Citation
Engineering properties of a sand-palygorskite mixture contaminated with methanol

Material Information

Title:
Engineering properties of a sand-palygorskite mixture contaminated with methanol
Creator:
Balint, Gregory G
Publication Date:
Language:
English
Physical Description:
xv, 110 leaves : illustrations ; 28 cm

Subjects

Subjects / Keywords:
Hazardous waste sites -- Leaching ( lcsh )
Sanitary landfills -- Linings ( lcsh )
Leachate ( lcsh )
Hazardous waste sites -- Leaching ( fast )
Leachate ( fast )
Sanitary landfills -- Linings ( fast )
Genre:
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )

Notes

Bibliography:
Includes bibliographical references (leaves 108-110).
General Note:
Department of Civil Engineering
Statement of Responsibility:
by Gregory G. Balint.

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:
40326485 ( OCLC )
ocm40326485
Classification:
LD1190.E53 1998m .B35 ( lcc )

Full Text
ENGINEERING PROPERTIES OF A SAND-PALYGORSKITE MIXTURE
CONTAMINATED WITH METHANOL
by
Gregory G. Balint
B. S., Colorado State University, 1987
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
1998


This thesis for the Masters of Science
degree by
Gregory G. Balint
has been approved
by


Balint, Gregory G. (M. S., Civil Engineering)
Engineering Properties of a Sand-Palygorskite Mixture
Contaminated with Methanol
Thesis directed by Assistant Professor Dunja Peric
ABSTRACT
Land disposal is the most common method of disposing of
municipal and hazardous wastes. Starting in the 1960's
and 70's, ground water contamination traced to landfills
was recognized as a problem that needed to be remediated
and prevented in the future. The first source is
hazardous liquid wastes placed in the landfill. The
second source is groundwater infiltrating into the
landfill or precipitation percolating down through the
landfill. This water leaches contaminants from solid
wastes as it moves through the landfill. This
contaminant-laden water, along with any liquid wastes
with which it mixes, is called leachate. The leachate
may move down into the water table resulting in a plume
of contaminated water. The strategy to mitigate this
problem is through regulation and engineering. The
engineering problems fall into two main categories.
First, to minimize the amount of ground water and
precipitation entering the landfill. Second, to minimize
in


the amount of leachate leaving the landfill. Use of
synthetic and compacted soil liners to inhibit flow of
ground water into and leachate out of landfills has
become a commonly used technique.
This study was conceived to explore the overall effects
of an organic chemical contaminant on the engineering
properties of a sand-palygorskite mixture. The purpose of
the research was threefold. First, to investigate the
engineering properties of a fine grained soil saturated
with water; second, to investigate the chemical-
mechanical influence of a liquid organic chemical
(methanol) on the state of the fine grained soil; and
third, to compare the results to verify if there were any
effects to the engineering properties of the soil.
A laboratory testing program was carried out to determine
the Atterberg limits, permeability, and undrained shear
strength of soil specimens saturated with both water and
methanol.
Results showed that methanol reduced the Atterberg
limits, increased the coefficient of permeability and
caused the sand-palygorskite mixture to swell. Results of
the undrained compression tests showed methanol decreased
cohesion, increased the angle of internal friction,
IV


increased the shear strength for specimens at similar
effective stresses and affected the dilatancy of the
soil.
This abstract accurately represents the content of the
candidate's thesis. I recommend its publication.
Signed
Dr, D. Peric
v


ACKNOWLEDGEMENT
The completion of this Masters Degree in Civil
Engineering ends a journey that has lasted almost ten
years. At times, as I looked ahead at all the deficiency
courses, graduate courses, research and writing the
thesis, I wondered if it was possible to stick it out.
Now that this endeavor is drawing to an end, I would like
to acknowledge the people who, without their assistance,
I could not have succeeded.
First, I would like to thank my manager, Kathy Beggane
who set the long-term professional development goal of
obtaining an engineering degree. She, along with my
current manager Dr. Eugene Nuccio, aided me immensely by
allowing the flexibility in my work schedule that made
continuing my education possible. Additionally, I wish to
thank my employers, EG&G Rocky Flats, Inc., and Tenera
Rocky Flats, LLC for their generous financial assistance
with tuition, fees and books.
I would like to extend my sincerest appreciation to Dr.
Dunja Peric, who supervised this study, for providing the
guidance and vision that have led to this most favorable
outcome. Her patience when my frustration level was high
was greatly appreciated. I would also like to thank Dr.
VI


N. Y. Chang and Dr. Jonathan Wu for reviewing this
research.
Additionally, I would like to thank Dr. Peric, Dr. Chang,
Dr. Wu, Dr. Mays, Dr. Guo, Dr. Hughes, and Dr. Hubly for
sharing their time, knowledge and experience with me in
the classroom. Finally, I would like to thank my two
fellow graduate students, Mark Shepard and Jan Chang.
Their encouragement and assistance in the classroom and
lab were invaluable.
Vll


CONTENTS
Chapter
1. Introduction.........................................1
1.1 Problem Description...............................1
1.2 Purpose...........................................4
1.3 Scope of Study....................................4
2. Literature Review....................................7
2.1 Introduction......................................7
2.2 Permeability Measurement Techniques...............8
2.3 Effects of Chemicals on Clay Hydraulic
Conductivity.....................................11
2.4 Effects of Organic Chemicals on Soil
Strength.........................................18
2.5 Applications.....................................20
3. Materials...........................................22
3.1 Soil.............................................22
3.1.1 Selection......................................22
3.1.2 Palygorskite Clay.............................22
3.1.3 Sand...........................................23
3.1.4 Palygorskite Clay and Sand Mixture............26
3.2 Methanol.........................................30
3.2.1 Selection......................................30
viii


Chapter
3.2.2 Description and Properties.....................30
4. Testing Program, Equipment, and Procedures.........33
4.1 General..........................................33
4.2 Testing Program..................................33
4.2.1 Soil Classification and Characterization......33
4.2.2 Isotropic Consolidation Test.............34
4.2.3 Undrained Shear Tests..........................34
4.3 Equipment........................................35
4.3.1 Flow Pump System...............................35
4.3.2 Loading Frame.................................4 8
4.4 Procedures.......................................48
4.4.1 Sample Preparation.............................48
4.4.2 Isotropic Consolidation Test.............53
4.4.3 Flow Phase.....................................56
4.4.4 Undrained Shear Tests..........................57
4.4.5 Equipment Calibration..........................59
5. Test Results........................................62
5.1 Data Analysis....................................62
5.1.1 Flow Test Data Analysis........................62
5.1.2 Undrained Shear Test Data Analysis.............65
5.2 Isotropic Consolidation Test Results.............67
5.3 Flow Test Results................................69
ix


Chapter
5.4 Undrained Shear Test Results....................81
5.5 Summary and Discussion..........................95
6. Summary, Conclusions, and Recommendations
for Future Research...............................101
6.1 Summary........................................101
6.2 Conclusions....................................103
6.3 Recommendations for Future Research............105
REFERENCES
x


FIGURES
Figure
3.1 Photograph of Palygorskite Clay..................24
3.2 Photograph of Air-Dried Sand.....................25
3.3 Photograph of Sand and Palygorskite Mixture......27
3.4 Moisture-Dry Weight Curve for Sand-Clay
Mixture..........................................29
3.5 Sand-Palygorskite Mixture Grain Size
Distribution.....................................29
4.1 Photograph of Flow Pump..........................40
4.2 Photograph of Bladder Accumulators...............41
4.3 Photograph of differential Pressure
Transducers DPT-1 and DPT-2......................42
4.4 Photograph of Burette and Volume Change
Differential Pressure Transducer.................45
4.5 Schematic Drawing of the Modified Flow
Pump System.....................................4 6
4.6 Schematic Drawing of the Portion of the Flow
Pump System Used in the Undrained Shear Test
Phase............................................47
4.7 Photograph of Compaction Tool and Mold...........50
4.8 Photograph of Isotropic Consolidation
Testing System...................................55
5.1 Isotropic Consolidation Line (log Scale)..........68
5.2 Isotropic Consolidation Line......................68
5.3 Pore Pressure Changes Due to Flow.................72
xi


Figure
5.4 Changes in Effective Stress State for
Flow Test M_52......................................7 5
5.5 Void Ratio Versus Time for Test M_52...............75
5.6 Hydraulic Gradient Versus Time for Flow
Test M_52...........................................76
5.7 Hydraulic Conductivity Versus Time for
Flow Test M_52...................................7 6
5.8 Volume Versus Time for Flow Test M_52............77
5.9 Volume Versus Time for Flow Test M_186...........77
5.10 Changes in Effective Stress State for Flow
Test M_186........................................78
5.11 Void Ratio Versus Time for Flow Test M_186........78
5.12 Hydraulic Gradient Versus Time for Flow
Test M_186........................................79
5.13 Hydraulic Conductivity Versus Time for
Flow Test M_186...................................79
5.14 Void Ratio Changes for Tests M_52 and
M_186 (log scale).................................80
5.15 Void Ratio Changes for Tests M_52 and
M_18 6............................................80
5.16 Deviator Stress Versus Axial Strain for
Test W_22CIU......................................85
5.17 Excess Pore Liquid Pressure Versus Axial
Strain for Test W_22CIU..........................85
5.18 Deviator Stress Versus Axial Strain for
Test W_44CIU......................................86
5.19 Excess Pore Liquid Pressure Versus Axial
Strain for Test W 44CIU..........................86
xii


Figure
5.20 Deviator Stress Versus Axial Strain for
Test W_187CIU.....................................87
5.21 Excess Pore Liquid Pressure Versus Axial
Strain for Test W_187CIU..........................87
5.22 Deviator Stress Versus Axial Strain for
Test M_52CIU......................................88
5.23 Excess Pore Liquid Pressure Versus Axial
Strain for Test M_52CIU...........................88
5.24 Deviator Stress Versus Axial Strain for
Test M_186CIU.....................................89
5.25 Excess Pore Liquid Pressure Versus Axial
Strain for Test M_186CIU..........................89
5.26 Stress Paths for the Water CIUC Tests.............92
5.27 Stress Paths for the Methanol CIUC Tests..........93
5.28 Water Test Mohr's Circle Plots, Shear
Stress Versus Normal Stress.......................94
5.29 Methanol Test Mohr's Circle Plots, Shear
Stress Versus Normal Stress.......................94
5.30 Af Versus Initial Effective Stress.................99
xiii


TABLES
Tables
3.1 Result of Index Property Tests of
Palygorskite Clay................................23
3.2 Summary of Laboratory Procedures Used.............28
3.3 Standard Proctor and Atterberg Limit Test
Results on Sand-Palygorskite Mixture.............28
3.4 Methanol Released to the Environment in
the United States in 1989........................31
3.5 Methanol Information..............................32
4.1 Flow Phase Control Modes..........................35
4.2 Summary of Triaxial and Flow test Program.........36
4.3 Undrained Shear Phase Control Mode................36
4.4 Relative Compaction of Soil Specimens.............49
4.5 Isotropic Consolidation Testing Program...........56
5.1 Isotropic Consolidation Test Data.................67
5.2 Isotropic Consolidation Test Void Ratio Data......69
5.3 Initial and Post Consolidation Data for
Flow Tests.......................................72
5.4 Intrinsic Permeability for Water and Methanol .... 74
5.5 Coefficients of Permeability for Water and
Methanol.........................................74
5.6 Coefficient of Permeability and Intrinsic
Permeability Increase Factors....................82
xiv


Tables
5.7 Undrained Shear Test Data.........................83
5.8 Summary of Undrained Shear Strengths at
Failure..........................................95
5.9 Modulus of Elasticity for Selected Specimens......95
5.10 Skempton's A Parameter at Failure for Water
Saturated Test Specimens.........................98
5.11 Skempton's A Parameter at Failure for Methanol
Saturated Test Specimens ............................
98


1.
Introduction
1.1 Problem Description
Disposal of municipal and industrial wastes has become
increasingly problematic during the last half of the
twentieth century. In 1992, the United States (U.S.) and
its territories contained 5,345 landfills (EcoWeb the
Program, 1998). The reasons for the problem are twofold.
First, the industrial development that has occurred over
the last 150 years has had the undesired side effect of
producing enormous guantities of waste. Second, the
constituents of the wastes have also become more diverse.
In many cases these wastes are hazardous. A partial list
would include municipal waste, demolition debris, waste
water treatment facility sludge, garbage, fly ash,
radioactive materials, and toxic chemicals (Fetter,
1994) .
Land disposal is and for some time has been the most
commonly used means of disposing of these wastes. In
1992, 67 percent or 118 million metric tons (130 million
tons) of waste was disposed of in U.S. landfills (EcoWeb
the Program, 1998). Starting in the 1960s and 70s,
1


ground water contamination traced to landfills was
recognized as a problem that needed to be remediated and
prevented in the future. There are a couple of mechanisms
responsible for groundwater contamination. They both
involve liquids in the landfill. The first more or less
obvious source was hazardous liquid wastes placed into
the landfill. The other source was water, either as
groundwater infiltrating into the landfill or
precipitation percolating down through the landfill. This
water leaches contaminants from solid wastes as it moves
through the landfill. This contaminant-laden water and
any liquid wastes that it mixes with is called leachate.
The leachate may move down into the water table resulting
in a plume of contaminated water. This water can in time
move to water supply wells or into rivers and streams.
The strategy to mitigate this problem is two pronged. The
first prong is regulatory. The disposal of liquid
hazardous waste is tightly controlled. Solid wastes now
must be in a leach resistant form before they can be
disposed of on land. The second prong is through
engineering. The engineering problems fall into two main
categories. First, to minimize the amount of ground water
and precipitation entering the landfill. Second, to
minimize the amount of leachate leaving the landfill.
2


Some of the engineering solutions are lowering the water
table, using synthetic or compacted soil liners and caps
to inhibit flow, and collecting and treating the
leachate. These solutions are used individually and in
combination to meet the requirements of a given site.
As previously mentioned, compacted soil is used for
landfill caps and liners. Clay is the soil of choice
because it has low permeability and also has numerous
sites available for ion exchange. A permeability of
lxlO"7 cm/sec or less is required. The ion exchange
sites naturally remove contaminants as the leachate
percolates through the clay. Compacted clay liners are
generally 0.9 to 3 meters (3 to 10 feet) in thickness
(Fetter, 1994). To date, numerous studies have been done
on the effect of chemicals and leachate on the hydraulic
conductivity of clays but little if any research has been
done as to their effect on the engineering properties of
the clay. This is important as the liners of a landfill
must support tons of waste material under conditions of
static and dynamic loading and finally support an
engineered cap.
Even small amounts of daily leakage through a liner can
potentially contaminate large quantities of ground water.
3


This leakage could take years to develop and would
continue for years. Current practices would require
remediation of the contaminated ground water, generally,
an interceptor system is constructed and the ground water
is treated above ground before being returned to the
environment. This costly activity would need to be
continued indefinitely or, until the source was removed
and the ground water remediated.
1.2 Purpose
This study was conceived to explore the overall effects
of an organic chemical contaminant on the engineering
properties of a sand-silt mixture. The purpose of the
research was threefold. First, to investigate the
engineering properties of a fine grained soil saturated
with water; second, to investigate the chemical-
mechanical influence of a liquid organic chemical
(methanol) on the state of the fine grained soil; and
third to compare the results to verify if there were any
effects to the engineering properties of the soil.
1.3 Scope of Study
The scope of this research consists of the following
principal tasks:
4


1. Performing a review of published literature to learn
about previous research relating to permeability
testing of soils contaminated with organic chemicals
and the results of such testing and results of strength
test on soils contaminated with organic chemicals if
any.
2. Selecting a soil for testing. Determining Atterberg
limits, grain size analysis, specific gravity of soil
solids, moisture-dry unit weight relationships
(compaction tests), and to classify this soil. An
isotropic consolidation test was also performed on the
soil.
3. Performing Atterberg limits on minus 40 sieve soil
using methanol instead of water.
4. Performing five isotropically consolidated, undrained
compression tests on the soil. Conducting two of these
tests on soil specimens contaminated with an organic
chemical. The organic chemical was introduced to the
soil by leaching in order to simulate in-situ
conditions. A comparison of the results of these test
will show changes, if any, in the engineering
properties of the soil attributable to the presence of
the organic chemical.
5


The results of this research are discussed in detail in
the following chapters. The plan of this study was to
develop and prove methods for performing research in this
area. Specifically, to identify any effect attributable
to methanol which was used as a representative organic
chemical. It is hoped that this and further research will
add to the understanding of overall effects chemicals and
leachate on compacted soil landfill liners.
6


2. Literature Review
2.1 Introduction
This study falls into the general discipline in civil
engineering often known as geo-environmental engineering.
Brumund (1995) noted that beginning in the 1970's,
legislation was enacted to protect the environment. The
U. S. Environmental Protection Agency (EPA) was founded
to develop, implement and enforce regulations to meet the
reguirements of the law. This created a need where none
had existed before. A new discipline within civil
engineering developed in response to the environmental
regulations being put into place. The application of
engineering fundamentals to environmental issues has
advanced the areas of groundwater flow and solute
transport in particular. He also noted that the
acceptability of a given solid or hazardous waste
disposal site from the standpoint of regulatory agencies
in the United States is usually based on the thickness of
a barrier material and the coefficient of permeability,
k, of the material.
EPA has set the maximum acceptable value for the
coefficient of permeability of 1 x 107 cm/sec. From
7


Darcy's law, the superficial velocity, v, is the product
of the coefficient of permeability and the hydraulic
gradient, i. Brumund (1995) emphasizes that the average
seepage velocity is actually higher than the calculated
superficial velocity because the flow paths through the
soil include only cross-sectional area of the pores.
2.2 Permeability Measurement Techniques
With the importance placed on the value of the
coefficient of permeability by regulatory agencies, the
measurement of this parameter has taken on increased
importance. There are two primary methods used in
determining k in fine-grained soils. The first is the
falling head method. Falling head method has two
disadvantages. First, steady state conditions are never
reached because head is always changing. Second,
volumetric flow rate, Q, is measured which decreases
accuracy. It is a relatively simple test and the cost of
the necessary equipment is reasonable. The second method
is with the use of a flow pump system. The equipment cost
is significantly higher for this method which has had the
effect of limiting its use.
A flow pump system generally consists of a flow pump,
flexible-wall or fixed-wall permeameter, pressure
8


regulation equipment and measurement equipment. All flow
pumps operate using basically the same principle. A motor
drives a worm screw which in turn drives a piston or
syringe plunger at a constant velocity along a single
axis. The liquid is thereby delivered at a constant
volumetric flow rate. A gear box between the motor and
worm screw and a rheostat to control motor speed give the
flow pump a wide range of deliverable flow rates. A main
advantage of the flow pump is that it is much easier to
precisely control volumetric flow rate than to measure it
accurately (Aiban and Znidarcic, 1989). They also noted
that a flow pump has the added benefit of providing
isolation between laboratory personnel and the hazardous
chemicals with which they are working.
The advantages of using flexible wall permeameters is
that it allows for back pressure saturation of the soil
specimen, allows the control of effective stresses and
reduces the potential for side-wall leakage (Redmond and
Shackelford, 1994). The additional advantage key to this
study is the use of a triaxial cell as the permeameter
which allowed undrained shear testing after the
permeability testing was completed. One key component is
the membrane in which the sample is confined. Latex is
used in most applications. Daniel et al. (1984) reports
9


that some chemicals, particularly organic solvents with
low dielectric constants, will destroy latex membranes.
Attempts to protect latex membranes have included coating
the latex membrane with silicon grease, and wrapping the
specimen in teflon tape. Thick butyl membranes have also
been substituted for the latex membranes. Only the teflon
tape has proven adequate.
When conducting permeability tests in this study, the
effective stress at the effluent end of the sample and
the volumetric flow rate, Q, of the flow pump are
controlled. A volumetric flow rate must be selected such
that the flow induced pore pressure does not exceed the
total stress. This is necessary in order to ensure
contact between the specimen and the membrane over the
full length of the specimen (Daniel et al., 1984).
According to Aiban and Znidarcic, the pressure gradients
in nature are generally quite low and higher imposed
hydraulic gradients produce a large variation of
effective stresses within a sample. High hydraulic
gradients may also cause local swelling making the sample
less homogeneous. By definition when swelling occurs,
there is an increase in void ratio. The delta in the
volume of voids is filled with pore liquid. The only
10


sources of this liquid inflow are from the flow pump or
back flow from the back pressure system. This introduces
some uncertainty as to what value of Q (based on total
inflow or outflow of the sample) to use when calculating
the value of k. When swelling occurs, the length and
width of the specimen change. Ideally, hydraulic
conductivity should be calculated at steady state. In
other words, while inflow and outflow are equal.
Lastly, Daniel et al. (1984) provides a useful, detailed
procedure for setting up a specimen for testing using the
flow pump method.
2.3 Effects of Chemicals on Clay Hydraulic Conductivity
A significant amount of research has been performed on
the effects of chemicals on the hydraulic conductivity of
several natural clays. Bowders and Daniel (1987) reported
on the use of methanol, acetic acid, heptane and
trichloroethylene in permeability testing of kaolinite
and illite-chlorite soils. Four pure, reagent grade
chemicals were used in the tests. The methanol and acetic
acid were mixed with distilled water in 20, 40, 60, and
80 % concentrations by volume. Both methanol and acetic
acid are miscible in water. A hydraulic gradient, i, of
11


200 was used for all of the tests. Each test was ended
after two pore volumes of permeant liquid had passed
through the specimen and the total organic carbon (TOC)
concentration in the effluent was approximately equal to
the TOC concentration in the influent. Of the various
concentrations of methanol tested, they reported that the
only significant increase in hydraulic conductivity
occurred with pure methanol permeant. The hydraulic
conductivity increased to 1.2 times that of water in a
test using a flexible wall permeameter.
A test using a rigid wall permeameter also conducted with
pure methanol as the permeant. Bowders and Daniel (1987)
reported a 44 fold increase in hydraulic conductivity.
They attributed the increase to the methanol, with a
dielectric constant of 33.6, collapsing the diffuse
double layer around the soil particles which resulted in
sample shrinkage. This caused cracking and sidewall
leakage. Furthermore they attributed the reduced increase
in hydraulic conductivity observed in the flexible-wall
permeameter to the confining pressure holding the
membrane against the specimen thus preventing sidewall
leakage and the effective stress providing resistance
against the development of cracks. They also determined
12


the liquid limit (LL) and plastic limit (PL) for both
types of clay using water and then all four of the
organic chemicals. Changes of +6 % to -11 % were noted.
They concluded that if the Atterberg limits were altered
by a neutral organic liquid then the hydraulic
conductivity would be adversely affected.
Foreman and Daniel (1986) reported the results of tests
conducted on kaolinite, hoytville clay and lufkin clay.
The clays were mixed with distilled water to a point
close to the optimum moisture content and then stored for
at least 24 hours to hydrate before compaction. Atterberg
limits were performed on the three clays mixed with
water, pure methanol and pure heptane. They reported that
the organic chemicals reduced or eliminated the
plasticity of the soils. Permeability tests were
conducted using, compaction-mold cell (fixed-wall),
consolidation cell (fixed-wall) and flexible-wall
permeameters. The flexible-wall tests were performed at
an average effective stress of 105 kPa (15 psi) at a back
pressure of 276 kPa (40 psi). The samples were allowed to
consolidate at these pressures for about one week to
saturate. The permeability tests were continued until
two pore volumes of permeant liquid had passed through
13


the specimens. They reported virtually identical
hydraulic conductivities with all three permeameter types
when water was the permeant liquid. When organic
chemicals were the permeant liquids the hydraulic
conductivities determined using the two fixed-walled
permeameter methods were higher than that determined
using a flexible-walled permeameter with hydraulic
gradients ranging from 2 to 300. This was attributed to
lower effective stresses and sidewall leakage in the
fixed-walled permeameter type caused by shrinkage of the
soil specimen. Finally, they suggest that clays subjected
to higher confining stress have reduced susceptibility to
changes induced by organic chemicals.
Fernandez and Quigley (1985) conducted high-gradient,
constant flow rate tests on a natural soil from Sarnia,
Ontario. A wide variety of pure liquids, including
methanol, were used in a series of permeability tests
using a fixed-walled permeameter with an inner diameter
of 5.38 cm. A rigid spacer prevented sample swelling in
the vertical direction. The specimens were prepared in
three lifts of approximately 8 mm each using a Harvard
miniature tamper with the soil wet of optimum. They
indicated that the excessive kneading of the soil as it
14


was prepared eliminated macropores and channels. Again,
two pore volumes of permeant liquid were leached through
the specimens at hydraulic gradients from 25 to 500.
These high hydraulic gradients are probably far in excess
of those that would probably be experienced in the field.
Ten-fold increases of hydraulic conductivity as compared
with water were reported with water-soluble alcohol
permeants.
In subsequent research, Fernandez and Quigley (1991)
explored the role effective stresses play in preserving
low hydraulic conductivities in a clay barrier. Again the
research was conducted on a natural clay from
southwestern Ontario permeated with various landfill
leachate mixtures and water-soluble organic chemicals
(ethanol and dioxane) at effective stresses from 0 to
~300 kPa (44 psi). The specimens were prepared using the
same method that has been previously described. The tests
were performed using a constant-flow rate, forced flow,
fixed ring permeameter system. The system was designed
with springs applying static vertical confining stress to
the porous stone on the top of the specimen. A
disadvantage of this method was that up to a 12 percent
reduction in the confining stress could occur due to the
15


relaxation of the springs as the volume of the sample
decreased. Up to 1000-fold increases in the coefficient
of permeability occurred when ethanol and dioxane
concentrations exceeded 70 percent of the permeant
mixture in unconfined tests. The results of their tests
showed the application of vertical effective stress prior
to permeation with chemicals reduced or eliminated the
large increases in hydraulic conductivity observed at
zero stress. Average vertical effective stresses of 70
kPa (10 psi) for ethanol and above 200 kPa (29 psi) for
dioxane were required to negate the increases in
hydraulic conductivity that occurred at zero effective
stress. Tests were also conducted to determine if static
vertical effective stress applied after permeation and
the resultant consolidation could reverse the increase in
hydraulic conductivity caused by the organic chemical.
For ethanol, 100 kPa (14.5 psi) was required. This was
about four times the original vertical effective stress
applied. Application of vertical effective stresses was
ineffective in reducing hydraulic conductivity after
permeation with dioxane suggesting greater difficulty in
closing the macropore flow paths that were created in the
reaction between the clay and the dioxane.
16


Hueckel et al. ( 1996) Reported that strains in clayey-
soils with low clay contents are affected by chemicals in
a way that is independent from the way permeability
changes are induced by chemicals. They suggest that in
low clay content soils, the course grains are in contact
with each other providing a skeleton that is responsible
for stress transmission and thus strain. The clay
fraction fills the voids between the contact points of
the course grains and is the primary factor controlling
the permeability. They concluded that in soils with high
clay contents, stress was strongly coupled with
permeability changes. This was the result of the absence
of the course-grained skeleton transmitting the stress,
which would in effect stiffen the specimen. In the soils
with lower clay content, the clay in the inter-granular
voids was somewhat sheltered from the stress transfer
mechanism.
Four of the papers reviewed in this section attributed
the larger increases in hydraulic conductivity observed
in fixed-wall as compared the flexible-walled
permeameters. This is due to shrinkage cracks and
sidewall leakage that occur as the specimen shrinks in
fixed-walled permeameter as a result of the organic
17


permeant leaching through it at essentially zero
confining stress. In theory, the shrinkage is caused by
the collapse of the diffuse double layer around the soil
particles caused by organic chemicals. The lower the
dielectric constant of the chemical, the larger the
increase in hydraulic conductivity. Fernandez and Quigley
(1985) used electron photomicrographs to show that
flocculation occurred when clay was molded with ethanol
and benzene. This suggests that the organic chemicals
cause changes in the soil fabric.
2.4 Effects of Organic Chemicals on Soil Strength
Anandarajah and Zhao (1996) reported on tests conducted
on commercial kaolinite. The stated goal of the research
was to gain a comprehensive understanding of the effects
of a variety of contaminants on the geotechnical
properties of kaolinite. The pore liquids tested included
two salts and ten organic chemicals, with methanol being
one. The specimens were prepared using two methods. In
the first method, a slurry with the desired pore fluid
was consolidated one-dimensionally. The second method
called for the preparation of a slurry using deionized
water. The desired permeant was then leached through the
specimen in a triaxial cell. The paper was very brief and
18


did not report any methanol test results. They did
confirm that organic chemicals, heptane in particular,
have an effect on the geotechnical properties of
kaolinite. They also noted that the effects of organic
chemicals were significantly larger than those caused by
the aqueous salt pore liquids.
Moore and Mitchell (1974) describe an analytical
technique expressing the electromechanical interaction
among the constituents of a soil system in terms of
physical properties of the individual components of the
soil system. The details of this technique will not be
discussed here. They also conducted a series of strength
tests on kaolinite clay leached with seven infinitely
miscible organic chemicals. Methanol was not one of the
chemicals tested. All of the specimens were prepared
using a 0.1 N aqueous solution of potassium chloride.
This ensured that the initial fabric of the soil was
determined by water. The specimens were then leached over
a period of six months with distilled water or the
desired pure, organic chemical. Triaxial tests and vane
shear tests were then conducted on the specimens. They
concluded that pore liquids other than water could be
used to improve the strength characteristics of a soil.
19


They speculated that soil initially mixed with pore
fluids other than water might exhibit different strength
characteristics from those determined in their study.
2.5 Applications
There are a couple of other considerations in the
application of clay soil as landfill liners. Stulgis et
al. (1995) described a site study for a vertical
expansion over an existing municipal solid waste
landfill. A liner and other environmental systems were to
be installed between the existing landfill and the
expansion. The primary concern addressed in the study was
the estimation of settlement of the material in the
existing landfill as the refuse decomposes and as
additional loading occurs when waste materials are placed
in the new landfill above. The settlement estimates were
used to develop a predicted settlement contour map. This
aided in the design of the landfill liner, leachate
collection and gas collection systems to safely
accommodate the expected settlement.
Mitchell et al. (1995) addressed material interactions
within landfills. They include waste-soil, geosynthetic-
soil, leachate-soil, leachate-geosynthetic and soil
20


geosynthetic interactions. At present little information
is available regarding these interaction and their
variation over time, but it is vital that each component
continue to perform its intended function in a
satisfactory manner. It is necessary to design compacted
clay, geomembranes and/or geosynthetic clay liners that
will resist biological and chemical attack, accommodate
settlement, resist excessive deformation under static and
dynamic loading that could breach leachate or gas
collection systems and meet all applicable local, state
and federal regulatory requirements.
There are a couple mechanisms singly or in combination
that could cause excessive deformation. Often landfills
are constructed in less desirable areas. This could
include areas possessing marginal foundation materials.
Local variations in the unit weight of the waste
materials placed above a liner could cause differential
settlement. As mentioned previously, excessive settlement
could breach gas and leachate collection systems. It
could also cause localized breaches of the liners and/or
membranes.
21


3.
Materials
3.1 Soil
3.1.1 Selection
A mixture of palygorskite clay and washed sand was
selected for study. This is a continuation of previous
work conducted at the University of Colorado. D. K.
Alierton conducted flow tests on three sand/palygorskite
clay mixtures containing different clay fractions. He
prepared the specimens by adding 10, 20, and 30 percent
palygorskite clay, by weight, to the sand (Alierton,
1995). The 20 percent by weight addition of palygorskite
clay to sand was chosen for this study. The two component
soils that make up the sand/clay mixture are discussed
individually in the following two sections.
3.1.2 Palygorskite Clay
The palygorskite clay is grayish-white, and fine-grained.
The Milwhite Mining Company of Attapulgus, Georgia
produces it. Palygorskite is also known as attapulgite
and sepiolite (Alierton, 1995). A summary of Atterberg
limits and grain size distribution test results of the
palygorskite clay are shown in Table 3.1. Palygorskite
22


clay powder in the form it is received from the vendor is
shown in figure 3.1.
Table 3.1 Result of Index Property Tests of
Palygorskite Clay (Chang, 1997)
Soil Property Value
Liquid Limit (LL) 263.9 %
Plastic Limit (PL) 118.7 %
Plasticity Index (PI) 145.2 %
Specific Gravity (Gs) 2.70
Percent Passing No. 200 Sieve 100 %
% finer than 5 p. 97 %
% finer than 2 (j. 90 %
Activity 1.61
USCS Soil Classification MH
3.1.3 Sand
The sand fill, listed as washed sand, was obtained from
Ainsworth Rock Sales, Inc., in Thornton, Colorado. The
sand was initially air dried and stored in sealed,
19.4-liter (5-gallon) plastic buckets until use. A
mechanical grain size analysis was performed to determine
the classification of the sand. From the plot of the
results the coefficient of curvature, Cc, and the
coefficient of uniformity, Cu, were determined to be 0.98
and 3.9 respectively.
23


Figure 3.1 Photograph of Palygorskite Clay
24


Figure 3.2
Photograph of Air-Dried Sand
25


According to the Unified Soil Classification System
(USCS), this sand is classified as poorly graded sand
(SP). The air-dried sand is shown in Figure 3.2.
3.1.4 Palygorskite Clay and Sand Mixture
The soil studied was a mixture of palygorskite clay and
sand. The mixture was prepared by mixing five parts by
weight of air-dried sand with one part by weight of
palygorskite clay. Five kg of sand and 1 kg palygorskite
were sealed 19.4-liter plastic bucket. Mixing was
accomplished by rolling, inverting and shaking the bucket
and its contents for a minimum of 10 minutes. Figure 3.3
shows the dry sand and palygorskite clay mixture.
3.1.4.1 Soil Properties
A battery of laboratory tests was conducted to classify
and characterize the sand/clay mixture. A summary of the
laboratory procedures used is shown in Table 3.2. The
results of these tests are summarized in the Table 3.3
below. Atterberg limits tests were performed on the minus
40-sieve fraction of the sand clay mixture. The moisture
content-dry unit weight relationship was determined using
standard Proctor test and the results are shown in Figure
3.4.
26


Figure 3.3 Photograph of Sand and Palygorskite Mixture
27


Table 3.2 Summary of Laboratory Procedures Used
Soil Property Procedure
Maximum Dry Density ASTM D 698-78 (Standard Proctor)
Optimal Moisture Content ASTM D 698-78 (Standard Proctor)
Liguid Limit (LL) ASTM D 4318 (method B)
Plastic Limit (PL) ASTM D 4318
Plasticity Index (PI) ASTM D 4318
Specific Gravity (Gs) ASTM D 854
Mechanical Grain Size Analysis ASTM D 422
USCS Soil Classification ASTM D 2487-85
Table 3.3 Standard Proctor and Atterberg Limit
Test Results on Sand-Palygorskite
Mixture
Soil Property Value
Maximum Dry Density 16.9 kN/m3 (107.6 pcf)
Optimal Moisture Content 16.9 %
Liquid Limit (LL) 83 %
Plastic Limit (PL) 48 %
Plasticity Index (PI) 35 %
Specific Gravity (Gs) 2.69
Percent Passing No. 200 Sieve 15.5 %
USCS Soil Classification SM
28


19.00 -
n
<
E
u>
0)
s
c
3
>
w
o
# Dry Unit Weight
B Zero Air Voids
Water Content, w%
Figure 3.4 Moisture-Dry Weight Curve for Sand-
Palygorskite Mixture
Particle Diameter, mm
Figure 3.5
Sand-Palygorskite Mixture Grain Size
Distribution
29


Atterberg limits were also determined using reagent grade
methanol instead of water. The Liquid Limit, LL, was
48.7 %. The soil mixture was non-plastic when saturated
with methanol.
3.2 Methanol
3.2.1 Selection
Methanol was selected as the chemical for use in this
study. It was selected for several reasons. First, as
with the sand clay mixture, Allerton had used methanol in
his research. Secondly, methanol is a good representative
of an organic chemical that is miscible in water. Lastly,
methanol is widely used and released to the environment
in the United States. In 1990, total U.S. production of
methanol was 7.99 billion pounds. In 1989, methanol
ranked third highest of 322 chemicals that comprise the
national Toxic Release Inventory (National (Safety
Council, 1998). Table 3.4 shows the breakdown for
methanol's release to the environment.
3.2.2 Description and Properties
Methanol, also known as methyl alcohol and wood alcohol,
is a clear, colorless liquid. Methanol is used in the
production of formaldehyde, acetic acid, and as a
30


Table 3.4
Methanol Released to the Environment in
the United States in 1989 (National Safety
Council, 1998)
Where Released Amount Released (millions of pounds)
Into Air 199.7
Into Public Sewage 109.0
Off-site 51.3
Underground 23.9
In Water 17.2
Onto Land 7.8
solvent. In recent years, methanol has been increasingly
used to produce methyl tert-butyl ether (MTBE) and oxinol
that are used to improve gasoline octane. Methanol is
also being used as a cleaner burning gasoline substitute
(Environmental Health Center, 1997). Methanol facts are
shown in Table 3.5.
Because of methanol's hazardous characteristics the
following precautions were taken. Methanol was only
handled in a well-ventilated room. Personal Protective
Equipment (PPE) included, wearing of a face shield,
gloves and a laboratory coat.
31


Table 3.5
Methanol Information (National Academy of
Sciences, 1995)
Formula CH3OH
Physical Properties Colorless liguid Boiling Point 65 C Melting Point -98 C Miscible with water in all proportions
Odor Faint alcohol detectable at 4 to 6000 ppm
Vapor Density 1.1 (air = 1.0)
Vapor Pressure 96 mm Hg at 20 C
Flash Point 11 C
Autoignition Temperature 385 C
Major Hazards Highly flammable (NFPA rating = 3) Low acute toxicity
Toxicity Ingestion or inhalation of high concentrations can produce headache, drowsiness, blurred vision, nausea, vomiting, blindness and death. It can also be absorbed in toxic amounts through the skin.
Warning Properties Not considered adeguate
32


4. Testing Program, Equipment, and Procedures
4.1 General
The objective of the test program was to explore the
changes in the engineering behavior of the palygorskite
clay and sand mixture when exposed to methanol. The basic
approach was to conduct undrained shear strength tests
and flow tests using some specimens saturated with water
and other specimens saturated with methanol. This
approach established the (water) baseline values.
Deviation from the baseline values would identify the
effects attributable to the presence of methanol. In the
flow tests, water was pumped through the sample first to
allow baseline permeability data to be collected.
4.2 Testing Program
The testing program consisted of four major components.
4.2.1 Soil Classification and Characterization
Atterberg limits, grain size distribution and other tests
were performed to characterize and classify the soil. The
results of these tests were previously discussed in
Chapter 3 (Section 3.1.3).
33


4.2.2 Isotropic Consolidation Test
An isotropic consolidation test was conducted on a soil
specimen saturated with water.
4.2.3 Undrained Shear Tests
Five Consolidated Isotropically, Undrained Compression
(CIUC) tests were conducted. The CIUC shear tests
consisted of the following three phases.
4.2.3.1 Isotropic Consolidation Phase
A desirable degree of saturation (of water) was achieved
at an effective stress of 7 to 14 kPa (1 to 2 psi). The
effective stress was then raised to the desired effective
stress for the test and allowed to consolidate.
4.2.3.2 Flow Phase
The flow phase consisted of pumping one pore volume of
water followed by two pore volumes of methanol through
the specimens. The flow Phase was conducted on only two
of the five specimens. Two different effective stresses
were selected. The flow phase was used to study the
effect of methanol on hydraulic conductivity. This also
provided soil specimens saturated with methanol for the
CIUC tests. Methanol induced changes in the volume of the
sample were also recorded. Table 4.1 shows the flow phase
control mode.
34


Table 4.1 Flow Phase Control Mode
Controlled (constant) Measured (variable)
Effective stress at the bottom of the sample, oB Pore pressure difference across the sample, Au
Volumetric flow rate, Q(1) Volume Change, Av
Note:
(1) The volumetric flow rate of the pump is constant. Volumetric
flow rate through the sample is constant only when the volume
of the sample is not changing.
4.2.3.3 Undrained Shear Phase
CIUC shear tests were conducted on five soil specimens.
Three of these were conducted on soil specimens saturated
with water. The remaining two shear tests were conducted
on the two specimens saturated with methanol. A summary
of the CIUC test program is shown in Table 4.2. Table 4.3
shows the controlled and variable parameters for the
undrained shear phase tests.
4.3 Equipment
4.3.1 Flow Pump System
D. K. Allerton (1996) initially developed the flow pump
system used for soil testing. A brief description of the
major components of the system designed by Allerton is
given below.
35


Table 4.2 Summary of the Shear and Flow Test Program
Test Number(1) 2 Specimen No. Consolidation Effective Stress, o'
W_22CIUC B3 22 kPa (3.2 psi)
W_187CIUC B5 187 kPa (27.1 psi)
M_52CIUC B6 52 kPa (8.1 psi)
M_186CIUC B7 186 kPa (31.3 psi)
W_44CIUC B8 44 kPa (6.4 PSI)
Note:
(1) The test numbers beginning with M were saturated with methanol.
Conversely, test numbers beginning with W were saturation with water.
Table 4.3 Undrained Shear Phase Control Mode
Controlled (constant) Measured (variable)
Confining Pressure, o3
Axial displacement rate, Ah/t Axial Force (rate)
Volume, V Pore Pressure, u
1. A Geotest triaxial cell was used as a flexible wall
permeameter.
2. A variable speed Harvard Apparatus model 909, positive
displacement flow pump capable of delivering constant,
low volume flow rates between 4.6xl0~6 cm3/sec to
36


4.6lxl0~1 cm3/sec. The flow pump is shown in Figure
4.1.
3. Four Geostore model S-470 bladder accumulators with
buna or viton diaphragms. Care was taken in selecting a
diaphragm material compatible with the chemical being
used. An incompatible chemical could deteriorate the
diaphragm and cause it to fail. The buna or viton
diaphragms provided a flexible, impermeable interface
between air and liquid (methanol or water) portions of
the system. The bladder accumulators were used to
provide back pressure, confining pressure, and as
reservoirs for the water and chemicals that were to be
pumped through the sample. The bladder accumulator used
to provide back pressure was also used to collect water
and methanol which had been pumped through the soil
specimen. This system effectively isolates hazardous
chemicals from the laboratory environment during flow
testing. The only occasions for exposure of laboratory
personnel occurred when filling the bladder accumulator
with a hazardous chemical, when emptying the back
pressure bladder accumulator after a hazardous chemical
had passed through the soil specimen, or when flushing
and draining the system after the completion of a test
using a hazardous chemical. Both assembled and
37


disassembled bladder accumulators are shown in Figure
4.2.
4. The system was instrumented with two Validyne
Differential Pressure Transducers (DPT) tied into a
personal computer-based data acquisition system. DPT-1
is a Validyne Model DP15-32 differential pressure
transducer with a range of 0 kPa to 138 kPa (20 psi).
DPT-2 is a Validyne Model DP15-52 differential pressure
transducer with a range of 0 kPa to 1379 kPa (200 psi).
DPT-1 measured the difference between pore liquid
pressure at the top and bottom of the sample during the
flow phase of the tests. DPT-2 was used to measure the
pressure difference between the cell pressure and pore
liquid pressure at the bottom of the soil sample. Cell
pressure is also called confining pressure. The
difference between the cell pressure and pore liquid
pressure at the bottom of the soil sample will be
referred to as effective stress hereafter. This is
justified by the assumption of nearly full saturation
as indicated by the satisfactory B parameters obtained
(See Table 5.4). DPT-1 and DPT-2 are shown in Figure
4.3.
38


As a new function needed for this thesis, both DPT-1
and DPT-2 were used for measurements during the
undrained shear strength phase. DPT-1 was used to
measure pore pressure and DPT-2 measured effective
stress.
These components were connected using 3.175 mm,
stainless steel tubing, and both two and three way-
valves. The system was designed to allow flow through
the sample from top to bottom or vice versa. The flow
tests could also be conducted in two modes. The first
test mode is with constant effective stress at one end
of the sample (drained). This is referred to as a
constant effective stress test. The second test mode is
with constant sample volume (undrained). All flow tests
conducted for this study were constant effective stress
tests.
5. The personal computer is an IBM 286 model with Validyne
model UPC-601-L LVDT and VRDT sensor interface card
installed. The Validyne Easy Sense software package
was used for sensor system management and data logging.
39


40


41


Figure 4.3 Photograph of Differential Pressure
Transducers DPT-1 and DPT-2
42


Several modifications were made to the Flow Pump system
received from Allerton. They are described below.
1. A 50 ml, bottom loading burette instrumented with a
Validyne model DP15-30 differential pressure transducer
(DPT-3) with a range of 0 kPa (0 psi) to 8.62 kPa (1.25
psi) to measure volume changes in the sample during
constant effective stress tests. DPT-3 was then
connected to the data acquisition system. DPT-3 was
used to measure the difference between the air pressure
and the air pressure plus the water pressure of the
water column. The air pressures on either side of the
diaphragm canceled each other out,"'thereby, allowing
the DPT to measure the pressure due to the water column
directly. The pressure of the water column was
directly proportional to the height of the water column
and also directly proportional to the volume in the
calibrated burette.
2. The volume change of a sample was determined indirectly
by measuring the volume of water that flows into or is
displaced from the cell pressure port. This
modification eliminated the need for the bladder
accumulator that had been used originally to provide
confining pressure. This bladder accumulator was
43


removed from the system. Cell pressure was directly
connected to the burette described above, via the cell
pressure port. Figure 4.4 shows the volume change
monitoring system. Initially, DPT-1 had a range of 0
kPa to 138 kPa (20 psi) which was not large enough to
permit the direct measurement of either back pressure
or cell pressure.
The 0 kPa to 138 kPa (20 psi) stainless steel diaphragm
was replaced with a 0 kPa to 862 kPa (125 psi)
stainless steel diaphragm. This permitted direct
measurement of both back pressure and confining
pressure for the remainder of the test program. This
change allowed both flow phase and shear phase testing
at higher effective stresses. The change also
permitted flow testing at larger pressure gradients.
Figure 4.5 is a schematic drawing of the flow pump
system, after modification, as it was used for the flow
phase of the test program. Figure 4.6 shows the portion
of the system used during the undrained shear phase.
44


45


Flow Pump
Figure 4 5 Schematic Drawing of the Modified Flow Pump System


Two way valve
9
Three way valve

r jn
Q>
Confining
Pressure
Regulator
Air-----(y
Burette
0
-d>

Figure 4.6 Schematic Draw ing of the Portion of Flow Pump System Used in
the Undrained Shear Test Phase


4.3.2
Loading Frame
Undrained shear tests were conducted using a 89 kN (20
kips-force) capacity hydraulic loading frame Material
Test System (MTS) Model 810 manufactured by Material
System Corporation of Minneapolis, Minnesota. Load and
vertical displacement output data from the loading frame
were tied into the data acquisition system. DPT-1 and
DPT-2 from the flow pump system were used to measure pore
pressure, u, and horizontal effective stress, <33',
respectively during the undrained shear phase.
Additionally, the axial displacement, Ah, and the force,
F, applied on the specimen by the loading ram of the MTS
were input from the MTS and recorded using the same PC
based data acquisition system.
4.4 Procedures
4.4.1 Sample Preparation
The test specimens were prepared two percent wet of the
Optimum Moisture Content (OMC) of 16.8 percent determined
using the Standard Proctor Compaction Test. Water was
mixed with the soil to increase the moisture content, w,
from the 1.8 percent initial content to approximately
18.8 percent. A precise mass of water was added to the
soil mixture incrementally using a spray bottle. As the
48


water was added, the soil mixture was kneaded by hand.
After all the water had been added and thoroughly mixed
in, the soil was sealed in a plastic bag for a minimum of
24 hours to cure.
The soil was then compacted into a 50.8 mm (2 inch)
diameter, 101.6 mm (4 inch) tall split mold. The mold and
compaction tool are shown in Figure 4.7. The soil was
compacted in three lifts of equal mass and height using
the MTS Model 810 loading frame to provide the compactive
effort. The mold was disassembled and the sample
extruded. Specimens of the sand/clay mixture were
prepared using this method for the isotropic
consolidation test and the five CIUC tests. Table 4.4
shows the relative compaction (R.C.) of the soil
specimens.
Table 4.4 Relative Compaction of Soil Specimens
Sample Isotropic Consolid- ation W_22 W_44 W_187 M_52 M_186
w0 18.5 % 18.1 % 18.0 % 17.7 % 17.8 % 17.9 %
R.C. 94.0 % 91.4 % 94.0 % 94.0 % 94.5% 93.9 %
49


Figure 4.7 Photograph of Compaction Tool and Mold
50


The samples were then carefully wrapped in plastic wrap,
then masking tape. The wrapped specimens were then coated
with paraffin to get an airtight seal. The soil specimens
were then placed in a commercially available, insulated
cooler for storage until needed. Lastly, the cooler was
stored in a climate controlled, soil specimen storage
room.
When a soil specimen was required for a test, the
paraffin and tape were pealed off. The plastic wrap was
then carefully removed. The mass of the specimen was
determined and recorded. The specimens were stiff enough
that they were not easily damaged if handled gently. The
soil sample was placed on top of filter paper and a
stainless steel porous stone that had been soaked in de-
aired water. The sample, filter paper, and porous stone
were all resting on the bottom platen of the triaxial
cell that had been previously flushed with de-aired,
demineralized water. A second, presoaked filter paper and
stainless steel porous stone were placed on top of the
sample. A membrane jacket was used to stretch a latex
membrane to a large enough diameter so it could be
lowered over the soil specimen. The vacuum source to the
membrane jacket was then isolated and the membrane
allowed to shrink around the sample. The membrane
51


stretcher was removed and the top platen was placed on
top of the upper porous stone. The membrane was now
secured to both the top and bottom platens using two
rubber o-rings at each platen. The height and diameter of
the sample were then measured at three locations each.
The vented triaxial cell top was then mated to the
triaxial bottom cell and the fastening rods screwed down
to finish assembly of the triaxial cell. Lastly, the
triaxial cell was filled with water. When full of water,
the vent was shut.
As a result of the method used to prepare the test
specimens, each on average had an air volume, Va, of
about 22 cm3. Confining pressure was maintained 7 kPa to
14 kPa greater than back pressure as back pressure was
raised to 414 kPa (60 psi). Both the confining pressure
and back pressure were raised in turn at 7 kPa
increments. The high back pressure was used to help
ensure that the air in the sample desolved into the water
allowing the sample to be as close to fully saturated as
possible. The sample was allowed to stand at these
pressures for several days to saturate. The confining
pressure was then raised to increase the effective stress
to the level dictated by the testing program. The sample
52


was allowed to consolidate for a day or two. Volume
change measurements were taken during consolidation.
Skempton's B parameter was measured to check on the level
of saturation being achieved. Skempton's B Parameter was
determined using the following equation (Bowles, 1992):
B =
A u
Act3
(4.1)
Where: Au = Change in back pressure/pore pressure
Ao3 = Change in confining pressure
When the B parameter was steady or was 0.95 or greater,
the isotropic consolidation phase, flow phase, or
undrained shear phase were conducted as per the testing
program.
4.4.2 Isotropic Consolidation Test
An isotropic consolidation test was conducted using a
triaxial cell and a single pressure transducer. A 4-way
valve allowed the selection for measurement of confining
pressure, pore pressure at the top of the sample, or pore
pressure at the bottom of the sample. Figure 4.8 shows
the isotropic consolidation testing system.
53


As with the other tests, back pressure was incrementally
raised to 414 kPa (60 psi) to ensure the sample was as
fully saturated as possible. Confining pressure was
maintained 34.5 kPa (5 psi) greater than the back
pressure. The sample was then allowed to saturate.
Skempton's B parameter was measured every couple of days.
When a steady B parameter of 0.93 was obtained, the
consolidation test was begun. For the remainder of the
test, back pressure was held constant at 414 kPa (60
psi). The test was conducted by raising confining
pressure every 24 hours for 5 days and then lowering the
confining pressure in the reverse order using the same
increments every 24 hours for the next 5 days. Table 4.5
shows the mean effective stress for each day of the
isotropic consolidation test.
The volume change induced by the pressure changes were
measured at the 0.1, 0.25, 0.5, 1, 2, 4, 8, 15, 30, 60,
120, 240, 480, and 1440 minute points after the confining
pressure was either increased or decreased. The volume
change of the sample was measured directly using a
burette connected to the back pressure port of the
sample. This measured the amount of water flowing into
54


Figure 4.8 Photograph of Isotropic Consolidation Testing
System
55


or out of the specimen. These readings were manually
recorded for later analysis.
Table 4.5 Isotropic Consolidation Testing Program
Day Mean Effective Stress, p'
Day 0 34 kPa (5 psi)
Day 1 69 kPa (10 psi)
Day 2 138 kPa (20 psi)
Day 3 276 kPa (40 psi)
Day 4 483 kPa (70 psi)
Day 5 689 kPa (100 psi)
Day 6 483 kPa (70 psi)
Day 7 276 kPa (40 psi)
Day 8 138 kPa (20 psi)
Day 9 69 kPa (10 psi)
Day 10 34 kPa (5 psi)
4.4.3 Flow Phase
Constant effective stress flow tests were conducted on
two samples at effective stresses of 56 kPa (8.1 psi) and
219 kPa (31.8 psi). The three DPTs were used to measure
the pressure difference across the sample, Au, radial
effective stress at the bottom of the sample, 03', and
56


changes in the volume of the specimen, AV. For these
tests, the direction of flow was from the top of the soil
sample toward the bottom of the sample. The back pressure
was controlled at the bottom of the sample. The
volumetric flow rate of the pump, Q, was set at 3.462E-4
cm3/sec. There was a brief (three to seven minute)
interruption daily to refill the flow pump.
The flow phase was conducted by pumping one pore volume
of water followed by two pore volumes of methanol through
the soil sample. After the three pore volumes had been
pumped through the specimen, flow was stopped. The sample
was left a minimum of 24 hours to equalize to a steady
state condition without flow. Drainage was allowed during
equalization. Skempton's B parameter was measured. When
the B parameter was 0.95 or greater, the triaxial test
was conducted according to the procedure described in
following section.
4.4.4 Undrained Shear Tests
After the flow portion of the test, the methanol
saturated specimens were isolated and disconnected from
the flow pump portion of the system. With the sample
isolated from everything except the DPTs, volume changes
were no longer possible. At this point, the measurement
57


of volume changes was discontinued. The remainder of the
system which included the triaxial cell, DPT-1, and DPT-2
were then carefully moved to the loading frame using a
cart. The triaxial cell was then positioned into the
loading frame for the triaxial test.
The three water undrained shear tests were set up without
using the flow pump. However, the triaxial cell, DPT-1,
and DPT-2 from the flow pump system were used during
these triaxial tests. As with the methanol saturated
specimens, the triaxial cell was placed into the loading
frame for the undrained shear test.
The data acquisition system was plugged into the Y-i and
Y2 inputs of the loading frames chart recorder to record
axial displacement of the loading ram and the force
applied. The Validyne Easy Sense software was configured
to log the date, time, u, 03', F, and Ah. With the
triaxial cell sitting on the loading frame's load cell
and the loading ram gently placed in contact with the top
platen of the triaxial cell, the offsets in the Validyne
software package were set to zero the force, F, and the
axial strain, ea. The software was then placed in the
58


logging mode to record test data at preset intervals.
The triaxial test was then begun.
The undrained shear tests were all conducted at a
constant axial strain rate of 1.862 percent per hour. The
loading frame was set to produce 15 percent axial strain
or 15.24 mm (0.60 inches) axial displacement in
approximately 8 hours. During the test, horizontal
effective stress, 03', pore pressure, u, axial
displacement, Ah, of the loading ram, the force applied,
F, date, and time were recorded in a text file.
4.4.5 Equipment Calibration
4.4.5.1 Differential Pressure Transducers
Differential pressure transducers DPT-1 and DPT-2 were
calibrated using another pressure transducer and high
precision gauges that had been previously calibrated.
First, the high precision pressure gauges were checked
against the pressure transducer. With these instruments
in agreement, each of the DPTs were checked from zero to
the top of their ranges. As the pressure was raised
incrementally, the pressure readings from the precision
gauges and the corresponding low voltage output from the
DPTs were recorded. Linear regressions were performed to
determine the slope of the calibration lines. The slopes
59


were entered into the EasySense Software allowing the low
voltage generated by the DPT to be converted to the
corresponding differential pressure reading.
4.4.5.2 Volume Change Differential Pressure Transducer
Differential pressure transducer DPT-3 was calibrated
using a bottom loading, 50 ml burette. As the water level
in the burette was raised in 10 ml increments, the level
readings of the burette and the corresponding low voltage
output from the DPT were recorded. A linear regression
was performed to determine the slope of the calibration
line. The slope was entered into the EasySense Software
allowing the low voltage generated by the DPT to be
converted to the corresponding burette reading over the
range of 0 to 50 ml.
4.4.5.3 Loading Ram Axial Displacement
The axial displacement of the loading frame's loading ram
was calibrated using a 0 to 25.4 mm (0 to 1 inch) dial
indicator with 0.0254 mm (0.001 inch) graduations. As the
vertical position of the loading ram was changed
manually, the dial indicator reading and the
corresponding low voltage output from the Material Test
System's chart recorder were recorded. Again, a linear
regression was performed to determine the slope of the
60


calibration line. The slope was entered into the
EasySense Software allowing the low voltage generated for
the Material Test System's chart recorder to be converted
to the corresponding axial displacement of the loading
ram.
4.4.5.4 Load Cell
The loading frame's load cell was calibrated using a set
of weights normally used for oedometer tests. The
weights, singly and in combination, were placed on the
loading frame's load cell. The total weight and the
corresponding low voltage output for the Material Test
System's chart recorder was recorded. Once again, a
linear regression was performed to determine the slope of
the calibration line. The slope was entered into the
EasySense Software allowing the low voltage generated for
the Material Test System's chart recorder to be converted
to the corresponding force being applied to the test
specimen and being measured by the load cell.
61


5.
Test Results
5.1 Data Analysis
The data from each test was transferred to a personal
computer for data reduction. The data, in the form of
text files, were imported into Microsoft Excel
spreadsheets. All computations were performed and the
results plotted. The equations used to convert the raw
data for interpretation are detailed in the following
sections.
5.1.1 Flow Test Data Analysis
The coefficient of permeability, k, is a constant of
proportionality relating hydraulic gradient and discharge
velocity. Darcy's law is given as (Holtz and Kovacs,
1981) :
v = ki (5.1)
The corresponding flow rate (or quantity per unit time)
is
Q = kiA
(5.2)
62


Where Q = quantity (volume) of fluid flow in a unit of
time.
k = Coefficient of permeability, or hydraulic
conductivity, in velocity units.
i = hydraulic gradient (Ah/L), head loss, Ah,
across a flow path of length L.
Ah = Total head difference across the flow path
of length L.
L = length of the sample or flow path.
A = Cross-sectional area of the soil sample
perpendicular to the direction of flow.
Two classical methods used in determining the coefficient
of permeability are the falling head test and the
constant head test. All the methods including the flow
pump system described earlier pass water through a mass
of soil, with known length of flow and cross-sectional
area. The head at the inlet and outlet can be measured or
controlled from any convenient datum. The temperature of
the fluid and the time elapsed are the last two
parameters measured. Using Darcy's law or equations
derived from Darcy's law, the coefficient of permeability
63


can be calculated. Assuming a constant, controlled flow
rate, measuring the pore liquid pressure difference
across the sample and measuring volume change, the
coefficient of permeability can be calculated as (Fetter,
1994)
k = Q
Am /
Yuq A
(5.3)
where Au is the pore pressure difference across the
sample; yLiq is the unit weight of the pore fluid
(liquid); Current height, I, and cross-sectional area, A,
are calculated by assuming that volumetric strain is
isotropic.
Intrinsic permeability, K, expresses the relative ease
with which a porous medium can transmit a liquid under a
hydraulic gradient. This is a property of the porous
medium and is independent of the nature of the liquid.
Intrinsic permeability is calculated as
K = k
(5.4)
where yliq is the unit weight and /j is the dynamic
viscosity of the liquid.
64


5.1.2 Undrained Shear Test Data Analysis
A series of equations were used to analyze the data
recorded during the triaxial testing program. The initial
length, 1Q, and diameter, d0, of the specimen were
measured as the specimen was loaded into the triaxial
cell.
The axial strain, ea, during the triaxial test can be
expressed as
-A/
where compression is positive (Wood, 1990). The axial
strain was used to correct the cross-sectional area, A,
as the test progressed. The following equation (Bowles,
1992) was used to calculate the area correction for the
undrained tests:
Aa
A=~rJL- (5.6)
The total axial stress, o-j, is the sum of the confining
or cell pressure, o3, and the axial stress due to force,
F, transmitted by the loading ram. The following equation
65


(Wood, 1990) was used to calculate the total axial stress
during the tests:
F
al= The effective stress, was then calculated using the
following equation:
cr,' = <7, u (5.8)
Where u is the pore pressure of the liquid within the
soil specimen. Similarly, the effective radial stress can
be calculated using the following equation (Wood, 1990):
cr3' = The mean effective stress, p', was then calculated using
the following equation (Wood, 1990):
P' =
a-, + 2ct3
(5.10)
The mean total stress, p, can be calculated using the
following equation (Wood, 1990):
p = p'+u (5.11)
66


5.2 Isotropic Consolidation Test Results
Isotropic consolidation test data is shown in Table 5.1.
The isotropic consolidation test was performed by
isotropically loading the specimen from mean effective
stresses ranging from 34.5 kPa (5 psi) to 689.4 kPa (100
psi). Figure 5.1 and Figure 5.2 show the plots of void
ratio, e, versus mean effective stress, p' on log and
linear scales respectively. Table 5.2 shows void ratio
data.
Table 5.1 Isotropic Consolidation Test Data
Initial Water Content 18.5 %
Initial Void Ratio (at zero total stress) 0.661
Initial Dry Density 15.89 kN/m3
B Parameter Before Test 0.93
Water Content at End of Test 22.9 %
Void Ratio at End of Test 0.616
Dry Density at End of Test 16.31 kN/m3
67


P' (kPa)
Figure 5.1 Isotropic Consolidation line (log scale)
Figure 5.2
Isotropic Consolidation line
68


Table 5.2 Isotropic Consolidation Void Ratio Data
Day Mean Effective Stress, p' Void Ratio, e
0 34 kPa (5 psi) 0.655
1 69 kPa (10 psi) 0.648
2 138 kPa (20 psi) 0.632
3 276 kPa (40 psi) 0.608
4 483 kPa (70 psi) 0.588
5 689 kPa (100 psi) 0.574
6 483 kPa (70 psi) 0.576
7 276 kPa (40 psi) 0.582
8 138 kPa (20 psi) 0.589
9 69 kPa (10 psi) 0.600
10 34 kPa (5 psi) 0.616
5.3 Flow Test Results
Two tests were conducted using the flow pump system.
Constant effective stress was maintained at the bottom of
the sample. Initial saturation and consolidation of the
specimen was achieved by setting back pressure and
confining pressure to get the desired effective stress.
Volume change of the sample was recorded and the current
height and radius were determined by assuming that the
sample was compressing isotropically, namely:
Aea = A cr
Ar Ah
r hn
o o
(5.12)
69


Initial and post consolidation data for the two flow
tests are shown in Table 5.3.
After saturation had been achieved as indicated by a B
parameter of 0.95 or greater, flow was initiated from the
top towards the bottom of the sample. As the pressure
head difference developed across the sample, pore
pressure at the top of the sample increased. This
decreased effective stress at the top of the sample, in
effect partially unloading it. Figure 5.3 graphically
shows the expected change in effective stress due to
flow. It is assumed that axial stress is equal to radial
stress during both consolidation and flow due to self
weight stress being negligible when compared to confining
stress. Therefore, from now on the term effective stress
refers to both axial and radial effective stresses.
Furthermore the recorded changes in the effective stress
state are shown in Figures 5.4 and 5.10 for tests M_52
and M_186 respectively.
The behavior of both specimens during the flow test was
similar. When the methanol was pumped through, both
specimens began to swell. The swelling continued even
after flow was stopped. As expected, a larger amount of
swelling was observed in the specimen at the lower
70


confining pressure. It is important to note that the
swelling occurred while the effective stress at the
bottom of the sample was constant and the effective
stress at the top was increasing (see Figures 5.4 and
5.10). In mechanical terms, the increase in effective
stress would indicate loading. With the sample swelling
while it was in effect being mechanically loaded, it can
be concluded that the observed mechanical response
(swelling) was chemically induced as the pore liquid was
changed to methanol. Figures 5.5 and 5.11 show the change
in void ratio versus time for each of the flow tests.
Figure 5.14 and 5.15 shows the void ratio versus the
average effective stress (on log and linear scales) for
both flow tests. The figures also shows the isotropic
consolidation line as a reference. The data presented in
Figure 5.14 and Figure 5.15 clearly indicate chemically
induced swelling. The final location of the void ratio-
mean effective stress points also indicate that some sort
of material transformation is taking place, perhaps
transformation from deflocculated towards a more
flocculated structure. Figures 5.6 and 5.12 show the
change in hydraulic gradient versus time for each of the
flow tests.
71


Table 5.3
Initial and Post Consolidation Data For
Flow Tests
Test No. M_52 M_18 6
Initial Area (cm2) 21.03 20.94
Initial Volume (cm3) 213.89 215.70
Initial Void Ratio 0.652 0.662
Initial Moisture Content (%) 17.8 17.9
Volume After Consolidation ( cm3) 212.74 208.91
Area After Consolidation ( cm2) 20.95 20.49
Void Ratio After Consolidation 0.643 0.610
Skempton's B Parameter 0.95 0.96
Ub o3
Figure 5.3 Pore Pressure Changes Due to Flow
72


Hydraulic gradient, i, and hydraulic conductivity, k,
were also determined and plotted continuously throughout
the flow tests (see Figures 5.6, 5.7, 5.12 and 5.13).
After one pore volume of water had been pumped through
the M_52 specimen, it appeared that hydraulic gradient
and effective stress at the top of the sample were at or
approaching steady state. The data does not indicate that
a steady state condition was being approached in the
M_186 test. In both tests, hydraulic conductivity began
to increase as soon as methanol introduction was started.
The rate of increase was higher at lower confining
pressure. After two pore volumes of methanol it appeared
that hydraulic gradient was again at or approaching
steady state. The intrinsic permeability, K, and
coefficients of hydraulic conductivity at the end of one
pore volume of water and two pore volumes of methanol can
be found in Table 5.4 and Table 5.5 respectively.
Methanol caused the intrinsic permeability to increase
and increased confining pressure tended to minimize the
amount of the increase. These were based on the
volumetric flow rate experienced by the specimen. Since
sample volume was changing at times, the volumetric flow
rate was the sum of the volumetric flow rate of the pump
and the volumetric flow rate into or out of the sample
73


necessary to account for the volume change during that
time increment.
Figures 5.7 and 5.13 show the change in the coefficient
of permeability versus time for each of the flow tests.
Table 5.4 Intrinsic Permeability for Water and Methanol
Test Number Water Methanol
M_52 3.77xl016 m2 9.47xl0"16 m2
M_18 6 l.llxlO'15 m2 l.45xl0"15 m2
Table 5.5 Coefficients of Permeability for Water and
Methanol(1)
Test Number Water Methanol
M_52 3.70xl0"7 cm/sec 1.25xl0-6 cm/sec
M_186 1.09xl0-6 cm/sec <2) 1.91xl06 cm/sec
Note: (1) The correct values of hydraulic conductivity can only
be reported for steady state conditions.
(2) This value was not at or approaching a steady state
condition.
Separate plots for water and methanol are included due to
the different densities of the two fluids. Both water and
methanol plots are shown for the second pore volume
because, at that time, both water and methanol are
present in the specimen. The spikes in the data for the
74


Effective Stress at Sample Ends
Top
a Bottom
Figure 5.4 Changes in Effective Stress State for Flow
Test M 52
Void Ratio vs. Time
Figure 5.5 Void Ratio Versus Time for Flow Test M_52
75


Hydraulic Gradient
70.00
60.00
50.00
40.00
30.00
20.00
10.00
0.00
0 50 100 150 200 250
Time (hours)
Figure 5.6 Hydraulic Gradient Versus Time for Flow Test
M 52
Hydraulic Conductivity
Methanol
Figure 5.7
Test M 52
Hydraulic Conductivity Versus Time for Flow
76


Figure 5.8 Volume Versus Time for Flow Test M_52
Figure 5.9 Volume Versus Time for Flow Test M_186
77


Effective Stress at Sample Ends
195.00
190.00
cl 185.00
165.00
Bottom
j Top j
300
Figure 5.10 Changes in Effective Stress State for Flow
Test M 186
Void Ratio vs. Time
Figure 5.11 Void Ratio Versus Time for Flow Test M_186
78


Hydraulic Gradient
VWter I
Methanol
Figure 5.12 Hydraulic Gradient Versus Time for Flow Test
M 186
Hydraulic Conductivity
-Water
-Methanol
Figure 5.13 Hydraulic Conductivity Versus Time for Flow
Test M 186
79


P'(kPa)
isotropic Consolidation Line
eVoid Ratio Changes for M_52
Void Ratio Changes for M_186
1 e
2 e
3 e
4 e
5 e
consolidation
1 pore votum* wale'
1 pore volgma mathanol
2 pore volgmas matnanol
frna*
Figure 5.14 Void Ratio
M_186 (log
Changes for Flow Tests M_52 and
Scale)
Figure
5.15
Void Ratio Changes for Flow Tests M_52 and
M 186
80


most part can be attributable to the daily isolation of
the sample during refilling of the pump. During this
short period of time, the DPTs were isolated from the
sample.
As shown in Table 5.4 and Table 5.5, both the coefficient
of permeability and intrinsic permeability increased when
the pore liquid was changed from water to methanol.
Equation 5.3 and 5.4 show that a change in the unit
weight, y, of the liquid will cause a change in the
calculated values of k and K. Additionally, a change in
the dynamic viscosity, /y, also has an effect on the
calculated value of intrinsic permeability. A comparison
of unit weights divided by the dynamic viscosities for
water and methanol shows a factor of increase of 1.35.
Since the factors of increase in the values of k and K
(see Table 5.6) are larger than can be attributed to the
difference in the properties of the two pore liquids, it
can be concluded that a significant portion of the change
can be attributed to changes in the geometry of the soil.
5.4 Undrained Shear Test Results
Five isotropically consolidated, undrained compression
tests were conducted. Initial consolidation and
81


Table 5.6 Coefficient of Permeability and Intrinsic
Permeability Increase factors
fr methanol methanol
Test Number k n water ^ water
M_52 3.38 2.51
M 186 1.75 1.31

saturation of the specimen was achieved by setting back
pressure and confining pressure to get the desired
effective stress. The volume change of each sample was
recorded and the current height and radius were
determined by assuming that the sample was compressing
isotropically using the method discussed in Section 5.3.
Table 5.7 provides data about the CIUC tests.
There are three item meriting special note. First, The
liquid content at the end of the tests of the two
methanol samples is lower than for the water samples.
Moisture content is a function of the mass of the liquid.
The difference is due to methanol having a density of
0.79 times that of water. Equivalent water content was
determined by dividing the liquid content by the specific
gravity of the liquid. Second, the test information for
methanol saturated specimens reflects the swelling that
82


occurred during the flow tests. Third, The B parameter
for the W_187 was only 0.84. The sample stood for 2
weeks but the B parameter did not improve. The back
pressure was raised an additional 138 kPa (20 psi) while
Table 5.7 Undrained Shear Test Data
Test Data W_22 W_44 W_187 M_52 M_186
Initial Water Content(%) 18.1 18.0 17.7 17.8 17.9
Initial Void Ratio 0.707 0.661 0.660 0.652 0.662
Initial Dry Density (kN/m^) 15.45 15.88 15.89 15.97 15.87
Initial Height (cm) 10.16 10.33 10.20 10.17 10.30
2 Initial Area (cm ) 21.75 20.84 21.09 21.03 20.94
Post Consolidation Void Ratio 0.703 0.632 0.610 0.643 0.610
Post Consolidation Dry Density (kN/m ) 15.50 16.17 16.39 16.06 16.39
Test Dry Density (kN/m^)^ 15.50 16.17 16.39 15.09 15.64
Test Void Ratio ^ 0.703 0.632 0.610 0.748 0.687
Test Initial Height (cm)^ 10.15 10.27 10.11 10.22 10.22
Test Initial Area (cm^)^ 21.72 20.59 20.59 22.14 21.42
Liquid Content After Test(%)<1) 25.0 22.6 21.8 18.5 17.6
Equivalent Water Content After Test ('/.) ^ 25.0 22.6 21.8 23.4 22.3
(1) B Parameter 0.95 0.95 0.84 0.98 0.95
Shear Rate (mm/min) .0315 .0315 .0315 .0315 .0315
Note:
(1) Values are post consolidation for water saturates specimens and
at isolation for undrained shear testing for the methanol
saturated specimens.
83


maintaining a constant effective stress. This also did
not cause the B parameter to change.
The results of each the triaxial compression tests are
presented in a pair of plots. The first plot is of the
deviator stress, q, which is defined as
q = o-i o3 (5.13)
versus the percent of axial strain, ea. The second plot
is excess pore liquid pressure also versus the percent of
axial strain, sa. As noted before, the pore liquid was
water for tests W_22, W_44, and W_187. The pore liquid
for tests M_52 and M_186 was methanol. There is one set
of plots for each of the five triaxial compression tests
of the testing program. These five sets of plots are
shown in Figures 5.16 through 5.25.
There are two details worth noting in the following
plots. The first is the randomly occurring unloading
most visible in tests W_22, W_187, and M_186. This was an
anomaly of the loading frame used for these tests. The
cause was not determined. The second is the small
oscillations in pressure in test W_44 and M_186. These
were the last two triaxial compression tests conducted.
84


Figure 5.16 Deviator Stress Versus Axial Strain for Test
W 22CIU
Axial Strain (%)
Figure 5.17
Excess Pore Liquid Pressure Versus Axial
Strain for Test W 22CIU
85