Citation
Feasibility of geosynthetic inclusion for reducing swelling of expansive soils

Material Information

Title:
Feasibility of geosynthetic inclusion for reducing swelling of expansive soils
Creator:
Vessely, Mark J
Publication Date:
Language:
English
Physical Description:
65 leaves : illustrations ; 28 cm

Subjects

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

Notes

Bibliography:
Includes bibliographical references (leaves 64-65).
General Note:
Department of Civil Engineering
Statement of Responsibility:
by Mark J. Vessely.

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:
49665218 ( OCLC )
ocm49665218
Classification:
LD1190.E53 2001m .V47 ( lcc )

Full Text
FEASIBILITY OF GEOSYNTHETIC INCLUSION
FOR REDUCING SWELLING OF EXPANSIVE SOILS
by
Mark J. Vessely
B.S., Colorado School of Mines, 1994
A thesis submitted to the
University of Colorado at Denver
in partial fulfillment
of the requirements for the degree of
Master of Science
Civil Engineering
2001


This thesis for the Master of Science
degree by
Mark J. Vessely
has been approved
by
QrToL r.r 2$2Pc> \
Date


Mark J. Vessely (M.S., Civil Engineering)
Feasibility of Geosynthetic Inclusion for Reducing Swelling of Expansive Soils
Thesis directed by Professor Jonathan T. H. Wu
ABSTRACT
A study was undertaken to examine the feasibility of geosynthetic inclusion for
reducing swelling of expansive soils. The study was conducted by performing
laboratory soil-geosynthetic swell tests on an expansive soil. The test specimen
measures 12 inches by 12 inches by 12 inches, with a sheet of geosynthetic
embedded horizontally at the mid-height of the soil. To prepare the test specimen,
the soil was first compacted, in 1 in. lifts, inside a wooden mold to the prescribed
density and moisture content. The soil was then allowed to swell subject to wetting
by soil suction The vertical and lateral deformations of the specimen were
monitored throughout the test. To assess the effect of geosynthetic inclusion, a test
without geosvnthetic inclusion was performed in otherwise identical conditions for
comparison purposes. This paper describes the test method and presents the test
results. Based on the test results, the feasibility of geosynthetic inclusion for
reducing swelling of expansive soils in practical applications is addressed.
This abstract accurately represents the content of the candidates thesis. I
recommend its publication.
in


CONTENTS
Figures.....................................................................vi
Tables......................................................................vii
Chapter
1. Introduction..............................................................1
1.1 Problem Statement.........................................................1
1.2 Objective.................................................................2
1.3 Scope of Study............................................................3
1.4 Engineering Significance..................................................3
2. Literature Review.........................................................5
2.1 Introduction..............................................................5
2.2 Expansive Soils..........................................................5
2.3 Previously Documented Applications of Reinforcement in Expansive Soils...7
2.3.1 Discrete and Random Fiber Reinforcement.................................7
2.3.2 Driven and Embedded Steel Reinforcement...............................8
2.4 Discussions Regarding Lateral Swell in Expansive Soils...................9
2.4.1 Studies Performed Using Non-Standard Research Methods...................9
2.4.2 Studies Performed using Standard Research Methods.....................19
2.4.3 Conclusions About Lateral Swell Obtained from Literature Review.......21
3. Test Program.............................................................24
3.1 General..................................................................24
3.2 Soil Properties.........................................................24
3.3 Geosynthetic Properties.................................................33
3.4 Sample Preparation......................................................33
3.5 Test Procedure..........................................................37
iv


4. Test Results and Discussion............................................41
4.1 General................................................................41
4.2 Swell Without Geosynthetic Inclusion...................................41
4.3 Swell With Geosynthetic Inclusion......................................43
4.3.1 Test GRS-1...........................................................46
4.3.2 Test GRS-2...........................................................46
4.4 Summary and Discussion of Test Results...............................51
4.5 Review of the Initial and Final Specimen Moisture and Densities.......54
5. Summary, Conclusions, and Recommendations for Future Research..........60
5.1 Summary................................................................60
5.2 Conclusions............................................................61
5.3 Recommendations for Future Research....................................62
References.................................................................64
v


FIGURES
Figure
2.1 Extent of Montmorillonite in Bedrock Exposures (From Chen, 1988).........6
2.2 Lateral Pressure with Time...............................................12
2.3 Relationship Between Surcharge Pressure and Equilibrium Lateral Pressure.14
2.4 Variation Between Surcharge Pressure and Ratio
of Lateral Pressure and Vertical Stress................................15
2.5 Lateral Pressure Distribution During Release of Surcharge Pressure.....16
3.1 Vicinity Map...........................................................25
3.2 Logs of Exploratory Borings............................................27
3.3 Gradation Analysis.....................................................28
3.4 Proctor Curve..........................................................29
3.5 Swell Percent and Pressure Versus Moisture Content for
Remolded Clay Samples..................................................30
3.6 Swell Percent and Pressure Versus Proctor Density for
Remolded Clay Samples..................................................31
3.8 Typical Specimen Setup.................................................38
3.9 Rate of Water Addition for Test Specimen...............................40
4.1 Specimen NO GRS Percent Vertical Swell with Time.......................42
4.2 Specimen NO GRS Percent Lateral Swell with Time........................44
4.3 Photographs of Specimen NO GRS at the Conclusion of Testing............45
4.4 Specimen GRS-1 Percent Vertical Swell with Time.......................47
4.5 Specimen GRS-1 Percent Lateral Swell with Time........................48
4.6 Photographs of Specimen GRS-1 at Conclusion of Testing................49
4.7 Specimen GRS-2 Percent Vertical Swell with Time.......................50
4.8 Specimen GRS-2 Percent Lateral Swell with Time........................52
vi


TABLES
Table
3.1 Results of Laboratory Test Data.......................................32
3.2 Some Index Properties of the Geotextile...............................34
4.1 Summary of Total Swelling Measurements................................53
4.2 Initial and Final Moisture Content Determination Worksheet
Specimen NO GRS........................................................55
4.3 Initial and Final Moisture Content Determination Worksheet
Specimen GRS-1.........................................................56
4.4 Initial and Final Moisture Content Determination Worksheet
Specimen GRS-2.........................................................57
4.5 Summary of Initial and Final Moisture Contents.........................58


1. Introduction
1.1 Problem Statement
Expansive soils problems are an international problem in many arid and
semi-arid regions where civil engineering works are constructed. Although
expansive soil problems have not been widely appreciated outside the area of their
occurrence, the damage due to the problem is alarming. It has been estimated that
the damage to buildings, roads, and other structures founded on expansive soils
exceeds two billion dollars annually (Chen, 1988). Furthermore, expansive soil
damage can be recurring where soil moisture conditions change throughout a year.
A number of methods are used to reduce the effect of expansive soil on
engineered structures. Commonly accepted methods include compaction control,
drainage control, chemical stabilization (with lime, cement, etc.), excavation and
replacement, pre-wetting, moisture barriers, and structural design (e.g., drilled piers).
These measures have varying degrees of success, yet the amount of damage
associated with expansive soils continues to increase.
Geosynthetics, such as geotextiles and geogrids, have been used successfully
as reinforcement in the construction of retaining walls, slopes, embankments,
roadways, and foundations (e.g., Koemer, 1997; Holtz, et al., 1997). This type of
construction is commonly referred to as Geosynthetic Reinforced Soil (GRS)
1


structures. The primary function of geosynthetic reinforcement in a GRS structure is
to restrain lateral movement of the soil, resulting in a stronger and more stable soil
mass (Wu, 1994).
The fact that a geosynthetic inclusion can effectively restrain deformation of
a soil mass through soil-geosynthetic interface adhesion plus the fact that a
geosynthetic inclusion can significantly facilitate drainage of a clayey soil make
geosynthetic inclusion a promising technique for reducing swell potential of an
expansive soil.
1.2 Objective
The scope of this study is to examine the feasibility of using geosynthetic
inclusion to reduce swell potential of expansive soils. The study was conducted by
performing laboratory tests on an expansive soil obtained from an actual
construction site where the material was used as fill. The test specimens were
prepared using special techniques to ensure uniformity within the compacted soil. A
sheet of non-woven geotextile was placed horizontally within a cubic soil mass at
the sample mid-height. The vertical and lateral deformations of the test specimen
during wetting were monitored throughout the test. To assess the effect of geosynthetic
inclusion, a test without geosynthetic inclusion was performed in otherwise identical
conditions for comparison purposes. Conclusions regarding the effect of geosynthetic
inclusion are developed based on the results of the laboratory program.
2


1.3 Scope of Study
Existing test methods are not available for evaluating swelling of a soil-
geosynthetic composite. As part of the scope of study, a new test method was
developed to examine the feasibility of geosynthetic inclusion for reducing swelling
of an expansive soil. An expansive soil was obtained from a local construction site
and index testing was performed to classify the material. After completing
preparation of the test specimens, the soil was wetted using soil suction, and the
vertical and lateral deformations of the specimen were monitored until the
deformations became negligible. Two tests were performed with a nonwoven
needle-punched geotextile placed at the mid-height of the specimen. The third test
was performed without geosynthetic inclusion.
1.4 Engineering Significance
The engineering significance that may be developed from this research thesis
includes the following.
-To determine if the test method developed in this study can be a viable and simple
method for evaluating swelling potential of soil-geosynthetic composites.
-To evaluate the effect of horizontal geosynthetic inclusion on swelling in the
horizontal and vertical directions.
3


- To determine if the use of geosynthetic inclusion may be a practical technique in
situations where reducing lateral swelling is of major concern, for example, a
basement wall.
4


2. Literature Review
2.1 Introduction
The literature review consisted of three parts. The first topic included a brief
review of expansive soil problems. Secondly, a review was performed for other
applications of reinforcement in expansive soil. The third item involved a review of
lateral swelling in expansive soil.
2.2 Expansive Soils
Expansive soils are a relatively new problem in soil mechanics and were not
recognized by geotechnical engineers until the 1930s. Expansive soils have become
a worldwide problem and the damage related to heaving soil is increasing as
development into arid and semi-arid regions continues. Expansive soil damage in
1970 was approximately 1.9 billion dollars and is estimated to increase to just over
4.5 billion dollars in 2000 (Chen, 1988).
Expansive soils result from a variety of geologic conditions that form the
clay mineral. Clay minerals such as montmorillonite and smectite have a structure
that can swell due to unsatisfied water absorption forces on specific surface areas.
The extent of montmorillonite in bedrock exposures throughout the United States is
presented in Figure 2.1. In the United States, the most severe expansive soils are
observed in Colorado, Texas, and Wyoming.
5


Figure 2.1 Extent of Montmorillonite in Bedrock Exposures (From Chen, 1988)
6


Approximately one-half of the expansive soil damage expenses are related to
transportation structures such as asphalt and concrete pavements. Buildings on
expansive soils utilize deep foundations and high dead load pressures to mitigate the
heave potential. However, pavement structures do not have sufficient dead load
pressure to resist heave. The pavement structures must span miles of expansive soils
and avoidance is not a possible remedy. Therefore, common mitigation involves
reducing the swelling potential of the clay through physical methods, such as pre-
wetting, compaction control, replacement, or chemical stabilization of the clay
mineral with cement, lime, fly ash, or electrochemical methods.
2.3 Previously Documented Applications of Reinforcement in Expansive Soils
A limited amount of published research is available regarding placing
reinforcement in an expansive soil to reduce the effects of swelling. The majority of
expansive soil treatment methods involve chemical additives, moisture control, or
excavation and replacement under controlled moisture and density conditions.
These methods have various disadvantages and do not necessary reduce expansive
soil movements to within tolerable limits. This research thesis examines the effect
of mechanical reinforcement.
2.3.1 Discrete and Random Fiber Reinforcement
A study performed by Puppala and Musenda (2000) examines the effect of
random oriented fiber reinforcement on two remolded expansive soils from
7


metropolitan areas in Texas. The reinforcement consisted of discrete and randomly
oriented polypropylene fibers. The effect of different quantities of 25- and 50-
millimeter fiber lengths was examined in the study. The testing program compared
the unconfined compressive strength, shrinkage, free swell, and swelling pressure for
reinforced and unreinforced samples.
With respect to expansive soil behavior, the fibers reduced the measured
vertical swelling pressures in the samples. As the quantity of fiber reinforcement
increased, the free vertical swell did increase. The authors speculate the fiber
reinforcement may have created preferential pathways for water movement within
the sample, and the greater distribution of water increased the free swell. The
reasons for a decrease in swelling pressure with fiber reinforcement are suggested to
include: the fiber reinforcement provides drainage for dissipation of pore pressures
and the fibers provide a tensile force within the sample to restrain swell pressures.
2.3.2 Driven and Embedded Steel Reinforcement
The research performed by Srinivasa Murthy et. al. (1987) examined the
effect of vertically oriented steel reinforcement on the swelling pressure of a
remolded expansive soil. The study utilized test specimens constructed in a
California Bearing Ratio test mold and a 1.2-meter cubic mold. Steel reinforcement
bars were driven or embedded into the clay soil. The embedded reinforcement was
placed in a bored hole filled with compacted sand. The reinforcement was rigidly
8


attached to a top plate and the specimens were submerged in water. The test results
were compared to specimens that were tested without reinforcement inclusion.
The test results indicate there is a reduction in the measured heave in the
direction parallel to the reinforcement. The measured heave also decreased
proportionally as the reinforcement length is increased. The specimens with
reinforcement surrounded by compacted sand or roughened steel indicated the
greatest reduction in heave. The authors suggest the horizontal swelling component
mobilizes a frictional force around the reinforcement that counters heave in the
direction parallel to reinforcement.
2.4 Discussions Regarding Lateral Swell in Expansive Soils
Much of the research for expansive soils discusses vertical deformation
rather than the lateral or volumetric movement. The following discussion reviews
previously published studies regarding volumetric or lateral swell in expansive soils.
2.4.1 Studies Performed Using Non-Standard Research Methods
Al-Amoud (1993) performed a study to examine the relationship between
lateral and vertical expansive stresses in an expansive soil. The research program
developed a testing box with a volume of 0.2 ft3. The box utilized hinged sides and
tops to measure the lateral and vertical swelling pressures. The stress measurements
were determined using the resulting forces calculated by summing the moments
associated with the hinges. The testing program evaluated the stress relationship for
9


a fill comprised of a local expansive soil and compacted at varying moisture and
density conditions.
The test results suggest the lateral and vertical swelling stresses increase
rapidly with time and then the rate decreases to an equilibrium stress. The clay fill
indicated a higher lateral swelling stress when compared to the vertical stress. The
ratio between lateral and vertical swelling stresses increased as the water content and
dry density increase. The measured swelling stress ratio (lateral/vertical) for the clay
soil ranged from 1.21 to 3.07.
Chen (1987) published the results of study to determine lateral expansion
pressure on basement walls. A model test was developed to measure the lateral
expansion pressure on a rigid wall. The test setup consisted of 15-inch by 5-inch
plan dimension and was 12 inches in height. A surcharge pressure of 800 psf was
applied and a constant volume was maintained. Load cells were used to measure the
lateral expansion pressures at the bottom, middle, and top of the specimen. The
vertical expansion pressure also was measured with a separate load cell.
The test results indicate lateral pressure increases during moisture infiltration
and then decreases towards an equilibrium value. The equilibrium lateral swell
pressure was 12 pounds per square inch (psi) (1700 psf). The vertical swell pressure
steadily increased towards an equilibrium swell pressure of near 12 psi.
If the at rest (ah) pressure is calculated for the specimen under an 800 psf
surcharge, the value is approximately 680 psf. The highest peak lateral pressure
10


during infiltration was approximately 34 psi (4900 psf) and was observed in the
middle of the specimen. The peak lateral pressures measured in the top and bottom
of the specimen were approximately 22 and 32 psi (3200 and 4600 psf), respectively.
The ratio of the lateral to vertical swelling pressures ranged from 3.6 during
infiltration and approached 1 as the equilibrium pressures were obtained.
Joshi and Katti (1980) used specially constructed equipment to study lateral
pressure development under surcharges. The study utilized large (3x4x9 feet) and
small (1 x 1 x 1.5 foot) scale containers to measure the lateral swell pressures from
an expansive black cotton soil during the saturation process. The tests were
completed with different surcharge pressures. Some tests also were performed with
a circular container to evaluate the influence of shape; however, the results indicated
insignificant edge effect.
The large scale testing indicated the lateral pressure increased to 6,000 PSF
(287 kN/m ) at a 3-foot depth. Below 3 feet, the lateral pressure measurements did
not increase, which suggests the lateral pressure distribution in an expansive soil is
different than that of a conventional soil. Additionally, the ratio between lateral
pressure and vertical stress at a depth of 3 feet was observed to be about 10 to 12.
The rate of lateral pressure development for all tests was observed to increase
rapidly when water was added. The rate decreases as the peak lateral pressure is
obtained and then decreases to a constant pressure. A graph from the paper
indicating lateral pressure with time is presented on Figure 2.2. The time
11


Figure 2.2 Lateral Pressure with Time
12


required for a constant pressure to develop appears to be dependent on surcharge
pressures. Constant lateral pressures were obtained more rapidly with lower
surcharge pressures.
Additionally, the relative lateral pressure magnitude decreases with
increasing surcharge pressures as shown on Figure 2.3. This type of behavior is
similar to that of a conventional soil. Similarly, the ratio between lateral pressures
and vertical stress decreases significantly as the surcharge pressure increases. This
relationship is illustrated in Figure 2.4.
The study also examined the lateral pressures that occur as the surcharge is
removed from the expansive soil. When the surcharge pressure is below the
swelling pressure range, the observed lateral pressures during surcharge pressure
relaxation increased above the equilibrium lateral pressure observed with the
original surcharge pressure. This relationship is illustrated in Figure 2.5.
Ofer (1980) developed instrumentation for laboratory and in situ
measurement of the lateral swelling pressure. Standard odometer rings modified
with strain gauges were used to measure lateral swelling pressure in the laboratory
and an in-situ lateral swelling pressure probe was developed to measure conditions
in the field. The study indicates that as the density of a sample increases, heave and
lateral swelling pressure also increase when water is introduced. The lateral
swelling pressure also appeared to decrease with time when measured with the field
probe.
13


INtTlAl SURCHARGE;^
Figure 2.3 Relationship Between Surcharge Pressure and Equilibrium Lateral
Pressure
14


INITIAL SURCHARGE 1S f
Figure 2.4 Variation Between Surcharge Pressure and Ratio of Lateral Pressure and
Vertical Stress
15


Figure 2.5 Lateral Pressure Distribution During Release of Surcharge Pressure
16


A research thesis by Phillips (1981) studied the lateral swelling pressure of
an expansive Colorado soil. The test program involved performing a series of lateral
and vertical swelling tests. For each of the tests, the strain was restrained in the
direction opposite to the swelling direction. Vertical swelling pressures were
obtained using a standard oedometer. Lateral swelling was performed using a
modified pressure vessel with a water-pressurized membrane.
The measured maximum lateral swelling pressures ranged from 2,088 to
3,816 psf and averaged 2,632 psf for ten tests. The maximum lateral swelling
pressures were developed during the first 30 minutes after conclusion of wetting.
The lateral swelling pressures remained at the maximum value for approximately 2
hours and then decreased to approximately 25 to 35 percent of the maximum value
at the conclusion of the test. The ratio of average lateral swelling pressure to
average vertical swelling pressure was approximately 0.60.
The author also suggests possible explanations why lateral swelling pressures
decrease from a peak value during testing. One offered explanation discusses the
creep behavior of materials subjected to continuous and uniform loads. This
principle would cause the soil to deform under the constant pressure of the
membrane. The strain would decrease the confining pressure surrounding the
membrane and therefore the measured lateral pressure. A second explanation
suggests the test equipment design may have allowed for some gradual drying of the
sample and therefore a decrease in swelling. This explanation is discounted by
17


author under the assumption that any drying would have been nearly constant for all
the samples.
A doctoral thesis by Snethen (1972) reviews the lateral swelling pressure
relationships for two Oklahoma clays. The research program involved developing a
device that can measure both lateral and vertical swelling. The device consisted of a
2.5-inch inside diameter and 1.25 inch tall Plexiglas cell with a top plate to restrain
vertical swelling. The pressures were measured using pressure transducers. Water
was added to the end of the samples under back pressure.
The maximum lateral pressures decreased with increasing initial moisture
contents. The maximum lateral swelling pressures for the two clays ranged from
approximately 1.6 to 8.7 psi (230 to 1,250 psf). The relationship between maximum
lateral and vertical swell pressures was approximately 1 for samples compacted at an
initial moisture content of 14 percent. When the initial moisture contents are above
14 percent the ratio decreases towards 0.5 at the optimum moisture contents. Above
the optimum moisture contents the ratios were nearly constant.
The study also examined the effect of lateral expansion and the effect on
lateral swelling pressure. This was accomplished by allowing for some deformation
to occur and then returning to restrained conditions. A reduction in lateral swelling
pressure of 50 percent or larger occurred when a lateral expansion of 1 percent was
allowed.
18


2.4.2 Studies Performed using Standard Research Methods
Al-Shamrani and Al-Mhaidib (2000) examined swelling behavior under
oedemetric and triaxial conditions. The study compares the results between swell
tests performed with a triaxial cell and tests performed with oedemeter rings that
restrain lateral swell. The triaxial specimens were tested under a multi-axial stress
state representative of field conditions. While lateral swell is not specifically
discussed, the ratio between vertical swelling and total volumetric swell is reviewed.
The tests were performed using intact specimens of a expansive clay shale in Saudi
Arabia.
The triaxial test results indicate the ratio between vertical swell and
volumetric swell ranged from 0.40 to 0.78 and averaged 0.55. The ratio increased
with increased confining pressure. The ratio also changed with time for each
specimen with the ratio increasing rapidly in the early stage of testing and then the
rate of increase slowed towards a constant value. The author also compared the
effect on vertical deformation when restraining lateral strains in the oedometer. The
results indicate the final vertical swell under a traxial state of stress is about half the
vertical swell observed when lateral strain is restrained.
A research thesis by Colby (1990) measured lateral and vertical swelling
pressures while developing a design for a laboratory test apparatus. The research
program involved performing laboratory tests to measure the lateral and vertical
using conventional testing equipment (triaxial cell, standard oedometer, and back
19


pressure consolidometer) and using the test results to design a new test apparatus.
The results for six tests were presented in the thesis report. Some of the data were
compromised due to inconsistent sample preparation and systematic error in the test
equipment.
The triaxial test results indicated a vertical swell pressure of 17,400 psf and
lateral swell pressure of 12,600 psf. This is a final lateral to vertical swell pressure
ratio of 0.72. During the test, the ratio of lateral and vertical swell pressure ranged
from 0.20 to 0.77. The vertical swelling pressures obtained from tests that restrained
lateral strain were approximately double that measured in the triaxial cell.
Parcher and Liu (1965) published research data reviewing swelling
characteristics of compacted clays. The study presents triaxial data that lateral
swelling is greater than vertical swelling when measured as total lateral volume
change. The test results suggest that the ratio between vertical and lateral swelling
for low moisture content samples indicates structural alterations in the sample may
occur. The research also documents different swelling values were observed for
different compaction methods. The authors compared the swelling for an
undisturbed and a disturbed sample. This comparison suggests that structural
alterations may not occur in the undisturbed samples. The conclusions of the
research theorize different swells are observed in disturbed samples because of the
different orientation of clay particles. The authors suggest there may be higher
potential for water to be absorbed on particle edges rather than on the faces.
20


2.4.3 Conclusions About Lateral Swell Obtained from Literature Review
A limited amount of published material is available regarding lateral swell in
expansive soil. This appears to be the result of two factors: the difficulty of
measuring lateral swelling and more damage to structures is the result of vertical
deformation. Much of the reviewed literature sources modified existing test
procedures or developed new equipment. Triaxial test results appeared to be most
reliable, but are difficult to perform for several samples and in a timely manner. As
a result, some of the research projects were attempting to establish reliable test
procedures that could be cost effective and easily performed, similar to the
established oedometer test methods for used for vertical heave determination.
The lack of standardized testing procedures and equipment limits any
definitive conclusions about lateral swelling that can be developed. With the
exception of the triaxial test results, the reviewed research data is provided from
methods developed specifically for the presented research project. Therefore, a large
database of comparable lateral swelling test results is not available for the same
laboratory methods.
All of the recent research has attempted to compare the ratio between lateral
and vertical swelling pressures. From the presented data, the ratio of vertical and
lateral swelling pressures change throughout testing. Furthermore, there is a
considerable variation between the reported final swelling ratios. A review of the
ratios at the conclusion of testing (swelling) indicates the ratio of lateral and vertical
21


swelling pressures ranged from 0.4 to as high as 10. However, most reported ratios
were less than 1. The ratios above 1 may be due to the test procedure or absence of
a surcharge pressure. A third difference, which is not thoroughly discussed in the
research, is the variability between soil types and the uncertainty associated with
reliable sample preparation for the same soils.
A similar conclusion was obtained in two of the studies that evaluated the
effect on vertical swell when restraining lateral strain. Both studies (Colby, Al-
Shamrani and Al-Mhaidib) determined the restraint of lateral strain effectively
doubles the measured vertical swell. This conclusion suggests vertical heave
estimates from oedometer type tests could significantly over-estimate vertical swell
and higher vertical swells may result in the field for expansive soils that are forced
maintain lateral at-rest conditions (adjacent to rigid walls). Another study (Snethen)
determined the lateral swelling pressures are reduced by a minimum of 50 percent
when 1 percent of lateral expansion is allowed.
Several studies observed the lateral swelling pressure develops towards a
peak during the wetting process and then decreases towards a constant value. In
these same studies, the ratio between lateral and vertical swelling pressures was
largest near the peak lateral pressure. As presented in the Chen study, the peak
lateral swelling pressure may be up to 8 times greater than the estimated at rest
pressure for the soil. Additionally, the Joshi and Katti large-scale study indicates the
lateral pressure increased to a maximum at a depth of 3 feet. Below this depth the
22


lateral pressure did not increase, which indicates a different pressure distribution for
expansive clays when compared to conventional lateral pressure distributions.
The decrease in lateral swelling pressures after obtaining a peak value is
speculated to be the result of creep behavior or laboratory error in the Phillips study.
Other investigations did not present theories for the decrease in lateral swelling
pressure after the peak value. A similar type of behavior was not observed in the
triaxial stress state tests performed by Al-Shamrani and Al-Mhaidib. This may
indicate a deficiency with the other developed methods, since the triaxial stress state
testing was most representative of field stress conditions.
In conclusion, lateral swelling appears to be an area of limited knowledge.
Future research programs could attempt to repeat test results for the same test
procedures and different soil types. Unfortunately, the variability in preparation of a
soil, particularly clay, will always be a source of uncertainty.
23


3. Test Program
3.1 General
The first part of the testing program involved classifying an expansive soil
for use in the research project. The purpose of the project is to determine the
feasibility of geosynthetic inclusion for reducing swelling deformation in an
expansive soil. After determining the soil properties, a non-woven geosynthetic was
selected for placement in the middle of a compacted cubic specimen. The specimen
was wetted using soil suction to maintain its integrity and the lateral and vertical
deformations were measured. The deformation between two specimens with
geosynthetic inclusion and a specimen without any reinforcement was compared and
discussed.
3.2 Soil Properties
The soil used for the test program was obtained from the excavation for a
residence foundation in Jefferson County, Colorado. The foundation excavation was
in the Bow Mar Subdivision where an engineering consulting firm had previously
performed a drilling investigation to characterize the site. A site vicinity map is
presented in Figure 3.1.
The site for soil collection was selected because the soil deposit appeared to
be relatively uniform. The soil consisted of sandy clay. The drilling investigation
24


Figure 3.1 Vicinity Map
25


results suggest the clay deposit extends to a depth of at least 35 feet below site grade.
A copy of the logs from the drilling investigation is presented in Figure 3.2.
The soil for the test program was collected from the foundation excavation at
a depth of 5 feet below the site grade. The soil was obtained by removing native
material from the excavation sidewall. Approximately 500 pounds of soil was
removed from the one location.
Index testing was performed on the soil to classify the material. Index tests
included gradation analysis, hydrometer analysis, Atterberg limit determination, and
specific gravity. A relationship between moisture density and one-dimensional
swelling potential also was established using samples obtained from a Proctor test
performed in accordance with ASTM D698. The results of laboratory testing are
summarized in Table 3.1, and Figures 3.3 and 3.4.
The relationship between moisture and density and the vertical swell and
pressure was established by performing one-dimensional swell tests on samples
extruded from the proctor mold. The swell testing was performed in a standard
oedometer ring and the samples where wetted under a surcharge pressure of 200
pounds per square foot (psf)- The moisture and density swell relationships for the
remolded soil indicate the highest swell percent and pressure occurs at an
approximate moisture content of 17 percent and a dry density of 97.5 pounds per
cubic foot (pcf). The relationship between moisture and density and the swelling
properties are presented in Figures 3.5 and 3.6.
26


-FEET
i
0
5
10
20
25
30
35
TH-1
131/12
MLL = 49
PI = 28
J 31/12
WC = 14.5
DD = 115
-200 = 62
] 50/11

48/12
WC = 17.9
DD = 111
-200 = 75
/ 1130/12
WC = 18.3
DD = 110
TH-2 TH-3
m o
] 18/12 WC 19.7 / /
/ / DD = 92* / i /
-200 = 78
/ r /
/ / ] 42/12 / " / ]50/11
/ / WC = 15.4 / , / WC = 15.8
'; DD = 101 s DD = 113
-200 = 83 -200 = 73
> . >
/ /
] 50/10 / / D4^12 10
/ / / / WC = 13.7
/ / DD = 113
/ * / -200 = 59
> /

] 50/12 15
' / WC 15.8
DD = 109
/ -200 = 74
/
'/j _
/ 20
' /
'/
/
/
f /
bJ ]5G/6 25
EXPLANATIONS:
g CLAY, SANDY, MOIST (TOPSOIL)
^1 CLAY, SANDY, VERY STIFF TO HARD,
' SUGHTLY MOIST TO MOIST, LIGHT
U BROWN TO DARK BROWN (CL)
| 31/12 INDICATES THAT 31 BLOWS OF A140 POUND HAMMER FALLING 30 INCHES WERE
J REQUIRED TO DRIVE A 2.5 INCH O.D. SAMPLER 12 INCHES
NOTES: 1. THE EXPLORATORY BORINGS WERE DRILLED ON 2-17-00 USING A 4-INCH DIAMETER
CONTINUOUS FUGHT AUGER POWERED BY CME 45 DRILLING RIG.
2. NO FREE WATER WAS ENCOUNTERED AT THE TIME OF DRILLING.
3. WC = WATER CONTENT (%) DD = DRY DENSITY (PCF)
-200 = PERCENT OF FINES PASSING THE NUMBER 200 SIEVE.
LL = UQUID UMTT, PI = PLASTIC INDEX
FIGURE 3.2
LOGS OF EXPLORATORY BORINGS
27
-FEET


PERCENT PMItNfc
Figure 3.2 Gradation Analysis
28
PCRCCNT RfTAlMEO


DRY DENSITY LBS /FT
MOISTURE CONTENT PERCENT OF DRY WEIGHT
Figure 3.4 Proctor Curve
I
29


12
u-10 --
1/5
0.
§8
3
to
s
0)
CJ
o;
CL
4 --
2 --
i--------1-------1-------1-------1-------1------1-------1------1
14.9% 16.6% 17.9% 20.9% 22.8%
Moisture Content (percent)
i--------1-------1-------1-------1-------1------1-------1------1
14.9% 16.6% 17.9% 20.9% 22.8%
Moisture Content (percent)
-r 8000
- 7000
- 6000
-- 5000
- 4000
- 3000
- 2000
- 1000
- 0
Figure 3.5
Swell Percent and Pressure Versus Moisture Content
For Remolded Clay samples
Swelling Pressure (PSF)


Percent Swell @ 200 PSF
12 T
10 ..
8 --
6 ~
4 -
2 --
0
90 9
+
-t-

97 98 5 102
Proctor Dry Density (PCF)
100 2
T 8000
-- 7000
-- 6000
5000
-- 4000
3000
2000
-l- 1000
0
Figure 3.6
Swell Percent and Pressure Versus Proctor Density For
Remolded Clay samples
Swelling Pressure (PSF)


Table 3.1
Results of Laboratory Test Data
Test Parameter Results
Percent Passing the No. 4 Sieve 100
Percent Passing the No. 200 Sieve 70
Percent Finer than 2p 52
Liquid Limit Range 46-49
Plastic Index Range 23-28
Specific Gravity 2.74
Activity Range 0.44-0.54
Maximum Dry Density (ASTM D698) 102 pcf
Optimum Moisture Content (ASTM D698) 20.8 %
USCS Designation CL
32


3.3 Geosvnthetic Properties
The geosynthetic used in the study was a nonwoven, needle-punched
geotextile. Some index properties of the geotextile, as provided by the
manufacturer, are presented in Table 3.2
3.4 Sample Preparation
The excavated clay fragments were contained in large plastic bin prior to
processing. The clay fragments were pulverized with hand-tools and then screened
through a No. 4 sieve. After all material was screened, the soil was repeatedly
mixed and returned to the plastic bin where the soil was allowed to cure. A
representative sample of the processed soil was selected for index and proctor tests.
The swell testing of proctor samples indicated the highest vertical swell
occurred at a moisture content of 17 percent. This moisture content was determined
to be the target moisture content for the test specimens in order to obtain a relatively
high swelling potential. The initial moisture content of the processed soil was below
17 percent. Water was added to the processed soil using a water bottle, and the soil
was then mixed and allowed to cure. This process was repeated four times until the
moisture content of the processed soil was just above 17 percent.
A cubical specimen mold with a volume of 1 ft3 was constructed with
plywood. A photograph of the specimen mold is presented on Figure 3.7. The
Standard Proctor dry density was 97.5 pcf for the soil with a target moisture content
33


Table 3.2 Some Index Properties of the Geotextile
Property Value
Weight (oz/yd2), ASTM D-3776 13.0
Thickness (mils), ASTM D-1777 155
Grab Strength (lbs), ASTM D-4632 375
Grab Elongation (%), ASTM D-4632 50
Trapezoidal Tear Strength (lbs), ASTM D-4632 130
Puncture Resistance (lbs), ASTM D-4533 155
Mullen Burst Strength (psi), ASTM D-4833 700
Water Flow Rate (gpm/ft2), ASTM D-3786 60
Permittivity (1/sec), ASTM D-4491 0.81
Permeability (cm/sec), ASTM D-4491 0.32
34


Figure 3.7
Test Specimen Mold
35


of 17 percent. The corresponding moist unit weight of the sample is approximately
114 pcf.
Based on a moisture content of 17 percent, the target moist weight for the test
specimens was 114 pcf. A pilot specimen was constructed using a comparable clay
soil. The compaction of the pilot sample was best obtained when compacted in 1-
inch lifts. As a result, the test specimens were compacted in 1-inch lifts. The
processed soil was split into 36 sub-samples, each with a weight of approximately
9.5 pounds. Each sub-sample was stored in a sealed container. An approximate 0.5-
pound soil mass also was obtained with each sub-sample for moisture content
determinations. The sub-samples were arranged into 3 groups, one group for each
specimen. The sub-samples were arranged so the soil for each specimen would be
similar. For example, the sub-samples from the top, middle, and bottom of the
storage bin where evenly split between the three specimens.
Each specimen was constructed by placing a sub-sample in the mold and
compacting the soil into a lift with a target thickness of 1-inch. Each specimen was
constructed with 12 lifts. The soil was compacted using a 5-pound Standard Proctor
hammer. The specimens with geosynthetic inclusion were constructed by placing a
12-inch by 12-inch non-woven geosynthetic at the 6-inch elevation of the specimen.
The soil above the geosynthetic was compacted outward from the center of the mold
to minimize slack in the reinforcement.
36


The completed specimen and mold were placed onto a smooth plexiglass
surface with lubricant between the soil and plexiglass. The mold was then
disassembled and the measurement devices (dial gauges and digital displacement
gauges) were mounted on the sides and top of the specimen to monitor vertical and
lateral movement. The top measurement device was placed in the middle of the
specimen. Two lateral measurement devices were placed at the middle of the 6-inch
elevation of the specimen (1 in the machine direction and 1 in the cross direction of
the geosynthetic reinforcement). A similar arrangement was used for the opposite
sides; however, the devices were placed at 3- and 9-inch elevations. A photograph
of a completed specimen with the measurement devices is presented in Figure 3.8.
3.5 Test Procedure
Following the sample preparation, the sides and top of the specimen were
surrounded with moist sponges to provide water for swelling. The initial test
procedure was to completely submerge the sample; however, pilot scale samples
would not maintain sufficient strength to complete the testing when submerged.
Therefore, alternative methods of wetting the sample were investigated. The
sponge method was developed to utilize the theory of soil suction.
The sponge method involved adding water to maintain relatively high
moisture content in the sponges. Water was added to the sponges at similar rate
using a syringe for the three specimens. The water was added to maintain high
37


Figure 3.8
T>pical Specimen Setup
38


moisture levels in the sponge, but also minimize any free water that may accumulate
at the base of the specimen. A plot of the volume of water added for each specimen
is presented in Figure 3.9.
The sponges were placed in direct contact with the sides and top of the soil
specimen. Water was added at select times and the deformation was recorded with
the measurement devices. The deformation and volume of water added were
monitored until the rate of deformation was near zero. At this point the test was
concluded. Approximately 72 hours was required to complete the testing.
After completion of the test, the specimens were broken down and samples
for moisture content analysis were taken at 1,3, 5, 7, 9, and 11-inch elevations. The
moisture content determinations were reviewed to establish a comparison between
the final moisture content of each specimen.
39


V olum e (m L)
o
Elaps e d Tim e (ho urs)
Figure 3.8
Rate of Water Addition for Test Specimens


4. Test Results and Discussion
4.1 General
Two tests were performed with horizontal geosynthetic inclusion in the
middle of the specimen. To assess the effect of geosynthetic inclusion, a specimen
also was tested without a geosynthetic. The specimens with geosynthetic are
designated GRS-1 and GRS-2. Conversely, the specimen without geosynthetic
inclusion is designated NO GRS.
The cubic specimens were allowed to swell without confining or surcharge
pressures so that the restraining effect of geosynthetic inclusion could be assessed.
The specimens were provided water using soil suction potential rather than
submerging the sample.
4.2 Swell Without Geosynthetic Inclusion
The vertical deformation results for the specimen without geosynthetic
inclusion are presented in Figure 4.1. The vertical swell rate decreased throughout
the duration of the test. During the first 24 hours, the swelling rate was
approximately 0.12 percent per hour and then decreased to 0.08 percent per hour in
the next 24 hours. The rate of swell further decreased after 48 hours and was less
than 0.01 percent per hour in the final 12 hours of the test. A total vertical swell of
5.09 percent (0.61 inch) was measured in this specimen.
41


Elapsed Time (hours)
Figure 4.1
Specimen NO GRS Percent Vertical Swell With Time


Lateral deformation measurements from 3, 6, and 9-inch elevations are
presented in Figure 4.2. Generally, the total lateral deformations were largest near
the top and bottom of the specimen. During testing, the rates and amount of lateral
swell were variable and some sample contraction also was observed on one side;
however, the contraction was recovered by the conclusion of the test. One possible
reason for sample contraction during the test may be due to a more dominant swell
in the opposite direction.
Similar to the vertical swell measurements, the rate of lateral deformation
approached zero after 48 hours. The total measured lateral swell ranged from 0.02
percent at the middle of the specimen on the east side to 3.63 percent at the 9-inch
elevation on the west side. The average of the four measurements from 3 and 9-inch
elevations was 2.84 percent. The average swell measured in the middle of the
sample was 0.97 percent. The average of all measured lateral deformation was 2.22
percent. An oblique photograph of the test specimen is presented on Figure 4.3.
4.3 Swell With Geosynthetic Inclusion
Two specimens were tested with a horizontal geosynthetic placed in the
middle of the specimen. The sides that intersect the machine direction in
geosynthetic are equivalent to the north and south sides of the specimen tested
without geosynthetic inclusion. Likewise, the east and west sides of the specimens
are parallel to the machine direction in the geosynthetic.
43


Percent Swell

Specimen NO GRS
Percent Swell With Time


North Elevation
Figure 4.3
Photographs of Specimen NO GRS at the Conclusion of Testing
45


4.3.1 Test GRS-1
The measured vertical swell measured in the GRS-1 specimen is presented
on Figure 4.4. The measured rate of vertical swell was 0.13 percent per hour for the
first 24 hours and 0.08 percent per hour in the next 24-hour period. After 52 hours
of wetting, the vertical swell stabilized and remained constant for the remainder of
the test, 72 hours total. The total measured vertical swell was 5.49 percent or 0.66
inches.
Similar to the patterns of the measured lateral in the NO GRS specimen, the
GRS-1 specimen also indicated some contraction in select locations during the first
half of the test. The contractions also were recovered by the completion of the test.
The lateral deformation measurements at the various locations are presented in
Figure 4.5.
The total lateral deformations ranged from 0.04 to 2.57 percent and averaged
1.52 percent. Measurements of total lateral swell measured at the middle of the
specimens averaged 0.66 percent. The average of the total lateral swell
measurements from the 3 and 9-inch elevations was 1.95 percent. A photograph of
the GRS-1 sample at the completion of testing is presented in Figure 4.6.
4.3.2 Test GRS-2
The vertical deformation results for the GRS-2 are presented on Figure 4.7.
The total vertical swell measured in the GRS-2 specimen was 4.27 percent (0.51
46


0:00 00 12:00:00 24:00:00 36:00 00 48 00 00 60:00 00 72:00:00 84:00:00
Elapsed Time (hours)
Figure 4.4
Specimen GRS-1 Percent
Vertical Swell With Time


Percent Swell
3
South @ 3 Inches
- South @ 9 Inches
West @ 3 Inches
* West @ 9 Inches
* North @ 6 Inches
East @ 6 Inches

84 00
-1
Figure 4.5
Specimen GRS-1 Percent
Lateral Swell with Time



Figure 4.6
Photographs of Specimen GRS-1 at Conclusion of Testing
49


Percent
Figure 4.7
Specimen GRS-2 Percent
Vertical Swell With Time


inches). The rate of swell was variable in the first 48 hours of the test. The average
rate of swell was 0.08 percent per hour in the first 24 hours and 0.9 percent in 24 to
48 hour period. After 48 hours testing, the rate decreased and only 0.14 percent
swell occurred in the last 24 hours of testing.
The patterns of lateral deformations in the GRS-2 specimen were similar to
the GRS-1 and NO-GRS specimens. The GRS-2 specimen indicated the lowest
average lateral deformation of the three specimens. The measured lateral
deformations are presented on Figure 4.8.
The rate of lateral swell measured in the GRS-2 specimen decreased
significantly after 48 hours. This was similar to the other two specimens. The total
measured lateral deformation ranged from 0.07 percent at the middle of the
specimen on the east side to 2.80 percent at the 9-inch elevation on the west side.
The average for all total lateral deformations was 1.35 percent. The average of the
four measurements from 3 and 9-inch elevations was 1.87 percent. The average
swell measured in the middle of the sample was 0.31 percent.
4.4 Summary and Discussion of Test Results
A summary of the total swelling measurements for the three specimens is
presented in Table 4.1. The data suggest the inclusion of the horizontal geosynthetic
does not influence vertical deformation. The vertical swelling ranged from 4.3
percent for specimen GRS-2 to 5.5 percent for specimen GRS-1. The NO-GRS
specimen indicated a total vertical swell of 5.1 percent.
51


Percent Swell
K>
Elapsed Time (hours)
Figure 4.8
Specimen GRS-2 Percent
Lateral Swell With Time


Table 4.1
Summary of Total Swelling Measurements
Measurement Location Sample ID
NO-GRS GRS-1 GRS-2
Top 5.1 5.5 4.3
South at 3 Inches 2.0 1.3 2.1
South at 9 Inches 2.7 1.6 1.6
West at 3 Inches 3.0 2.6 1.0
West at 9 Inches 3.6 2.3 2.8
North at 6 Inches 1.9 1.3 0.6
East at 6 Inches 0.0 0.0 0.1
Average of 3-Inch Measurements 2.5 2.0 1.6
Average of 9-Inch Measurements 3.2 1.9 2.2
Average of 3 and 9-Inch Measurements 2.8 1.9 1.9
Average of 6-Inch Measurements 1.0 0.7 0.3
Average of All Lateral Measurements 2.2 1.5 1.4
53


The total lateral deformations do appear to be influenced by inclusion of the
geosynthetic. The total lateral swell ranged from 0.02 to 3.63 percent and averaged
2.22 percent. Conversely, total lateral swell measured in the GRS-1 and GRS-2
specimens ranged from 0.4 to 2.80 percent. The average of all lateral measurements
for the GRS-1 and GRS-2 specimens was 1.52 and 1.35 percent, respectively. These
values are approximately 68 and 61 percent of the average lateral swell measured in
the NO-GRS specimen.
When reviewing the lateral swelling measurements, it is important to note the
vertical swell observed in the NO-GRS specimen was between total vertical swell
measured in the GRS-1 and GRS-2 specimens. Without geosynthetic inclusion in
the three specimens, it would be reasonable to expect a similar pattern to the
distribution of lateral deformation. However, both specimens with geosynthetic
inclusion indicated smaller lateral deformations when compared to the specimen
without geosynthetic inclusion. This suggests the inclusion of geosynthetic in an
expansive soil can reduce swelling in the direction parallel to the geosynthetic.
4.5 Review of the Initial and Final Specimen Moisture and Densities
Tables 4.2 through 4.4 present the initial and final moisture content
determinations for each specimen. Table 4.5 summarizes the data for all three
specimens. The initial dry densities for the three specimens were within 0.1 pcf,
which calculates to a percent relative standard deviation of 0.06 percent. This
54


Table 4.2
Initial and Final Moisture Content Determination Worksheet
Specimen NO-GRS
Mold Compaction Sheet
Number of Lifts: 12
Approximate Moisture Content: 17%
Target Dry Density (Ibs/ft3): 97.5
Target Moist Density (Ibs/ft3): 114.1
Mass of Soil/Lift (lbs): 9.51
Actual Moisture Conditions Prior to Testing
Mass of Moist Soil + Dish (gr) Mass of Dry Soil + Dish (gr) Weight of Water (gr) Dish Weight (gr) Moisture Content (%)
Lift 1 280.3 259.7 20.6 132 16.1%
Lift 2 422.9 392.9 30 216 17.0%
Lift 3 449.1 414.8 34.3 213.5 17.0%
Lift 4 383.7 359 24.7 213.3 17.0%
Lift 5 506 465.4 40.6 225.6 16.9%
Lift 6 336.9 308.1 28.8 134 16.5%
Lift 7 337,8 320.4 17.4 213.4 16.3%
Lift 8 318.1 295.3 22.8 155 16.3%
Lift 9 322.8 295.6 27.2 131 16.5%
Lift 10 302.6 281.8 20.8 155.2 16.4%
Lift 11 412.6 384.9 27.7 213.8 16.2%
Lift 12 492.5 452.9 39.6 210 16.3%
Average Moisture Content (%): 16.5%
Moist Weight of Soil (lbs): 114.1
Dry Density of Mold Soil (Ibs/ft3): 97.9
Final Moisture Contents at Conclusion of Testing
Lift 1 512.3 441.8 70.5 213.4 30.9%
Lift 3 293.1 256.6 36.5 134.0 29.8%
Lift 5 333.5 287.7 45.8 130.9 29.2%
Lift 7 537.8 466.3 71.5 213.7 28.3%
Lift 10 461.4 401.9 59.5 210.0 31.0%
Lift 12 343.6 293.3 50.3 133.3 31.4%
Average Final Moisture Content (%) 30.1%
55


Table 4.3
Initial and Final Moisture Content Determination Worksheet
Specimen GRS-1
Mold Compaction Sheet
Number of Lifts: 12
Approximate Moisture Content: 17%
Target Dry Density (lbs/ft3): 97.5
Target Moist Density (Ibs/ft3): 114.1
Mass of Soil/Lift (lbs): 9.51
Actual Moisture Prior to Testing
Mass of Moist Soil + Dish (gr) Mass of Dry Soil + Dish (gr) Weight of Water (gr) Dish Weight (gr) Moisture Content (%)
Lift 1 358.9 338.1 20.8 212 16.5%
Lift 2 327.5 300.2 27.3 134 16.4%
Lift 3 425.3 395.2 30.1 216.4 16.8%
Lift 4 420.7 390.9 29.8 213 16.8%
Lift 5 383.1 358.8 24.3 210.4 16.4%
Lift 6 299.40 275.8 23.6 133.9 16.6%
Lift 7 373.4 350.6 22.8 213.5 16.6%
Lift 8 342.1 312.2 29.9 134 16.8%
Lift 9 362.7 341.9 20.8 216.5 16.6%
Lift 10 308.4 283.4 25 134.2 16.8%
Lift 11 402.1 375.4 26.7 213.7 16.5%
Lift 12 365.8 344.2 21.6 211.9 16.3%
Average Moisture Content (%): 16.6%
Moist Weight of Soil (lbs): 114.1
Dry Density of Mold Soil (lbs/ft3): 97.9
Final Moisture Contents at Conclusion of Testing
Lift 1 515 447.4 67.6 213.5 28.9%
Lift 3 329.4 287.5 41.9 154.7 31.6%
Lift 5 503.7 435.5 68.2 213.6 30.7%
Lift 7 455 401.3 53.7 213.2 28.5%
Lift 10 507.9 441.8 66.1 210.6 28.6%
Lift 12 510.90 439.7 71.2 212.1 31.3%
Average Final Moisture Content (%) 29.9%
56


Table 4.4
Initial and Final Moisture Content Determination Worksheet
Specimen GRS-2
Mold Compaction Sheet
Number of Lifts: 12
Approximate Moisture Content: 17%
Target Dry Density (Ibs/ft3): 97.5
Target Moist Density (Ibs/ft3): 114.1
Mass of Soil/Lift (lbs): 9.51
Actual Moisture Conditions Prior to Testing
Mass of Moist Soil + Dish (gr) Mass of Dry Soil + Dish (gr) Weight of Water (gr) Dish Weight (gr) Moisture Content (%)
Lift 1 296.3 273.1 23.2 134.3 16.7%
Lift 2 329.7 302 27.7 133.9 16.5%
Lift 3 317.8 291.8 26 133.8 16.5%
Lift 4 312.2 286.6 25.6 131.9 16.5%
Lift 5 337.6 309 28.6 133.9 16.3%
Lift 6 362.1 341.3 20.8 213.8 16.3%
Lift 7 386.6 362.2 24.4 213.4 16.4%
Lift 8 413.9 385.4 28.5 211.9 16.4%
Lift 9 404.9 379.8 25.1 225.6 16.3%
Lift 10 392.5 367.7 24.8 213.3 16.1%
Lift 11 429.4 399.3 30.1 216.6 16.5%
Lift 12 439.5 407.9 31.6 213 16.2%
Average Moisture Content (%): 16.4%
Moist Weight of Soil (lbs): 114.1
Dry Density of Mold Soil (Ibs/ft3): 98.0
Final Moisture Contents at Conclusion of Testing
Lift 1 510.9 439.7 71.2 212.1 31.3%
Lift 3 507.9 441.8 66.1 210.6 28.6%
Lift 5 455 401.3 53.7 213.2 28.5%
Lift 7 503.7 435.5 68.2 213.6 30.7%
Lift 10 329.4 287.5 41.9 154.7 31.6%
Lift 12 515.00 447.4 67.6 213.5 28.9%
Average Final Moisture Content (%) 29.9%
57


Table 4.5
Summary of Initial and Final Moisture Contents
Sample ID Average Initial Moisture Content Dry Density of Specimen (pcf) Average Final Moisture Content Change in Moisture Content
NO-GRS 16.5% 97.9 30.1% 13.6%
GRS-1 16.6% 97.9 29.9% 13.3%
GRS-2 16.4% 98.0 29.9% 13.5%

Statistical Summary
Average 16.5% 97.9 30.0% 13.5%
Standard Devation 0.1% 0.058 0.1% 0.2%
% RSDV 0.61% 0.06% 0.39% 1.13%
^Relative Standard Deviation
58


indicates the dry densities of the specimens are nearly identical within the limitations
of the measurement. A similar conclusion can be formulated from the initial
moisture content data.
The moisture content data also indicate the initial and final moisture contents
were within 0.2 percent for the three specimens. The change in initial and final
moisture contents ranged from 13.3 percent for specimen GRS-1 to 13.6 percent
specimen NO-GRS. The percent relative standard deviation for the final moisture
contents was 0.39 percent, which indicates the final moisture contents are practically
the same within the limitations of measurement.
The initial moisture and density measurements and the final moisture content
data indicate the measured distribution in swell values is not the result of errors in
sample preparation or test procedures.
59


5. Summary, Conclusions, and Recommendations for Future Research
5.1 Summary
The purpose of this study was to determine the feasibility of geosynthetic
inclusion for reducing swelling deformation in an expansive soil. Prior to beginning
the test program, a literature review for similar studies was performed. The
laboratory program involved establishing a test method to allow for the measurement
of free swell in the vertical and horizontal dimensions. The results of laboratory
investigation were compiled to determine the effects of geosynthetic inclusion on
free swell.
The laboratory test program involved compacting three, 1 ft3 specimens at
the same moisture and density conditions. Two of the specimens were constructed
with a horizontal, non-woven geosynthetic placed at the 6-inch elevation in the
sample. Lateral and vertical deformation measurement devices were setup and the
specimens were surrounded with sponges. Water was added to the sponges at select
times to maintain a relatively high moisture condition in the sponges. The swelling
occurred via soil suction from the moist sponges into the expansive soil. A
statistical review of the initial moisture and density conditions and the final moisture
contents suggests the measured differences in swelling for the three specimens are
not the result of errors in sample preparation or test procedures.
60


5.2 Conclusions
Based on the review of the test program data, the following conclusions can
be developed.
1. The test method developed in this study appears to be a viable yet simple
method for evaluating swelling potential of soil-geosynthetic composites.
Fairly consistent results were obtained from tests GRS-1 and GRS-2,
performed under similar conditions.
2. The inclusion of a horizontal geosynthetic layer does not appear to affect
swelling in the vertical direction. The rates of swell decreased significantly
after 48 hours for the three specimens. The distribution in vertical swell
appears to be random and independent of horizontal geosynthetic inclusion.
The differences in measured total vertical swell may be due to variability
within the soil.
3. The inclusion of a geosynthetic did appear to reduce the amount of measured
lateral deformation. This conclusion is based on two observations. First, the
average of the lateral swell measurements in specimens with geosynthetic
inclusion were 32 and 39 percent less than the similar average from the
specimen without geosynthetic inclusion. Secondly, the vertical swell
observed in the specimen without a geosynthetic was between total vertical
swell measured in specimens with a geosynthetic. Without geosynthetic
inclusion in the three specimens, it would be reasonable to expect a similar
61


pattern to the distribution of lateral deformation. However, both specimens
with geosynthetic inclusion indicated smaller lateral deformations when
compared to the specimen without geosynthetic inclusion.
4. The use of geosynthetic inclusion may be a practical technique in situations
where reducing lateral swelling is of major concern, such as a basement wall.
5. It may be possible to reduce vertical swelling of an expansive soil by placing
geosynthetic in a vertical direction. The installation of a vertical
geosynthetic is difficult for most constructed earth structures; however, a
wick drain with geosynthetic filter, which can be inserted into the ground of
an expansive soil, may be an effective means for reducing the vertical swell.
5.3 Recommendations for Future Research
Based on the literature review, a limited amount of study has been performed
regarding the inclusion of a geosynthetic in expansive soil. The only similar study
has examined fiber reinforcement of expansive soil. Other studies have established
the role of geosynthetic as a drainage layer in clay soil.
Any future research could involve examining the effect of placing multiple
layers of geosynthetic into an expansive clay fill. The spacing of geosynthetic layers
may further reduce the lateral deformation of a swelling soil. Additionally, there
may be a change in the rate of swell with multiple geosynthetic layers because the
facilitation of drainage that can occur with a non-woven geosynthetic.
62


A testing program that examines the effect of a geosynthetic on swelling
pressures also could be developed. This study determined the feasibility of a
geosynthetic in an expansive soil. The observed reduction in the lateral deformation
for samples with a geosynthetic should be used as a basis to determine if a
comparable conclusion can be developed for swelling pressures.
Finally, field scale test programs should be developed to determine the
relationship between laboratory and field scale test results. If lateral deformation
and swelling pressures can be reduced using geosynthetic reinforced soil, potential
applications could include placing a geosynthetic material in retaining wall and
foundation wall backfill that consists of expansive soil. Also, it may be possible to
reduce vertical swelling deformations and pressures if a method for installing
geosynthetic layers in a vertical direction can be developed.
63


References
Al-Amoud, Sulaiman I., 1993 The Relationship Between Lateral and Vertical
Expansive Stresses for a Compacted Clay. Master Thesis, Arizona State Unversity.
Al-Shamrani, Mosleh A. and Al-Mhaidib, Abdullah I., 2000. Swelling Behavior
Under Oedemetric and Triaxial Loading Conditions. Advances in Unsaturated
Geotechnics, Proceedings of Sessions of Geo-Denver 2000, Denver, Colorado,
August. 344-359.
Chen, F. H., Foundations on Expansive Soils, Elsevier Science Publishing Company,
1988, 463 p.
Chen, Fu Hua, 1987. Lateral Expansion Pressure on Basement Walls.
Proceedings of the Sixth International Conference on Expansive Soils, New Delhi,
India, December. 55-59.
Colby, Craig A., 1990. Research and Design of Test Apparatus for Laboratory
Measurement of Lateral and Vertical Swell Pressures. Master Thesis, University of
Colorado.
Holtz, R. D., Christopher, B. R., and Berg, R. R., Geosynthetic Engineering, BiTech
Publishers, 1997,451 p.
Joshi, R.P. and Katti, R.K. (1980). Lateral Pressure Development Under
Surcharges. Proceedings of the Fourth International Conference on Expansive
Soils, Denver, Colorado, June, 227-241.
Koemer, R. M., Designing with Geosvnthetics, Fourth Edition, Prentice Hall
Publisher, 1997, 761 p.
Ofer, Z. (1980). Instruments for Laboratory and In Situ Measurement of the
Lateral Swelling Pressure of Expansive Clays. Proceedings of the Fourth
International Conference on Expansive Soils, Denver, Colorado, June, 45-53.
Parcher, James V., and Liu, P.C. (1965). Some Swelling Characteristics of
Compacted Clays. Journal of the Soil Mechanics and Foundation Division, ASCE,
Vol. 91, No. SM3, 1-17.
64


Phillips, Richard A., 1981. The Lateral Swelling Pressure of an Expansive
Colorado Soil. Master Thesis, University of Colorado.
Puppala, Anand J. and Musenda Chisha, 2000. Effects of Fiber Reinforcement on
Strength and Volume Change in Expansive Soils. Transportation Research Record
1736, Paper No. 00-0716.
Snethen, Donald R., 1972. Lateral Swelling Pressure Relationships for Two
Oklahoma Clays. Doctoral Thesis, Oklahoma State University.
Srinivasa Murthy, B.R., Sridharan, A., Bindumadhava, Revana Siddappa, K., 1987.
Soil Reinforcement Technique for Control of Heave. Proceedings of the Sixth
International Conference on Expansive Soils, New Delhi, India, December. 195-
199.
Wu, J. T. H., Design and Construction of Low Cost Retaining Walls The Next
Generation in Technology, Colorado Department of Transportation, 1994, 152 p.
65