Citation
Performance of a large scale fabric wall during load test

Material Information

Title:
Performance of a large scale fabric wall during load test
Creator:
Billiard, John William
Publication Date:
Language:
English
Physical Description:
130 leaves : illustrations ; 29 cm

Subjects

Subjects / Keywords:
Walls -- Testing ( lcsh )
Weight (Physics) ( lcsh )
Walls -- Testing ( fast )
Weight (Physics) ( fast )
Genre:
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )

Notes

Bibliography:
Includes bibliographical references (leaves 101-103).
General Note:
Submitted in partial fulfillment of the requirements for the degree, Master of Science, Department of Civil Engineering.
Statement of Responsibility:
by John William Billiard.

Record Information

Source Institution:
University of Colorado Denver
Holding Location:
Auraria Library
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
20960351 ( OCLC )
ocm20960351
Classification:
LD1190.E53 1989m .B54 ( lcc )

Full Text
PERFORMANCE OF A IARGE SCALE
FABRIC WALL DURING LOAD TEST
by
, John William Billiard
B.S. Civil Engineering, University of Colorado at Denver, 1985
A thesis submitted to the
Faculty of the Graduate School of the
University of Colorado in partial fulfillment
of the requirements for the degree of
Master of Science
Department of Civil Engineering
1989


This thesis for the Master of Science degree by
John William Billiard
has been approved for the
Department of
Civil Engineering
by
Date
28, mg


Billiard, John William (M.S., Civil Engineering)
Performance of a Large Scale Fabric Wall During Load Test
Thesis directed by Associate Professor Jonathon Tzong H. Wu
Geotextile-reinforced earth walls present an economical
alternative for the construction of retaining walls.
Geotextile-reinforced earth walls already constructed have
performed successfully, but there is a lack of understanding
regarding the reinforcing mechanism when a geotextile fabric is
placed within the backfill. To obtain a better understanding of
the reinforcing mechanism, a controlled test wall was designed,
instrumented and constructed.
A five foot 3-inch tall geotextile-reinforced earth
wall was constructed inside a test apparatus designed and built
for the study of earth structures in a plane strain condition.
The wall was constructed using a spunbonded nonwoven
polypropylene geotextile fabric in five layers, ranging from
0.85 feet thick at the bottom to 1.24 feet thick at the top.
The wall was designed using the Forest Service design method to
have a factor of safety in rupture of 1.0 at 850 pounds per
square foot surcharge pressure.
The geotextile fabric layers were successfully
instrumented for measurement of strain using a newly developed
method of attaching a strain gage to the geotextile surface.
Along with strain measurement, lateral and vertical deflections
were obtained along with information on the internal movement of
the soil mass.


iv
The strain and displacement measurements provided
excellent controlled test data for subsequent verification of
analytical models in investigation of the performance of
geotextile-reinforced earth walls.
The form and content of this abstract are approved. I recommend
its publication.
Signed ______
F


ACKNOWLEDGEMENTS
It is very rare when a project is accomplished and the
persons that helped to bring that project to a successful close
can be thanked in a formal manner. I would like to take this
opportunity to thank all of them.
This research was performed under the supervision of
Jonathon T. H. Wu. His positive attitude, constant support and
friendship played a paramount role in the successful completion
of this venture. To him, I express my deepest appreciation.
Sincerest gratitude is extended to my other committee
members, N. Y. Chang, and Judith J. Stalnaker, for their
friendship, guidance and understanding throughout my academic
career. To my colleges and friends in the University
Geotechnical/Structural laboratory and at the Colorado Highway
Department, I express my thanks for many fruitful conversations.
I am eternally indebted to James Crofter and his staff for all
of their help and ideas, for without them, the success of this
project would not have been possible.
With greatest pleasure, I thank my father Francis E.,
and mother Patricia A., for a lifetime of love, support and
encouragement. Words cannot describe the depth of my gratitude.
Finally, I cannot thank my wife Judith A. enough for
all of her love, support and understanding. Without her, none
of this would be possible.


CONTENTS
CHAPTER PAGE
1. INTRODUCTION ................................... 1
1.1 Problem Statement ................... 1
1.2 Study Objective ............................ 2
1.3 Method of Research ................... 2
2. DESIGN METHOD OF THE TEST WALI.................... 4
2.1 Description of the Design Method ........... 4
2.1.1 Lateral Earth Pressure ... 4
2.1.2 Fabric Tension and Layer
Thickness ..................... 5
2.1.3 Factor of Safety Versus
Layer Thickness ............... 6
2.1.4 Resistance to Pullout of
the Fabric .................... 6
2.2 Test Wall Design ......................... 8
2.2.1 Factor of Safety Versus
Layer Thickness ............... 9
2.2.2 Factor of Safety Versus
Pullout Resistance ........... 11
3. TEST APPARATUS AND INSTRUMENTATION ............... 14
3.1 Description of Test Apparatus ............. 14
3.1.1 Design of Steel Box .......... 14
3.1.2 Timber Facing ................ 15
3.1.3 Acrylic Sheeting ............. 15
3.1.4 Back Wall .................... 18


vii
CHAPTER PAGE
3.1.5 Teflon Sheeting ............ 18
3.2 Description of Load Frame and
Hydraulic Jack .......................... 34
3.3 Description of Instrumentation .......... 36
3.3.1 Deflection Indicators ........... 36
3.3.2 Yarn Grid ....................... 36
3.3.3 Hydraulic Load Cells ............ 44
3.3.4 Strain Gages ............... 50
4. MATERIAL PROPERTIES ............................. 60
4.1 Soil Properties ......................... 60
4.2 Geotextile Fabric Properties ............ 65
4.3 Soil-Fabric Interface Properties ........ 66
4.4 Acrylic Plastic ......................... 66
4.5 Teflon Sheets ........................... 66
5. TEST WALL CONSTRUCTION .......................... 69
5.1 Test Box Preparation .................... 69
5.2 Fabric Preparation ...................... 70
5.3 Strain Gage Attachment .................. 70
5.4 Temporary Support During Construction . 73
5.5 Backfilling ............................. 74
5.6 Overlapping ............................. 76
5.7 Test Wall Geometry ...................... 78
5.8 Surcharge Load Application .............. 80
6. TEST RESULTS AND DISCUSSION OF RESULTS .......... 82
6.1 Strain and Deflection Results ........... 82
6.2 Strain Versus Surcharge ................. 90


viii
CHAPTER PAGE
6.3 Lateral and Vertical Deflection
Versus Surcharge ................. 95
7. SUMMARY AND CONCLUSIONS ................. 98
7.1 Summary .................................. 98
7.2 Conclusions .............................. 99
BIBLIOGRAPHY ............................................. 101
APPENDIX
A. INTERFACE PROPERTY DIRECT SHEAR DATA .......... 104
B. STRAIN GAGE CALIBRATION DATA ..................... 115
C. TRIAXIAL TEST PLOTS .............................. 117
D. PULLOUT TEST DATA ................................ 120
E. LOAD TEST STRAIN GAGE AND DEFLECTION DATA ........ 122
F. ANCILLARY DATA AND INFORMATION ................... 129


FIGURES
FIGURE PAGE
2- 1 Test Wall Geometry ............................... 10
3- 1 Structural Steel Supports And Timber Facing
Installed On Test Apparatus ..................... 16
3-2 Installation Of Acrylic Sides .................... 17
3-3 Completed Test Apparatus ....... 19
3-4 Interface Friction Angle For The
Soil/Teflon/Acrylic Interface ................... 20
3-5 Direct Shear Test Results Normal Stress
= 3.95 PSI On Soil/Teflon/Acrylic Interface ... 21
3-6 Direct Shear Test Results Normal Stress
= 7.9 PSI On Soil/Teflon/Acrylic Interface .... 22
3-7 Direct Shear Test Results Normal Stress
= 15.8 PSI On Soil/Teflon/Acrylic Interface ... 23
3-8 Direct Shear Test Results Normal Stress
= 31.6 PSI On Soil/Teflon/Acrylic Interface ... 24
3-9 Interface Friction Angle For The
Teflon / Teflon Interface ....................... 26
3-10 Direct Shear Test Results Normal Stress
= 7.9 PSI On Teflon / Teflon Interface .......... 27
3-11 Direct Shear Test Results Normal Stress
= 15.8 PSI On Teflon / Teflon Interface ......... 28
3-12 Direct Shear Test Results Normal Stress
= 31.6 PSI On Teflon / Teflon Interface ......... 29
3-13 Interface Friction Angle For The
Teflon / Acrylic Interface ...................... 30
3-14 Direct Shear Test Results Normal Stress
= 7.9 PSI On Teflon / Acrylic Interface
31


X
FIGURE PAGE
3-15 Direct Shear Test Results Normal Stress
= 15.8 PSI On Teflon / Acrylic Interface ......... 32
3-16 Direct Shear Test Results Normal Stress
= 31.6 PSI On Teflon / Acrylic Interface ......... 33
3-17 Installation of Teflon Sheeting ..................... 35
3-18 Installation Of Movement Measurement Grid ........... 38
3-19 Measurement Grid Marks On Side Of Apparatus ... 39
3-20 Internal Movement Of Soil Mass .................... 40
3-20-A Internal Movement Of Soil Mass, Point A ........... 41
3-20-B Internal Movement Of Soil Mass, Point B ........... 42
3-20-C Internal Movement Of Soil Mass, Point C ........... 43
3-21 Construction Of Steel Casing Load Cells ............. 46
3-22 Hydraulic Calibration Of Load Cells ................. 47
3-23 Equipment Used To Calibrate The Load
Cells In Sand ...................................... 48
3-24 Hydraulic Load Cell Attached To The Fabric
Layer Face ......................................... 49
3-25 Strain Gages Prepared For Attachment
By Standard Glue Procedure To Geotextile
Fabric ............................................. 52
3-26 Strain Gage Being Glued To Fabric Using
Strain Gage Glue In Accordance With Accepted
Strain Gage Attachment Procedure ................... 53
3-27 Strain Gages Being Glued Down Using Blocks And
Weights ............................................ 54
3-28 Typical Strain Gage Glued By Standard Glue
Procedure .......................................... 55
3-29 Strain Gage Glued Only At The Ends .................. 56
3-30 Fabric Pieces Used For Strain Gage
Calibration ........................................ 57


xi
FIGURE PAGE
3-31 Planar Pull Test Calibration Equipment ........... 58
3- 32 Calibration Curve For Strain Gage Glued
At Ends ......................................... 59
4- 1 Gradation Curve For "Squeegee" Material .......... 61
4-2 Calibration Of Squeegee Placement Density
Raining Down Into A Bucket Of Known Volume .... 63
4-3 Placement Of Squeegee Into Test Wall ........... 64
4- 4 Fabric Pull Out Test, Typar 3301 .............. 67
5- 1 Test Apparatus With Squeegee Glued To
The Floor ..................................... 71
5-2 Marked Fabric Pieces With Strain Gages
Attached ..................................... 72
5-3 View Of Test Wall After Failure in Pull Out
In Layer Number 1 ............................ 75
5-4 Layer 2 Under Construction ...................... 77
5- 5 Test Wall Geometry As Constructed .............. 78
6- 1 Test Wall Response At A Surcharge
Pressure = 295 PSF ............................. 83
6-2 Test Wall Response At A Surcharge
Pressure = 850 PSF ............................. 84
6-4 Test Wall Response At A Surcharge
Pressure = 1380 PSF ............................ 86
6-3 Test Wall Response At A Surcharge
Pressure = 1950 PSF ............................ 88
6-5 Test Wall Response At A Surcharge
Pressure = 2660 PSF ............................ 89
6-6 Strain Versus Surcharge, Layer 1 .............. 91
6-7 Strain Versus Surcharge, Layer 2 .............. 92
6-8 Strain Versus Surcharge, I.ayer 3 .............. 93


xii
FIGURE PAGE
6-9 Strain Versus Surcharge, layer 4 .............. 94
6-10 Locus Of Peak Strain At Surcharge
Pressures Greater Than 2,000 PSF .................. 96
6-11 Lateral Deflection Versus Surcharge ............... 97


CHAPTER 1
INTRODUCTION
1.1 Problem Statement
Geotextile-reinforced earth walls have been designed
and constructed with excellent performance characteristics (Al-
Hussaini, 1977; Barrett, 1985; Bell, et al., 1975; Bell, et al.,
1977; Bell, et al., 1983; Delmas, et al., 1988; Douglas, 1982;
Mohney, 1977). In actual construction, geotextile-reinforced
retaining walls have repeatedly demonstrated many advantages
over conventional gravity walls and to a lesser extent, over
other types of flexible walls. Advantages include 1) a flexible
wall system; 2) minimum excavation; 3) strong resistance to
corrosion and bacterial action; 4) backfill can contain fines;
5) little drainage problem; 6) unskilled labor can be used; and
7) low cost. This superior wall system however suffers a major
disadvantage. When geotextiles are placed into backfill, there
is a total lack of understanding as to the reinforcing mechanism
once the walls are constructed.
The design of geotextile-reinforced earth walls in
engineering practice is currently conducted using a number of
diverse design methods. These design methods are typically
based on relatively simple analytical models from classical


2
earth structures, rather than on more realistic models which
account for the complex soil-geotextile interaction. Since a
relatively small number of controlled tests on the behavior of
these structures has been conducted to date (Wichter, et al.,
1986; Christopher, 1988), the analytical models used are based
more on assumptions than on supporting empiricism. The result
is diverse results among the various methods (Claybourn, 1989).
1.2 Study Objective
The objective of this study was to establish controlled
test data for investigation of the geotextile-reinforced earth
wall under an increasing uniform surcharge load in a plane
strain condition. Controlled data was obtained by conducting a
large-scale controlled laboratory test on a geotextile-
reinforced earth wall designed using the Forest Service method
(Steward, et al., 1983).
Reliable strain data on embedded fabric has been
difficult to obtain (Delmas, et al., 1988), therefore as part of
this study, a reliable method of strain measurement was
investigated.
1.3 Method of Research
In this study, the geotextile-reinforced earth wall was
designed in accordance with the Forest Service design method to
have a factor of safety in rupture equal to 1.0 at a surcharge
of 850 pounds per square foot (psf), while the factor of safety
in pullout was larger than 1.0.


3
The wall was constructed inside a test apparatus built
for testing earth structures in a plane strain condition. The
constructed wall was composed of five layers of geotextile and
stood 5' 3" inches tall by 30" wide by 7' in length. Surcharge
loads were applied in increments by using a hydraulic jack
reacting against a steel beam. Deformations of the embedded
geotextile were measured using high elongation strain gages
attached by a method unique to the author. Lateral and vertical
displacement measurements were made using dial indicators. In
addition, a grid system installed on the transparent side wall
was employed to track the internal movements of the soil mass.
A complete set of data was obtained for each surcharge load
increment.


CHAPTER 2
DESIGN METHOD OF THE TEST WALL
The Forest Service design method was used for the
design of the test wall. The method was originally published by
the Federal Highway Administration, Report number FHWA-TS-78-205
in June, 1977 and subsequently revised in June, 1983 (Steward et
al.,1983) The Forest Service design method was chosen for this
test due to the popularity of the method in actual construction.
The design method uses a limiting equilibrium approach to
evaluate the horizontal forces due to lateral earth pressures to
the stabilizing in or along the geotextile reinforcement.
2.1 Description of the Design Method
2.1.1 Lateral Earth Pressure
Lateral pressure at the face of the wall is derived
from the earth pressure behind the wall face and from live loads
at the top of the wall. Lateral pressure due to the weight of
the fill material at depth is assumed to act in a triangular
distribution at the face of the wall. For the Forest Service
design method, the lateral earth pressure is determined using
the at-rest lateral earth pressure constant, K0. Lateral
pressure due to live point loads at the top of the wall is
assumed to act at the face of the wall as calculated from the


5
Boussinesq equations.
2.1.2 Fabric Tension and Layer Thickness
To compute the average fabric tension for the fabric
face, or the fabric layer thickness by the Forest Service
method, the following is accomplished.
The lateral pressure midway between the top and the
bottom of a layer is determined and then multiplied by the layer
thickness to calculate the fabric tension. If the allowable
fabric tension is known at this point, the equation can be used
to determine the thickness of the fabric layers.
The layer sizing portion of the design is an iterative
procedure. Given the allowable tension for the fabric, a depth
to mid-layer is estimated with a corresponding average lateral
earth pressure and then a layer spacing is calculated from:
( o- h) (F.S.)
where,
x = Layer spacing
S = Fabric Strength (Wide Width Tensile
strength was used in this study)
"h = Lateral Earth Pressure
F.S. = Factor of Safety Against Rupture
Unless a considerable amount of good fortune prevails,
the calculated spacing has a depth to mid-layer different than
the trial spacing used to calculate the average lateral earth
pressure. The new calculated depth to mid-layer thickness is


6
then used to determine the new average lateral earth pressure
and a new fabric layer spacing is calculated. The above
procedure is repeated until the depth to mid-layer converges
upon a layer thickness.
In practice, the thickness of all of the layers are
calculated, and then one thickness is often chosen for a
particular range of depths. The chosen thickness is usually the
smaller thickness for that depth range. This is done for a more
conservative design and easier field construction.
2.1.3 Factor of Safety Versus Layer Thickness
The Forest Service design method applies a safety
factor to the calculation of layer thickness to account for the
allowable tension to be applied to the fabric, and also to
account for unanticipated loading of the wall, effects from
creep, variations in soil parameters and other unknown
parameters.
2.1.4 Resistance to Pullout of the Fabric
Pullout of the fabric from the soil is resisted by the
embedment of the fabric into the backfill material. The length
of embedment term describes the bottom length of the fabric
layer embedded beyond the assumed Rankine failure plane. The
length of overlap term describes the top length of the fabric
that is embedded to form the top of the fabric layer.
For the bottom length of embedded fabric, only the
length of fabric beyond the assumed failure plane is assumed to


7
offer resistance to pullout. This is due to the fact that if
the failure plane develops, the fabric inside the failure mass
will be of no use in resisting pullout.
The length of embedment is calculated as:
(K0) (x) (F.S.)
Le = ---------------------------
(2) (tan (2/3 4> ))
where,
Le = Embedment Length
K0 * At-Rest Lateral Earth Pressure Coefficient
x = Layer Spacing
F.S. = Factor of Safety Against Pullout
$> = Angle of Internal Friction, Soil
The (2/3 $> ) term is an estimate of the soil/fabric
angle of friction at the interface.
The embedment length used must be either the calculated
value, or a minimum of 3 feet beyond the assumed failure plane.
The assumed Rankine failure plane is located at an angle of
(45 +<£/2) upward from the horizontal surface starting at the
toe of the wall face for level backfill.
Kq can be calculated by:
K0 = (1 sin $')
Overlap length is critical to a successful construction
and performance of the fabric retaining wall. Overlap is
required to maintain a stable wall facing.


8
Overlap length is calculated as:
( crh) (x) (F.S.)
(2) (df) ( y ) (2/3 tan 4> )
* Overlap Length
= Lateral Earth Pressure at the Overlap Depth
= Layer Spacing
* Factor of Safety Against Overlap Pullout
= Depth to Overlap
= Bulk Unit Weight of the Soil
= Angle of Internal Friction, Soil
The length of overlap used must be either the
calculated value or a minimum of 3 feet. Generally, the
embedment length and the overlap length are made a uniform
length in practice.
2.2 Test Wall Design
The fabric test wall used in this study was designed
using the Forest Service method. Where possible, safety factors
were set equal to unity in order to assess the safety margin of
the Forest Service method. However, the safety factor for the
resistance to fabric pullout was designed to be greater than
1.0. This was done to prevent a premature failure during
construction of the test wall.
where,
^h
x
F.S.
r
*
2.2.1 Factor of Safety Versus Layer Thickness


9
For purposes of this study, the safety factor for layer
thickness calculation was held equal to one, which required the
factor of safety for allowable fabric tension to also be held
equal to one. In this manner, the test wall was designed to
fail in rupture according to the Forest Service design method.
Rupture failure would occur at 850 psf uniform surcharge
pressure applied over the top of the wall.
The iterative procedure to calculate the layer spacing
was used to determine the exact layer thickness that would be
constructed. The layer thicknesses used in the test wall sample
varied from 0.85 thick at the bottom layer number 5, to 1.24
feet thick at the top layer number 1. See Figure 2-1 for the
wall geometry.
For the layer thickness calculations, the Wide Width
Strip tensile fabric strength (ASTM D4595-86) was used. No
factor of safety was used to reduce the allowable fabric
strength.


1' SQUARE
Figure 2-1
Test Wall Geometry


11
2.2.2 Factor of Safety Versus Pullout Resistance
Longer lengths than those calculated with F.S. equal to
1.0, for fabric embedment length and overlap length were used in
the construction of the test wall. The embedment lengths
corresponding to a factor of safety of one were calculated to be
only a few inches. This indicated that pullout was likely to
occur if the test wall was constructed with all of the factors
of safety equal to 1.0. The factors of safety against pullout
were chosen to be larger than 1.0 to allow for successful
construction of the test wall.
The minimum embedment length of 3 feet, as recommended
by the Forest Service design criteria, was used in the
construction of both Test Walls 1 and 2.
Overlap lengths corresponding to a F.S. equal to 1.0
was also calculated to be only a few inches. Since this did not
seem reasonable and would also pose a potential pullout problem
during wall construction, the length of overlap was extended.
However, the recommended overlap length of 3 feet was not used.
For Test Wall 1, an overlap length of 1 foot was used
for all five layers.
The top layer number 1 in Test Wall 1, failed in
pullout immediately after construction and before any
appreciable load was applied.
A length of 1.5 feet was used as the overlap length in
Test Wall 2. The overlap length of 1.5 feet was used to lower


12
as much as possible, the number of fabric sheets intersecting
the assumed potential failure plane. An overlap distance of 1.5
feet was used for Test Wall 2 for layers 2, 3, 4 and 5. The top
layer number 1, had an overlap length of 6.5 feet. This was a
corrective measure to counter any pullout of fabric in the top
layer until a sufficient load could be introduced.
With the factor of safety for rupture held equal to
1.0, the constructed layer thickness was the thickness as
calculated. No inherent strength of a layer was developed from
using a layer thickness that was smaller than the theoretically
calculated value.
The factor of safety for rupture at 850 psf was 1.00
for layers 1, 2, 3, 4 and 5. However, by using longer embedment
and overlap lengths, the factor of safety against pullout was
increased to more than one. In fact, the factor of safety for
pullout at a uniform pressure of 850 psf for layers 1, 2, 3, 4
and 5 were calculated to be 2.5, 5.3, 8.2, 11.1 and 14.2
respectively.
Consequently, according to the Forest Service design
method, the wall would not fail in pullout but would fail in
rupture at a uniform surcharge pressure of 850 psf.
To assess the inherent factor of safety in rupture of a
geotextile-reinforced earth wall, a failure has to be reached.
Once a failure is obtained during a load test of a fabric wall,
the inherent factor of safety can be assessed by comparing the


13
actual failure surcharge to the theoretical failure surcharge.


CHAPTER 3
TEST APPARATUS AND INSTRUMENTATION
For this study, a large-scale geotextile-reinforced
earth wall was constructed. The wall was instrumented to
monitor its performance under incremental loads. Measurements
were taken after each load increment was applied.
3.1 Description of Test Apparatus
3.1.1 Design of Steel Box
The five foot tall test wall was constructed inside a
steel ribbed test box. The overall interior dimensions of the
box were 66" tall by 30" wide by 96" long. Each side of the box
had 8 vertical standard A36 C 3 X 4.1 channels 69" tall, with
varied horizontal center to center spacings. The channels were
spaced closer together in the central portion of the box with a
wider spacing near each end of the box. Lateral support for the
channels was provided by A36 2" X 2" X 3/16" angles spaced
vertically apart between 12" and 18".
The side walls were separated using 2" X 2" angles, 38"
long, fastened to the vertical channel supports by 3/8" diameter
mild steel bolts. The angle irons were fastened to the top of
every vertical channel except for the center channel. An angle
iron was not bolted across the center channel because it would


15
have interfered with the loading plates and hydraulic jack.
Angle irons were also bolted at 12" to 18" vertical intervals
across each end of the box. In this manner, the walls were
laterally supported at each end and at the top. Four 3/8"
thread rods were bolted to the four center channels to provide
lateral spread resistance at the bottom of the box. See Figure
3-1 for a view of the steel support structure.
3.1.2 Timber Facing
Timbers of 2" by 4" were cut and fastened with the edge
against the angle iron and timber materials 1/2" thick by 1.5"
wide were fastened along the edge of the channels that faced the
interior of the box. This was done to provide a timber facing
for the acrylic plastic sheets that were placed against the
interior sides. See Figure 3-1 for a view showing the timber
facing.
3.1.3 Acrylic Sheeting
Acrylic plastic sheets 1/2" thick were fastened to the
entire interior sides of the metal walls along the top edge and
the bottom edge in as few places as practical. The limited
bolted connections were in an effort to prevent the cracking of
the plastic during the loading from stress concentrations, and
to limit the number of bolt heads protruding into the interior
of the box so that a smooth side would result. See Figure 3-2
showing the installation of the acrylic sheeting.


16
i
Figure 3-1 : Structural Steel Supports And Timber Facing
Installed On Test Apparatus
I
!
I
I


17


18
3.1.4 Back Wall
A 30" wide by 66" tall piece of 3/4" plywood was bolted
to the interior of the box along the back wall. Plywood was
used because no visual information need be obtained from the
rear wall. See Figure 3-3 for a view of the completed test
apparatus.
3.1.5 Teflon Sheeting
Two layers of teflon sheets, 3 mills (0.003") thick by
12" wide were placed along the full interior of the acrylic
sides. This was accomplished to eliminate side friction between
the soil and acrylic as much as possible.
Figures 3-4, 3-5, 3-6, 3-7, and 3-8 show the results of
direct shear type friction tests of soil/teflon/acrylic
interface. The friction angle of soil/teflon/acrylic was
measured to be approximately 4.3 degrees.
During the direct shear tests, it was noted that one
sheet of the teflon would slip with respect to the other sheet
of teflon on tests whose results are shown in Figures 3-5 and 3-
7. During the other two tests shown in Figures 3-6 and 3-8, one
sheet of teflon slipped with respect to the acrylic surface.
This indicated that the friction angle of teflon/tef1 on and
teflon/acrylic was close and slippage between teflon/tefIon and
teflon/acrylic depended on the soil grain configuration beneath
the teflon sheets. The acrylic was well protected using the two
sheets of teflon and showed only very minor damage after the


19
Figure 3-3 : Completed Test Apparatus
I
I
i


SHEAR STRESS. PSl
20
Figure 3-4 :
Interface Friction Angle For The
Soil/Teflon/Acrylic Interface


SHEAR STRESS, PSI
Figure 3-5 : Direct Shear Test Results Normal Stress
= 3.95 PSI On Soil/Teflon/Acrylic Interface


SHEAR STRESS. PSI
Direct Shear Test Results Normal Stress
* 7.9 PSI On Soil/Teflon/Acrylic Interface
Figure 3-6 :


SHEAR STRESS. PSI
23
Figure 3-7 : Direct Shear Test Results Normal Stress
15.8 PSI On Soil/Teflon/Acrylic Interface


SHEAR STRESS. PSI
24
DEFLECTION. INCHES
Figure 3-8 : Direct Shear Test Results Normal Stress
' 31.6 PSI On Soil/Teflon/Acrylic Interface


25
tests were completed even with a high confining pressure.
Direct shear tests for the friction angle on the
teflon/teflon interface was performed. The results of these
tests are shown in Figures 3-9, 3-10, 3-11, and 3-12. The
friction angle of the teflon/teflon interface was measured to be
approximately 9.5 degrees.
Direct shear tests for the friction angle on the
teflon/acrylic interface was also performed. The results of
these tests are shown in Figures 3-13, 3-14, 3-15, and 3-16.
The friction angle of the teflon/acrylic interface was measured
to be approximately 10.0 degrees.
The friction angle of the soil/teflon/acrylic interface
system was measured to be about one half that measured in the
teflon/teflon and teflon/acrylic interface tests. This can be
accounted for due to the type of teflon used in the
teflon/teflon and teflon/acrylic surface interface tests. The
teflon used for these tests was an industrial teflon that was
acid etched in the factory on one side and the other side was
not perfectly smooth. The teflon used in the
soil/teflon/acrylic interface system tests was 0.003" thick
teflon sheets with a very smooth surface.
The two stacked sheets of teflon were placed between
the acrylic and soil to eliminate side friction as much as
possible. Elimination of side friction would also simulate the
idealized plane strain condition and protect the acry]ic plastic


SHEAR STRESS. PSt
26
Figure 3-9 :
Interface Friction Angle For The
Teflon / Teflon Interface


SHEAR STRESS, PSI
27
Figure 3-10
Direct Shear Test Results Normal Stress
= 7.9 PSI On Teflon / Teflon Interface


SHEAR STRESS, PSI
28
Figure 3-11
Direct Shear Test Results Normal Stress
= 15.8 PSI On Teflon / Teflon Interface


SHEAR STRESS. PSl
29
Figure 3-12
Direct Shear Test Results Normal Stress
= 31.6 PSI On Teflon / Teflon Interface


SHEAR STRESS,
30
Figure 3-13 : Interface Friction Angle For The
Teflon / Acrylic Interface


SHEAR STRESS. PSI
31
Figure 3-14 : Direct Shear Test Results Normal Stress
=7.9 PSI On Teflon / Acrylic Interface


SHEAR STRESS. PSI
32
8.00
7.00
6.00
5.00
4.00
3.00
2.00
1.00
0.00
0.00 0.02 0.04 0.06 0.08 0.10 0.12 0.14
DEFLECTION, INCHES
Figure 3-15 : Direct Shear Test Results Normal Stress
= 15.8 PSI On Teflon / Acrylic Interface


SHEAR STRESS. PSI
33
Figure 3-16
Direct Shear Test Results Normal Stress
* 31.6 PSI On Teflon / Acrylic Interface


34
sides. See Figure 3-17 for a view depicting placement of the
teflon sheets.
See Appendix A for direct shear test data for the
direct shear tests performed on the soil/teflon/acrylic,
teflon/tefIon and teflon/acrylic materials.
3.2 Description of Load Frame and Hydraulic Jack
The loading frame for the test wall consisted of two 4"
diameter A36 steel posts, each bolted with four 3/4" diameter
high strength bolts to deadman connections in the laboratory
floor. Each dead-man mounted in the laboratory floor was
capable of providing more than 30,000 pounds of reaction force.
A reaction beam was bolted to the top of the posts
using four 3/4" diameter high strength bolts at each end of the
beam. The beam was an A36 W 12 X 19 steel beam 78" long with
1/2" by 6" plates 60" long welded full length along the top and
bottom of the beam flanges to provide the necessary stiffness
required for the experiment. A donut shaped ring of steel was
welded to the center of the bottom of the bottom flange of the
beam to provide a jacking point for the hydraulic ram that would
not allow slippage between the ram and beam during the test.
The loading was applied by using a hydraulic jack
capable of producing 60 kips of load. A load gage was mounted
on the jack and pump assembly with a dial graduated in 500 pound
increments. The hydraulic jack and gage was calibrated prior to
use in the test by means of a 120 kip capacity MTS. The gage


Figure 3-17
Installation of Teflon Sheeting


36
and the MTS load cell reading corresponded very well.
3.3 Description of Instrumentation
3.3.1 Deflection Indicators
For each load increment during the test, lateral and
vertical displacements of the test wall were obtained using dial
indicators. The dial indicators are graduated to 0.001"
increments and are capable of 1.000" measurement without being
reset. Each dial indicator was mounted onto a magnetic base
tool holder commonly used in machine shops. These magnetic base
tool holders were capable of holding the dial indicators at
specific points of interest throughout the test. A total of
seven dial indicators were used to obtain deflection data. Two
dial indicators were located at the top of the test wall to
measure the vertical deflection at the front and back of the
wall. Five dial indicators were used to measure the lateral
deflection at the center of each of the five geotextile layers.
3.3.2 Yarn Grid
A grid made up of yarn was placed against the interior
side of one wall prior to placement of fill in order to trace
the movements of the soil mass. The yarn was placed three
inches center to center in the vertical plane and three inches
center to center in the horizontal plane. The intersections of
the yarn lines were stapled together to prevent deformation of
the grid during placement of the fill. The grid was temporarily
held in place at the top and bottom and at the front and back


37
during construction by means of tape. The tape was removed
prior to the start of the test so that the grid would move
without restriction during the test.
The grid could be seen through the transparent acrylic
plastic after the construction of the test wall, even though
there was 0.006" of teflon sheeting on the interior of the
acrylic wall. The location of the grid was marked on the
exterior side of the acrylic wall by means of a water soluble
marker pen. These locations marked the intersections,
connecting lines and edge of the face of each of the five layers
at the start of the test. During the test, the locations of
select intersections were marked by a water soluble pen of a
different color and the locations were labeled as to the load in
pounds that was presently applied. At the end of the test, the
final locations of the intersections, connecting lines and the
layer faces were marked by yet another color. Figure 3-18 shows
the movement measurement grid installed along the interior wall
of the test apparatus. Figure 3-19 depicts the side of the test
apparatus showing the marks.
Figure 3-20, with Figure 3-20-A, Figure 3-20-B and 3-
20-C shows the movement of the soil mass as traced by the yarn
grid during the test. It can be seen that the movement occurred
at an angle of approximately 45 degrees to the horizontal.
Movement at the front of the test wall occurred at a constant
angle downward, while the movement towards the rear of the wall


38
Figure 3-18
Installation Of Movement Measurement Grid


39
Figure 3-19 : Measurement Grid Marks On Side Of Apparatus


40
Figure 3-20 : Internal Movement Of Soil Mass,
Reference Marks


SCALE
41
H"
1,380 PSF
980 PSF
+'
1,580 PSF
2,465 PSF

2,660 PSF
2,855 PSF
I
A
SCALE : 1" 1"
Figure 3-20-A
Internal Movement Of Soil Mass, Point A


SCALE
42

/
/
+
/ 1,580 PSF
/
*
2,140 PSF
2,855 PSF
SCALE : 1" 1"
Figure 3-20-B : Internal Movement Of Soil Mass, Point B


SCALE
43

£ 1,580
PSF
1,950 PSF

*
2,325 PSF
2,855 PSF
l
J
SCALE : 1" 1"
Figure 3-20-C : Internal Movement Of Soil Mass, Point C


kk
was more vertical at the beginning of the test and angled toward
horizontal as the test progressed.
3.3.3 Hydraulic Load Cells
Hydraulic load cells were invented and glued to the
exterior face of layers 3 and 5 for Test Wall 1. These cells
were capable of accurately measuring the actual pressure applied
to the cell face. The fabric for the layers involved was marked
and a small hole was cut in the center of the layer face to
provide a place to insert the load cells. As mentioned
previously, the Test Wall 1 reached a failure condition prior to
application of surcharge loads.
During the removal of Test Wall 1 from the test
apparatus, it was discovered that there was a rip in the fabric
face at the location of the hydraulic load cell in layer number
3. Due to the time and effort required to construct a test wall
and the potential for a tear in the fabric creating a failure,
it was decided not to attempt to mount the load cells in Test
Wall 2. However, a successful mounting of the load cells in
future tests is possible. This would provide for measurement of
the earth pressure at the face of the layer. This measurement
could then be compared to the pressure assumed to act at the
face due to the soil self weight and live loads at the top of
the wall.
Figure 3-21 shows the construction of the steel rimmed
version of the load cells. After the steel rimmed cells were


45
constructed and tested in soil, it was discovered that a
substantial portion of the load imparted into the sand was
carried by the steel rim. The solution to this problem was to
use a much thinner material as the casing for the load cell
bladder. A plastic tube was chosen as the casing material and
several more cells were manufactured and calibrated.
Figure 3-22 shows the method of hydraulic calibration
used for both versions of the load cells. Figure 3-23 shows the
equipment used to calibrate both versions of the cells in sand.
Figure 3-24 shows the plastic load cell attached to the fabric
layer face.


46
Figure 3-21
Construction Of Steel Casing Load Cells


47
(


48
Figure 3-23 : Equipment Used To Calibrate The Load Cells In
Sand


9
Figure 3-24 : Hydraulic Load Cell Attached To The Fabric Layer
Face


50
3.3.4 Strain Gages
Strain gages were used to measure the strain of the
geotextile during the test. The strain in the fabric was
expected to exceed 10%, therefore, gages were selected to
provide strain measurement up to 20%. The high elongation
strain gages were self-temperature-correcting, allowing for a
1/4 bridge connection. The gages were glued to the fabric and
calibrated at the room temperature. It should be noted that the
soil used in the test was brought into the lab for several weeks
and had equilibrated to room temperature prior to its use. This
was done to avoid large errors in strain measurement due to
effects of temperature on the strain gages.
The gages were soldered to strain gage leads that were
soldered to terminal strips. Three conductor, 22 gage, fully
shielded cable with shield ground was used as the
instrumentation cable for this project. The cable was soldered
to the terminal strip for a 3-wire, 1/4 bridge connection to a
Baldwin-Lima-Hamilton (BLH) direct readout device. The BLH
meter provided direct readings in micro-strain. The micro-
strain readings were then converted to strain (%).
Strain measurements at select points on the fabric for
layers 1, 2, 3, and 4 were obtained for each increment of load
during the test. The measurements were made by attaching the
high elongation strain gages to the fabric by means of a unique
method of attachment. The strain gages and the method of


51
attachment to the fabric were calibrated in a planar tensile
fabric pull test. Figures 3-25, 3-26, 3-27 and 3-28 depict the
standard method of attachment of strain gages to the fabric.
This method however, proved not to be reliable. A new method
was devised using the same glue down procedures shown in Figures
3-25, 3-26 and 3-27, except the glue that was used was a 2-ton
epoxy glue obtained in a hardware store. The epoxy glue was
mixed and applied only to the very ends of the strain gage.
This procedure allowed the strain gage to accurately record the
strain in the fabric.
Figure 3-29 shows a picture of a gage glued only at the
ends. Figure 3-30 shows a view of the fabric pieces used for
calibration. Figure 3-31 shows the strain gage calibration pull
test equipment. Figure 3-32 displays the calibration curve
developed by the planar tensile pull test.
Along with the strain gages mounted on the fabric,
measurement marks were drawn on the fabric with an indelible
marker prior to the installation of the fabric in the test
apparatus. These marks allowed for measurements of strain and
movement after the test was completed and the test wall
disassembled.


i
Figure 3-25
Strain Gages Prepared For Attachment By Standard
Glue Procedure To Geotextile Fabric
|
}


53
Figure 3-26 : Strain Gage Being Glued To Fabric Using Strain
Gage Glue In Accordance With Accepted Strain Gage
Attachment Procedure


54
Figure 3-27 : Strain Gages Being Glued Down Using Blocks And
Weights


55
Figure 3-28 : Typical Strain Gage Glued By Standard Glue
Procedure


56
Figure 3-29 : Strain Gage Glued Only At The Ends


57
Figure 3-30 : Fabric Pieces Used For Strain Gage Calibration


58
I
Figure 3-31 : Planar Pull Test Calibration Equipment
I


NIVH1S 30V0
59
Figure 3-32 : Calibration Curve For Strain Gage Glued At Ends


CHAPTER 4
MATERIAL PROPERTIES
In order to establish controlled test data for a
geotextile-reinforced earth wall, it is necessary to document
all relevant material properties. The material properties
reported herein include soil properties, geotextile properties,
interface properties between the soil and geotextile and
interface properties among the soil, teflon and acrylic plastic.
4.1 Soil Properties
The soil used in this study was a fine gravel and sand
soil commonly referred to as "Squeegee." Squeegee material was
selected for this study due to the fact the material density
does not change substantially over a range of placement
techniques. The grain size distribution curve is shown in
Figure 4-1. The material has 100% passing the 3/8" screen and
0.5% passing the # 200 screen. The Coefficient of Curvature,
Cc, is 1.25, and the Uniformity Coefficient, Cu, is 2.4. The
plasticity index (PI), is 0.0. The particles have a subangular
to subrounded shape, due in part to the fact the material was
obtained as pit-run and washed.
A Relative Density test was performed on the material
yielding a maximum dry unit weight of 108.10 lbs/ft3 and a


61
Figure 4-1
Gradation Curve For "Squeegee"
Material


62
minimum dry unit weight of 91.73 lbs/ft3 (ASTM D 4253, D 4254
and D 2049-69).
When placed into a bucket by raining down from a chute
at the bottom of a suspended bulk bag, the unit weight of the
material was approximately 103.5 lbs/ft3. a unit weight of
103.5 lbs/ft3 corresponds to a relative density of approximately
70%. Figure 4-2 shows the placement of the material into the
bucket for calibration of placement unit weight. Figure 4-3
shows placement of the material into the test wall from a bulk
bag.
The bulk density of the material as placed in the test
wall was lower than that of the calibration bucket. This is due
to the fact that the material was hand shoveled into place
within the test wall after it was rained down out of the bulk
bag. An in-place bulk density test of the soil was not
performed, but it is estimated that the in-place bulk density
was approximately 95 lbs/ft3. A more accurate measurement of
the unit weight as placed will be determined upon disassembly of
the test wall.
An isotropic compression drained triaxial test (CID)
was performed on the squeegee material. The test was run at a
density close to 102.4 lbs/ft3 corresponding to a relative
density of 65%. The triaxial test was run on a 6" diameter by
12" tall sample. The sample was saturated to a Skempton "B"
parameter of 0.96. The sample was then sheared using a computer


63
i
t
i
i
i
!
i
i
Figure 4-2
Calibration Of Squeegee Placement Density Raining
Down Into A Bucket Of Known Volume


64
Figure 4-3
Placement Of Squeegee Into Test Wall


65
controlled MTS machine in the strain control mode. The sample
was sheared at 10, 20 and AO psi confining pressures using a
multi-stage test procedure. The cohesion intercept was 0.00 and
the phi angle at the lower confining pressures was 39. See
Appendix C for the triaxial test results.
4.2 Geotextile Fabric Properties
The geotextile fabric used for this study was TYPAR
3301, as manufactured by Reemay Incorporated. TYPAR 3301 is a
spunbonded nonwoven polypropylene geotextile fabric. The
primary uses for the fabric include erosion control, protection,
separation, drainage, filtration, stabilization, pipe wrap,
composite cover, and landscaping. The fabric was chosen because
it has a low enough strength to allow for failure to be reached
at a relatively low surcharge load.
A standard roll 36" wide, 100 feet long was used for
this study. The TYPAR 3301 Mass Per Unit Area is 2.9 oz/yd2.
The apparent opening is between the # 60 and # 70 U.S. sieve
screen. The Grab Tensile/Elongation by ASTM D4632-86 is 120
pounds/inch at 60% elongation and the Wide Width Strip
Tensile/Elongation by ASTM D4595-86 is 35 pounds/inch at 60%.
This corresponds to 420 pounds/foot at 60% elongation.
The Wide Width strength value was used in the design of
Test Wall 1 and 2. It is interesting to note that the Wide
Width Strip Tensile values are the same for both machine
direction and cross machine direction. The fabric was placed


66
into the test wall in the machine direction.
4.3 Soil-Fabric Interface Properties
The fabric / soil friction angle as determined from a
pullout test is approximately 27 degrees. Figure 4-4 displays
the curve developed during the fabric / soil interface pullout
test. Test data are presented in Appendix D.
4.4 Acrylic Plastic
For this study, smooth walls were required to eliminate
side wall friction as much as possible. Along with smooth
sides, the walls had to be of sufficient strength to withstand
the high loadings without appreciable deflection. This was
necessary to properly simulate the idealized plane strain
condition. To accomplish this, three 1/2" thick 4' by 8' cast
acrylic plastic sheets were used. The acrylic has a compressive
strength between 11,000 psi and 19,000 psi, a tensile strength
between 8,000 psi and 11,000 psi, and a flexural bending stress
between 12,000 and 17,000 psi.
The lower values of strength correspond to extruded
acrylic while the upper values correspond to cast acrylic. The
1/2" thick acrylic sheets used were probably cast, however this
could not be confirmed by the manufacturers representative,
therefore the lower values were used for strength parameters in
the design.
4.5 Teflon Sheets
Teflon sheeting was used to protect the acrylic sheets


PULL OUT FORCE. LBS/UNEAL INCH
67
Figure 4-4 : Fabric Pull Out Test,
Typar 3301


68
and to provide an even smoother surface of the side walls. The
teflon sheets were 0.003" thick by 12" wide and were stacked in
double thickness with overlapping edges against the interior
walls of the test apparatus in a vertical orientation. Figures
3-4 through 3-16 illustrate the results of direst shear tests of
soil / teflon / acrylic and results of teflon and acrylic in
different configurations.


CHAPTER 5
TEST WALL CONSTRUCTION
The test wall was constructed using a temporary support
by the same procedure as that adopted by the Colorado Highway
Department construction method (Derakhshandeh and Barrett,
1987). After the test wall construction was completed, the
surcharge loading was initiated within hours in order to
minimize creep effects.
5.1 Test Box Preparation
The box used in this study was constructed of steel and
timber with acrylic walls facing the interior of the apparatus.
The steel box apparatus was placed on top of one 3/8" thick by
4' by 8' piece of plywood. The weight of the apparatus itself
was the only anchor to the concrete floor.
Two teflon sheets, 0.003" thick by 12" wide were
stacked one on top of the other for a total thickness of 0.006".
These teflon sheets were held in place at the top and bottom by
using tape. The edges of the teflon sheets were held together
by tape in several places.
A yarn grid with a spacing of 3" by 3" was placed
against one interior side of the apparatus. The grid work was
held in place temporarily during construction by tape at the


70
top, bottom and each end. The intersections of the yarn grid
work were stapled together using a small stapler.
On the floor of the steel box was exposed plywood.
Glue was spread on the plywood using a tile mastic trowel and
Squeegee material was dropped onto the glued surface by the
handful. The glued squeegee was placed in the area of the
bottom of the fabric for layer 5. After the glued Squeegee had
dried, a small amount of loose Squeegee was spread evenly over
the glued Squeegee. This was accomplished to more closely
simulate placing a layer of fabric onto a hard soil surface in
the field. See Figure 5-1 for a view of the test apparatus with
the glued squeegee.
5.2 Fabric Preparation
The TYPAR 3301 fabric was used in this study, the
fabric was cut from a standard 36" wide roll to have a width of
31" in the exposed area of the wall face, and cut to 30" wide in
the embedded areas of the fabric. The fabric was marked with an
indelible marker across the width of the fabric. Marks were
placed at the points of interest such as the top and bottom of
the wall face, and the points where the strain gages were to be
glued. See Figure 5-2 for a view of the marked fabric pieces.
The test wall consisted of a total of five layers of geotextile.
The length and geometry are illustrated in Figure 2-1.
5.3 Strain Gage Attachment
One of the most difficult portions of this study was to


71
Figure 5-1
Test Apparatus With Squeegee Glued To The Floor


72


73
determine a method of accurate measurement of fabric strain. To
accomplish this, strain gages were glued to the fabric. The
method of attachment to fabric is critical to the accuracy.
Two methods of strain gage attachment were attempted,
however, only one method proved viable. Both methods consisted
of soldering the leads to the strain gage and then soldering the
leads to a terminal strip. Three conductor, 22 gage, fully
shielded cable with ground was used as the instrumentation cable
in this study. The instrumentation cable was soldered to the
terminal strip for a 3-wire, 1/4-bridge connection. The strain
gage, leads and terminal strip with cable was then considered
one complete assembly. The strain gage was then attached to the
fabric. Figure 2-1 shows the locations where strain gages were
placed on the geotextile layers.
Strain gage readout was accomplished by using a direct
readout meter. A Baldwin-I.ima-Hamilton (BLH) 1200 meter with
two BLH 1225 multiple strain gage switching and balancing
stations were hooked up in unison to provide readout on twenty
strain gages.
5.4 Temporary Support During Construction
Temporary supports were built and used during the
construction of the test wall. In Test Wall 1, the temporary
support allowed for the construction of the entire wall prior to
removal of the support. This proved not to be an effective
means of construction. When the temporary support was removed,


74
the top layer instantly failed in pullout mode. This pullout
could have been foreseen at a lower layer had the supports been
made to be removed after each layer was constructed. Also,
removal of the supports after the entire construction is
completed is not commensurate with accepted field construction
procedures. See Figure 5-3 for a view of the failed test wall.
The construction procedure commonly used in practice
was used on Test Wall 2. For this construction procedure, a
temporary support was be removed after each layer was
constructed. This method was used with excellent results. The
support for the bottom layer 5, was removed when layer 4 was
completed. The support for layer 4 was removed when layer 3 was
completed, with repetition until all layers were completed.
Upon completion of the top layer number 1, the support was left
in place until a seating load of 2,000 pounds corresponding to a
surcharge of 155 psf was applied.
5.5 Backfilling
Backfilling each layer was accomplished by means of
raining Squeegee material down from a bag suspended from a
crane. This method was adopted to ensure a uniform placement of
materials in each layer. The method consisted of using a Bulk
Bag with a downspout sewed onto the bottom. The downspout was
8" in diameter and 3' long. Each bag was suspended from the
overhead rail crane by means of a steel frame and chain. The
distance of drop versus bulk density was determined by raining


75
Figure 5-3
View Of Test Wall After Failure in Pull Out In
Layer Number 1


76
material into a bucket of known volume beneath the suspended
bag.
Each bag had a capacity of 2,500 pounds and was
suspended over the top of the Test Wall. Material was released
from the bag, placed using a hand shovel to the proper elevation
and leveled using a straight board. Level was checked using the
straight board and a Torpedo level.
Because the material had to hand shoveled into place,
variable placement density was possible. Placement density was
estimated to be 95 pcf.
It should be noted that when a small amount of material
was placed onto the bottom portion of the fabric layer, the tail
of the fabric was pulled tight to remove any wrinkles. This
allowed for strain of the fabric immediately after load
application. In this manner, fabric strain could more
accurately be measured. See Figure 5-4 for a view of a layer
partially filled with backfill.
5.6 Overlapping
The overlap portion of the fabric was embedded into the
top of the fill material by smoothing the overlap horizontally
over the top of the fill material for 6", and then embedding the
overlap at a shallow angle for a distance of l'-0". The total
overlap length was 18" for layers 2, 3, 4 and 5. For the top
layer 1, the overlap was not buried but was spread over the top
of the entire layer surface to the back of the test apparatus.


77
Figure 5-4 : Layer 2 Under Construction


78
The first 6" of the fabric was placed onto the level
fill at the elevation of the top of the layer to provide a seal
against the layer above. Without this seal, the Squeegee
material would have a chance of dribbling out during the test,
and would cause measurement errors due to uneven load transfer.
See Figure 5-5 for an elevation view of the test wall sample as
constructed.
5.7 Test Wall Geometry
The five layer test wall geometry can be seen in Figure
5-5. This figure was drafted to represent the geometry of the
wall immediately after the wall was constructed. The figure
depicts as accurately as possible, the bulges of each layer
face, as well as the dip of the top layer near the front. Also
shown is loading plate system used for this study. Scales are
provided to accurately depict the length and height of the test
apparatus and layer geometry. Locations where strain gages were
mounted are shown along with the corresponding strain gage
numbers. Layers are numbered from number 1 at the top to number
5 at the bottom.


ELEVATION.
79
r
5 -
4 -
£
3 -
2 -
1 -
0 L
Figure 5-5
i_i_i_i_i_i_i-1-1
012345678
SCALE IN FEET
: Test Wall Geometry As Constructed


80
5.8 Surcharge Load Application
The load was applied as uniformly as possible by means
of a hydraulic jack pushing between a reaction beam and loading
plates. The reaction beam was bolted to two reaction tension
posts that straddled the test wall apparatus. The two posts
were bolted to two dead men anchors installed in the laboratory
floor. The reaction beam was a W 12 X 19 with plate stiffeners
at the middle and the ends. Plates 1/2" X 6" X 60" were welded
on the top and on the bottom of the beam flanges providing for a
center point load of more than 80 kips. Since a maximum center
point load of 60 kips was expected, this beam was determined to
be strong enough.
I.oading plates atop the test wall sample consisted of
two 3/4" X 29" X 78" pieces of high density particle board
bolted together, one 1/2" X 29" X 48" plate, two 1/2" X 29" X
18" plates, and one 1" X 29" X 18" plate. One-inch square stock
was used to provide a simple support configuration used to
distribute the load as evenly as possible given the maximum
height of the loading plate constraint of 6". See Figure 5-6
for an elevation view of the loading plate configuration.
The hydraulic jack was a 60 kip ram with a hydraulic
jacking mechanism and load gage located at a remote location
from the ram by means of a quick disconnect hose. The jack
system was calibrated just prior to this study and was found to
be in good working order with accuracy within expected limits.


81
The hydraulic ram was placed on a "ball and socket" plumb
compensator device, used to keep the jack ram plumb under a
variety of loading conditions.
The test wall was loaded using 2,000 pound increments.
An initial load of 500 pounds was applied to seat the loading
plates. The load was brought to 2,000 pounds and the layer
number 1 temporary face was removed. The load was increased by
2.000 pounds thereafter until a load of 20,000 was reached. The
load increment was increased from 2,000 pounds to 3,000 pounds.
The loading increment was 3,000 pounds from 20,000 pounds to
44.000 pounds when the test was stopped due to hydraulic jack
leakage.


CHAPTER 6
TEST RESULTS AND DISCUSSION OF RESULTS
In this study, a five layer geotextile-reinforced earth
wall was constructed within a test apparatus and loaded
incrementally to a total of 2,854 psf surcharge over the top of
the wall. Placement unit weight of the granular backfill was
estimated to be approximately 95 pcf. Strain gage, lateral
deflection and vertical deflection data were obtained for every
load increment. Presented within this chapter are results
obtained during the load test. The test data presented in this
chapter are shown in Appendix E.
6.1 Strains and Deflections
Figures 6-1 through 6-5 show composite plots of strain
gage, lateral deflection and vertical deflection readings
obtained at five different surcharge loads.
Figure 6-1 (at a load surcharge of 295 psf) shows small
strains in all of the instrumented layers. Other than the top
layer, all of the other layers exhibited negligible strains.
I.ateral deflections measured at the center of each layer shows
that the largest deflection was in the second layer. Subsequent
figures show this to be a trend throughout the test.
Figure 6-2 (at a load surcharge of 850 psf) shows


LATERAL FACE DEFLECTION
83
VERTICAL DEFLECTION
13 0 r .
* 2
s *
DEFLECTION. INCHES DISTANCE. INCHES
Figure 6-1 : Test Wall Response At A Surcharge
Pressure = 295 PSF


LATERAL FACE DEFLECTION
84
VERTICAL DEFLECTION
i 2---------------------------------------------
G
l§
DEFLECTION. INCHES DISTANCE. INCHES
Figure 6-2 : Test Wall Response At A Surcharge
Pressure = 850 PSF


85
larger strains recorded for each layer. This load increment
corresponds to the failure load predicted by the Forest Service
design method.
Note that the top layer shows the maximum amount of
strain. Note also that strains are larger toward the rear of
layer 4 than at the front. Subsequent figures will show a
reversal of this trend. Lateral deflections of the layers are
more pronounced with a significant deflection developing in
layer 2. Note the larger vertical deflection at the front of
the test wall. This trend will become more pronounced with
additional loading.
It is to be noted that there was an increase in loaded
area as each load increment was applied, resulting from
settlement of the sand backfill and lateral deflection of the
wall face which closed the gap at the top front of the loading
plates. The reported surcharge pressures were corrected based
on the actual contact area at the given load increment.
Closing the gap also induced tilting of the loading
plates. The ball and socket compensator at the top of the
loading plates, however, allowed for subsequent surcharge loads
to be applied perpendicular to the loading plates even after the
plates were dipping toward the front of the test wall.
Figure 6-3 (at a load surcharge of 1380 psf) shows a
trend developing in the recorded strains for each layer. Layer
1 still shows the largest amount of strain. The locations where


LATERAL FACE DEFLECTION
86
VERTICAL DEFLECTION
DEFLECTION, INCHES
0
2
4
*-
DISTANCE. INCHES
Figure 6-3 : Test Wall Response At A Surcharge
Pressure = 1380 PSF


87
maximum strain occurred in a layer moved toward the front of the
wall in layer 1, layer 2 and layer 3. The observation that the
strains in layer 2 were smaller than layer 3 below, is believed
to be associated with the large lateral deflection of layer 2.
Layer 4 still shows slightly larger strains at the rear of the
layer, although the strain at the front of the layer has
increased. Lateral deflections are more pronounced with the
maximum deflection in layer 2. Vertical deflection at the front
has also shown an increase.
As may be expected, the strains and deflections
illustrated in Figure 6-4 (at a load surcharge of 1950 psf) are
larger than those of Figure 6-3. The same trend in the strain
distribution and deflection patterns as in Figure 6-3 was
observed. Layer 4 shows a nearly uniform strain along its
length. Lateral and vertical deflections have become more
pronounced than the previous load increment. The shapes of
their distributions, however, remain unchanged.
Figure 6-5 (at a load surcharge of 2660 psf) presents
the recorded strains and deflections just prior to the cessation
of the load test. Again, the same trends observed in the
previous load increments were observed. Notice the strain
recorded for gage 16 jumped suddenly from about 8% to about 18%.
It is not known exactly what caused this sudden jump in strain.
Note also that gage number 6 did not record any further strains
beyond the previous load increment. The lateral and vertical


LATERAL FACE DEFLECTION
88
Figure 6-4
Test Wall Response At A Surcharge
Pressure = 1950 PSF