.A., "Some Effects of Reinforcement on the

Mechanical Behavior of Soils," Ph.D. Thesis, University
of Cambridge, 1980.

70.

Juran, I., and Chen, Ch.L., "Soil-Geotextile Pullout

Interaction Properties: Testing and Interpretation",
Transportation Research Board, 67th Annual Meeting,
Paper No. 87-0159, 1988.

71.

Littlejohn, G.S. and Bruce, D.A., "Rock Anchors

State-of-the-Art," Part I:
Design and Part II:
Construction, Ground Engineering, May, 1975.

72.

Peck, R.B., "A study of the Comparative Behavior of

Friction Piles", Highway Research Board Special Report
No. 36, p 72, 1958.

73.

Woodward, R.J., Lundgren, R., and Boitono, J.D., "pile

Loading Tests in stiff Clays", Proc., 5th ICSMFE, Vol.
2, pp 177-184, 1961.

74.

Louis, C., "Theory and Practice in Soil Nailing

Temporary or Permanent Works," ASCE Annual Conference,
Boston, Oct., 1986.

156

75.

Louis, C., "Nouvelle Methode de soutennement des Sols

en Deblais," Revue Travaux No. 533, 1981.

76.

Elias, V., and Juran, I., Draft Manual of Practice for

Soil Nailing, prepared for U.S. Department of
Transportation, FHWA, Contract DTFH-61-85-C-00142, May,
1988.

77.

LCPC-SETRA, "Reglis de Justification des Foundations

err Pioux", 1985.

78.

Singh, A., and Mitchell, J.K., "General

Stress-Strain-Timi Function for Soils", 1968, J. SMFD.,
ASCE, Vol. 94, SMI, pp 21-46.

79.

Bustamante, M., "Capacite de' Ancrage et Comportement

des Tirants Injectes, Scelles dans une Angile
Plastique," These' Docteur-Ingenieur ENPC, Paris.

80.

Wilding, M.A. and Ward, I.M., "Tensile Creep and

Recovery in Ultrahigh Modulus Linear Polyethylene", J.
Polymer, 19, pp 969-976, 1978.

81.

McGown, A., Andrawes, K.Z., and Kabir, M.H.,

"Load-Extention Testing of Geotextiles Confined in
Soil", Proc. Second International Conference in
Geotextiles, Vol. 3, pp 793-798, Las Vegas, Nevada,
1982.

82.

John, N., Johnson, P., Ritson, R., Petley, D.,

"Behavior of Fabric Reinforced Soil Walls, Proceedings
of the Second International Conference on Geotextiles,
Las Vegas, Nevada, IFAI, st. paul, MN, 1982.

83.

Jarrett, P.M. and McGown, A., "The Application of

Polymeric Reinforcement in Soil Retaining Structures,"
Proc. NATO Advanced Research Workshop, Kluwer Academic
publishers, Kingston, Ontario, Canada, 1987.

84.

Adib, Mazen, E., "Internal Lateral Earth Pressure in

Earth walls", Dissertation Submitted in Partial
Satisfaction of the Requirements of the Degree of
Doctor of Philosophy, Department of Civil Engineering,
University of California, Berkeley, 1988. Note: This
research was supported primarily by the Institute for
Transportation Studies, University of California, 1988.

85.

Schmertmann, G., Chew, S.H., Mitchell, J., "Finite

Element Modeling of Reinforced Soil Wall Behavior,"
Geotechnical Engineering Report No. UCB/GT/89-01,
University of California, Berkeley, September, 1989.

86.

Bastick, M., Schlosser, F., Amar, S., and Canepa, Y.,

"Instrumentation d'un Mur Experimental en Terre Armee a
Armatures Courtes", Proc. 12th ICSMFE, 1988. To be
published.
157

87.

Bastick, M., and schlosser, F., "Comportement et

Dimensionnement Dynamique des Ouvrages en Terre Armee,
ler ColI. Nat. de Genie Parasismique,
Saint-Remy-les-Chevreuse France, January, 1986.

88.

Tomlinson, M.J., "The Adhesion of Piles Driven in Clay

Soils", Proc., 4th ICSMFE, Vol. 2, pp 61-71, 1957.

89.

Ward, I.M., "The Orientatin of Polymers to Produce High

Performance Materials," Proc. Symp. on Polymer
Performance in Civil Engineering, London, 1984.

90.

Juran, I., Shafiee, S., and Schlosser, F., "Numerical

Study of Nailed Soil Retaining Structures," Proc., 11th
ICSMFE, Vol. 4, pp 1713-1717, San Francisco, 1985.

91.

Shafiee, S., "Simulation Numerique du Comportement des

sols Cloues: Interaction Sol-Renforcement et
Comportement de L'ouvrage", Ph.D. Diss, ENPC, Paris,
1986.

92.

Frank, R., Guenot, A., and Humbert, P., "Numerical

Analysis of Contact in Geomechanics," Proc., 4th Int.
Conf. on Numerical Methods in Geomechanics, Edmonton,
1982.

93.

Zaman, M.M., Desai, C.S., and Brumm, E.C., "Interface

Model for Dynamic Soil-Structure Interaction", Journal
of Geotechnical Engineer Division, ASCE, Vol. 110, SM9,
pp 1257-1273, September, 1984.

158