Thermal treatment of expansive soil

Material Information

Thermal treatment of expansive soil
Shepard, Mark D
Publication Date:
Physical Description:
141 leaves : illustrations ; 28 cm


Subjects / Keywords:
Swelling soils -- Analysis ( lcsh )
Swelling soils -- Measurement ( lcsh )
Soil stabilization ( lcsh )
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )


Includes bibliographical references (leaves 146-138).
General Note:
Department of Civil Engineering
Statement of Responsibility:
by Mark D. Shephard.

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:
40337049 ( OCLC )
LD1190.E53 1998m .S54 ( lcc )

Full Text
Mark D. Shepard
B.S., University of Idaho, 1976
M.S., University of Texas at El Paso, 1984
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

This thesis for the Master of Science
degree by-
Mark D. Shepard
has been approved
//, /ff DATE

Shepard, Mark D. (M.S., Civil Engineering)
Thermal Treatment of Expansive Soil
Thesis directed by Professor N. Y. Chang
Despite extensive research on their behaviour, expansive
soils continue to cause enormous damage to engineered
structures. Naturally occurring expansive soils contain
varying amounts of highly swelling clay minerals, such as
montmorillonite. These soils will heave and contract under
changing moisture conditions and exert pressure on adjacent
structures including foundations, walls, pavements,
pipelines, and canals. Contemporary remediation and
structural design practices are not always effective in
countering the problems; thus, continued research attention
is needed.
Objectives of this research are three fold: (1) select an
expansive soil and determine its swelling characteristics;
(2) determine the effectiveness of thermal treatment to
reduce its swelling pressure and percent swelling
potential; and (3) formulate functional relationships where

swelling pressure and percent swelling are expressed in
terms of temperature, treatment duration, and index
properties of soils.
A laboratory testing program was carried out to determine
the swelling characteristics of an expansive soil and its
derivatives from thermal treatment. A total of 10 samples
were heated from 150C to 800C (302F to 1,472F) for
durations ranging from 24 to 240 hours. Liquid limit,
plastic limit, Proctor compaction and specific gravity
tests were performed on the heated soils. Swell tests were
conducted on all samples for their percent swelling
potential and swelling pressure at an overburden pressure
of 23.94 kPa (500 psf).
Results show that the thermal treatment technique is
effective in reducing the percent swelling and swelling
pressure at 400C (752F) and eliminating these properties
at 600C (1,112F). Dehydroxylation, the irreversible
release of structurally held water in clay minerals,
occurred between 600 and 800C and was accompanied by a
marked decrease in specific gravity of soil particles.
Between 400 and 600C the soil was converted to a
nonplastic granular material and the color was transformed
from its original olive-gray to bright reddish brown.

Statistical analyses using STATGRAPHICS revealed strong
linear correlation between heating temperature, swelling
characteristics, and soil index properties. The percent
swelling potential, swelling pressure, liquid limit, and
plastic limit all decreased with increasing temperature.
Correlation with heating duration were inconclusive due to
the nature of the sample population.
This abstract accurately represents the content of the
candidate's thesis. I recommend its publication.

I have always held a love and general curiosity of the
engineering discipline and the attainment of this Masters
Degree in Civil Engineering has been a long-time personal
goal. I now have a deep appreciation and respect for those
many individuals who are striving to advance their
education while holding down a full-time position, and for
their spouses and family who support them.
I extend my sincere appreciation to Dr. N.Y. Chang, who
supervised this study, for providing a constant source of
encouragement, enthusiasm, and inspiration. His patience
and wisdom bestowed over the course of this work was also
greatly appreciated. I would also like to thank Dr.
Jonathan T-H. Wu and Dr. Dunja Peric for sharing their
knowledge and experience in the classroom and for reviewing
this research.
I am grateful to Mr. Tim Spencer of GTG Geotech. Inc. for
sharing key information on soil conditions at the Fairway
Vista Subdivision and for his guidance on swell-pressure
testing parameters used at his company.

Special thanks to Mr. Bill Kepler, Mr. Kurt Mitchell, and
Mr. David Harris of the Materials, Engineering, and Earth
Science Laboratory, U.S. Bureau of Reclamation, for
providing access to their muffle furnace.
Greg Balint and Jan Chang provided good company and
assistance throughout those long hours spent in the UCD
soils laboratory. Greg in particular was instrumental in
developing the static compaction technique used to prepare
test specimens in this study.
Finally, I would like to dedicate this work to my wife,
Anita, in return for her love, affection, and endorsement
of this endeavor. She is as good a friend and lab partner
as anyone could ask for, and my sense of achievement was
all the more fulfilled in her presence.

1. INTRODUCTION ................................ 1
1.1 Problem Description........................1
1.2 Purpose....................................5
1.3 Scope of Study.............................7
2. LITERATURE REVIEW ........................... 9
2.1 Introduction...............................9
2.2 Characteristics of Expansive Soils........11
2.2.1 Clay Mineralogy.........................13 Octahedral and Tetrahedral Layers .. 14 Isomorphous Substitution ...........19 Clay Particle Attraction ...........21 Relation of Structure to
Properties ..........................22
2.2.2 Initial Moisture Content .............. 24
2.2.3 Initial Density.........................25
2.2.4 Surcharge Pressure .................... 28
2.2.5 Soil Structure and Fabric...............31
2.2.6 Availability of Water...................32


2.2.7 Type and Concentration of
Electrolytes ......................... 34
2.2.8 Thickness of the Soil Stratum...........36
2.2.9 Time Available for Swelling to Occur .. 38
2.2.10 Swelling Fatigue ...................... 39
2.3 Thermal Treatment Effects..............40
2.3.1 Strength................................40
2.3.2 Pulverization and Stabilization ....... 45
2.3.3 Fusion of Soils.........................47
3. TESTING PROGRAM.............................49
3.1 General...................................49
3.2 Test Approach.............................50
3.3 Sampling Location and Geologic
3.4 Soil Properties...........................56
3.5 Sample Preparation and Testing
Procedures............................... 61
3.5.1 General................................ 61
3.5.2 Thermal Pretreatment .................. 62
3.5.3 One-Dimensional Swell Test.............65 Specimen Preparation ................ 65 Oedometer Cell .................69 Procedure 76

3.5.4 Equipment Calibration.................. 81
4. TEST RESULTS................................ 83
4.1 General...................................83
4.2 Experimental Results......................84
4.2.1 Physical Changes ...................... 89
4.2.2 Swell-Pressure Tests .................. 94
4.2.3 Atterberg Limits ...................... 96
4.2.4 Moisture-Density Relationships ........ 98
4.3 Summary and Discussion...............100
5.1 Introduction.............................103
5.2 Description of Statistical Procedure.... 104
5.3 Results of Statistical Analyses......108
5.3.1 Total Sample Population .............. 108
5.3.2 Low-Temperature Subgroup.............. 116
5.3.3 Single-Temperature Subgroup........... 120
5.4 Summary and Discussion...............124

FOR FUTURE RESEARCH......................128
6.1 Summary..................................128
6.2 Conclusions..............................131
6.3 Recommendations for Future Research.....133
REFERENCES......................................13 6
Values of Swelling Properties and
Correlation Coefficients in the
Low-Temperature Subgroup Corrected
for Initial Moisture and Density
Conditions................................ 139

2-1 Diagram of the Octahedral, or Gibbsite
Sheet of Phyllosilicate Structures.......15
2-2 Diagram of the Tetrahedral, or Silica
Sheet of Phyllosilicate Structures.......16
2-3 Diagram of the Structure of Kaolinite.... 17
2-4 Diagram of the Structure of
2-5 Effect of Varying Moisture Content
on Volume Changes from Constant
Density Samples....................................25
2-6 Effect of Varying Density on Volume Change
for Constant Moisture Content Samples.... 27
2-7 Effect of Varying Density on Swelling
Pressure for Constant Moisture Content
2-8 Effect of Varying Surcharge Pressure for
Constant Density and Moisture Content
2-9 Effect of Varying Sample Thickness on
Volume Changes for Constant Density and
Moisture Content Samples..................37
2-10 Effect of Heat Treatment on Dry Strength
of Kaolinite...............................42
2-11 Effect of Heat Treatment on the Rewet
Strength of Kaolinite......................42

3-1 Location Map of Sample Collection Site
for Expansive Soil Used in This Study.... 54
3-2 Moisture-Density Curve for Untreated
Soil Used in the Thermal Treatment
3-3 Typical Gradation Curve for Untreated
Soil Used in the Thermal Treatment
3-4 Relation of Test Soil to the Positions
of Other Clay Minerals on Casagrande's
Plasticity Chart...................................60
3-5 Photograph of Stabil-Therm Oven Used
for Pretreatment of Soil Samples to
250C.................................... 63
3-6 Photograph of Thermo Electric Mfg. Co.
Muffle Furnace Used to Pretreat Soil
Samples above 2 50C..................... 66
3-7 Photograph of Load Cell Used to Prepare
Pretreated Soil Specimens for Swell
Testing.................................. 68
3-8 Disassembled Standard Fixed Ring
Oedometer Cell............................70
3-9 Stainless-Steel Confining Ring Placed
on Bottom Porous Stone Inside the
Oedometer Cell............................71
3-10 Photograph of Locking Ring Installed
Over the Steel Confining Ring and
Secured with Thumbscrews...........................72
3-11 Top View of the Oedometer Cell, Fully
Assembled with Top Porous Stone and Cap
in Position on Top of the Specimen........73

3-12 Close-up View of the Consolidometer
3-13 Photograph of Standard Oedometer Cell
Placed on Consolidometer Platform,
Ready for Testing.........................75
3-14 Measurement of Swell Potential and
Swelling Pressure Using ASTM D 4546,
Method B..................................77
3- 15 Calibration of the Consolidometer and
Standard Oedometer Cell...................82
4- 1 Histogram of Initial Moisture Contents
of Swell-Pressure Test Specimens..........88
4-2 Histogram of Initial Dry Densities of
Swell-Pressure Test Specimens.............89
4-3 Color Gradation in Thermally Pretreated
Samples of Expansive Soil.................91
4-4 Photograph of "Bricked" Expansive Soil... 92
4-5 Change in Specific Gravity of Soil
Solids as a Result of Thermal
4- 6 Maximum Dry Density vs. Optimum
Moisture Content of Thermally
Pretreated Expansive Soil Samples........100
5- 1 Plot of the Fitted Linear Model for
Heating Temperature vs. Percent
Swelling Potential.......................110
5-2 Plot of the Fitted Linear Model for
Heating Temperature vs. Swelling

5-3 Plot of the Fitted Linear Model for
Heating Temperature vs. Plasticity
5-4 Plot of the Fitted Linear Model for
Heating Temperature vs. Liquid Limit.... Ill
5-5 Plot of the Fitted Model for Liquid
Limit vs. Percent Swelling Potential.... 113
5-6 Plot of the Fitted Model for Liquid
Limit vs. Swelling Pressure.............114
5-7 Plot of the Fitted Model for Plastic
Limit vs. Percent Swelling Potential.... 114
5-8 Plot of the Fitted Model for Plastic
Limit vs. Swelling Pressure..............115
5-9 Plot of the Fitted Linear Model for
Initial Moisture Content vs. Percent
Swelling Potential.......................119
5-10 Plot of the Fitted Linear Model for
Initial Dry Density vs. Swelling
5-11 Plot of the Fitted Linear Model for
Heating Time vs. Liquid Limit............123
5-12 Plot of the Fitted Linear Model for
Heating Time vs. Plasticity Index........123

2.1 Results of Heat Treatment on Zettlitz
Kaolin IA.................................44
3.1 Thermal Pretreatment Conditions for
Test Samples..............................51
3.2 Post-Treatment Soil Properties
Measured in the Laboratory Test
3.3 Results of Index Property Tests on
Soil Used for Thermal Treatment
4.1 Results of Thermal Pretreatment
Testing on Expansive Soil.................85
4.2 Results of Compaction Tests on
Thermally Treated Expansive Soil..........99
5.1 Test Variable Correlation Matrix;
Total Sample Population..................109
5.2 Test Variable Correlation Matrix;
Low-Temperature Subgroup.................117
5.3 Test Variable Correlation Matrix;
Single-Temperature Subgroup..............121

1.1 Problem Description
Expansive soils are widely recognized as a worldwide
problem with the potential to cause extensive damage to
civil engineering structures. These soils have the ability
to exhibit unusually large volume changes as a result of
moisture variations often associated with changing
environmental conditions. Expansive soils are typically
characterized by (1) a high percentage of clay particles
(effective diameter of 0.002 mm or less) and (2) the
presence of clay minerals having a particular chemical
composition that will expand when coming into contact with
water. Since the driving mechanism for these soils is
changes in moisture content, it should be recognized that
the potential for shrinking and settlement due to
contraction exists as well.
The swell and shrinkage of expansive soil can cause
extensive damage to structures such as foundations and
walls of buildings, retaining walls, canal and reservoir
linings, pavements, and roadways. Types of damage may

include distortion and cracking of pavements and on-grade
floor slabs; cracks in grade beams, walls, and drilled
piers; and jammed or misaligned doors and windows. The
damage is caused by differential movement of the foundation
soil. Differential movements redistribute the structural
loads and cause large changes in moments and shears not
accounted for in design. These in time cause changes in
stress which result in cracking and/or undesireable
relative deformations in the structure.
Research on expansive soils has been ongoing for several
decades and much has been learned about their nature and
destructive tendancies. Numerous soil modification and
construction techniques have been tried in an attempt to
limit the amount of swelling that occurs, although many of
these have met with limited long-term success. Even so,
structural damage continues to occur to civil engineering
structures founded on and in these soils. The damage
typically results in some form of legal action taken
against the builder with expensive repairs and mitigative
measures to follow.
In the United States (U.S.) alone, annual damages
attributed to expansive soil movement have been estimated
to exceed 2 billion dollars, over half of which occurs to

transportation facilities such as highways and streets
(Jones and Holtz, 1973). A study conducted for the
National Science Foundation in 1978 projected that
residential losses in the U.S. due to expansive soils could
approach 1 billion dollars annually by the year 2000 (J. H.
Wiggins Co., 1978). This figure arguably could be
increased by a factor of 2 or 3 when the cost of damage to
commercial/industrial buildings and transportation
facilities is considered. As pointed out by Nelson and
Miller (1992), these cost figures are somewhat subjective.
The documented evidence of expansive soil damage is limited
due to the fact that these losses are not generally
considered for federal/state agency disaster funds or
covered by insurance.
Certain geographical areas are more prone to damage from
expansive soils than other areas, especially those areas
that have large surficial clay deposits and climates
characterized by periods of rainfall and drought. In the
U.S., the major concern with expansive soils exists
generally in the western and Gulf States, particularly
Wyoming, Colorado, California, and Texas. The expansive
soil problems in the northern and central U.S. are
primarily related to highly overconsolidated shales (Chen,
1988). In the Denver, Colorado Metropolitan Area, a region

of heaving bedrock exists within 1-3 miles of the Front
Range piedmont. The region has been called the Designated
Dipping Bedrock Area (DDBA) in Jefferson County and the
Dipping Bedrock Overlay District (DBOD) to the south in
Douglas County. The combination of steeply dipping
sedimentary bedrock and presence of expansive claystones
produces a particularly destructive linear style of ground
deformation in these districts (Noe and Dodson, 1995) .
A variety of soil treatment methods and/or construction
techniques have been used to alleviate the detrimental
volume change of expansive soils. These include the
(1) sub-excavation and removal of expansive soil and
replacement with nonswelling soil
(2) application of heavy applied load to counteract the
swelling pressure
(3) preventing moisture changes by encapsulating the
swelling soil
(4) flooding the in-place soil to achieve swelling prior
to construction
(5) decreasing the soil density by compaction control

(6) changing the properties of the expansive soils using
chemical admixtures
(7) explosive treatment to disrupt bedded shale
orientation and decrease its density
(8) avoid the expansive soil altogether
According to Ardani (1992) many of these techniques have
been tried on various Colorado Department of Transportation
(CDOT) projects. In certain CDOT districts, some of these
methods have met with poor success and costly repair or
overlay has been required.
1.2 Purpose
This study was conceived in an effort to explore additional
soil remediation techniques that could potentially be used
to solve the expansive soil problem. The purpose of this
research is essentially twofold: (1) to investigate the
capability of thermal pretreatment to reduce, and
potentially eliminate, the swelling properties of an
expansive soil; and (2) to establish empirical
relationships between treatment parameters, namely the
heating temperature and heating duration, and widely used
indicators of expansive soil behaviour, namely the soil's
swelling potential, swelling pressure, and index properties

such as the liquid limit, plastic limit, and plasticity-
index .
To the author's knowledge, the deliberate application of
heat energy to eliminate the swelling properties of a soil
has never been studied before. Several researchers have
reported on the fact that cyclic wetting and drying of an
expansive soil leads to diminished swell potential. This
has been demonstrated through direct field observations as
well as laboratory tests where soil specimens were allowed
to dry passively at room temperature prior to testing and
rewetting. Through review of scanning electron
micrographs, several authors have suggested that the
wetting/drying cycle leads to a continuous rearrangement of
clay particles, which in turn, inhibits the soil's ability
to swell. The author believes that thermal treatment of
expansive soil, at sufficiently high temperatures, goes
well beyond merely drying the soil and promotes fundamental
changes to the clay mineral structure. These changes are
irreversible and prohibit the mineral's ability to readsorb
moisture. Thus a thermal treatment technique offers the
potential for an effective and permanent solution to the
expansive soil problem.

1.3 Scope of Study
The scope of this research consists of the following
principal tasks:
1. Performing a review of the published literature to
document the physical and chemical characteristics of
expansive soils, and to determine the extent to which
previous researchers have studied the the effects of
heat treatment on soil engineering properties.
2. Selecting samples of a naturally occurring expansive
soil and determining its swelling characteristics and
index properties.
3 Thermally pretreating the expansive soil samples over a
range of temperatures and for various durations. Then
conducting swell-pressure, Atterberg limit, and specific
gravity tests on the pretreated soil specimens.
4. Performing a statistical comparison of test results to
determine the relative strength of correlations between
test variables. This comparison was then used to
display the most significant relationships between
thermal treatment parameters and induced soil

The results of each task are discussed in the chapters that
follow. The overall plan for this study was to prove, in
principal, the capability of thermal treatment and to
determine the most effective treatment conditions. If the
method is to be developed further, the results obtained in
these laboratory studies need to be transferred to the
field. Field trials using a variety of application
techniques (e.g., in-situ or ex-situ) would provide
valuable data on the suitability and economics of this soil
remediation method. Ultimately it is hoped that heat
treatment could provide a viable and cost-competitive
alternative to current foundation design and soil
modification techniques used in areas of expansive soils.

2.1 Introduction
The problems associated with expansive clay soils have long
been recognized and there is a considerable body of
literature published on the subject. Over the past 35
years, the results of laboratory testing and field
observations have yielded an extensive amount of knowledge
on the nature and behaviour of expansive soils. Excellent
descriptions of the engineering properties of expansive
soils and foundation design practices are provided in Chen
(1988); Gromko (1974); Nelson and Miller (1992); Parcher
and Liu (1965); Seed, Woodward, and Lundgren (1962); and
Snethen, Johnson, and Patrick (1977) among others.
By comparison, the effects of temperature on the
engineering properties of expansive soils is not as well
studied and there is much smaller pool of literature
available on this topic. Several authors have investigated
the thermal treatment of pure clays and naturally occuring
clay soils from the aspect of improving strength,
compression, and pulverization properties. However,

suprisingly little work has been done on thermal treatment
as a means to alter the swelling potential of clay soils.
One notable contribution to the technical literature is a
collection of papers from an international conference on
the "Effects of Temperature and Heat on the Engineering
Behaviour of Soils" held in Washington, D.C. in January,
1969. These papers were published as Special Report 103 by
the Highway Research Board of the National Academy of
This literature search focused on prior investigations
where soils were thermally pretreated; that is, heat energy
was deliberately applied to a soil in an attempt to raise
the soil temperature up to and beyond its fusion
temperature. The temperature range of interest for this
study was from greater than 100C (212F) to less than
1,000C (1,832F). The fundamental objective of the
pretreatment is to gain beneficial engineering effects from
reactions that occur in the soil as a direct result of the
heat energy applied. Other investigations were found where
soil properties were examined within a temperature range
that might be experienced under extreme environmental
conditions, say from less than 0C to over 60C. While
these studies may have a great deal of practical
application, for instance, in highway subgrades, they were

not considered germaine to the scope of this project and
were excluded from further consideration.
A complete review of the innumerable studies on expansive
soils is beyond the scope of this paper; however, based on
a review of above-referenced work, the most important
characteristics of expansive soils, the factors affecting
their expansion potential, and their behaviour when
subjected to thermal pretreatment are addressed in this
2.2 Characteristics of Expansive Soils
Expansive soils are naturally occurring materials broadly
characterized by a high percentage of clay particles (less
than 2 microns in diameter) and the ability to undergo
volumetric expansion when exposed to water. Volumetric
expansion is the result of changes to the soil-water
system; without these environmental changes, swelling will
not take place. The mechanism of swelling in expansive
soils is complex and is influenced by a variety of factors.
Seed, Woodward and Lundgren (1962) discuss the difference
between predictions of "swell" and "swelling potential" in
soils, pointing out that the amount of expansion of a soil
due to swell is influenced by a number of factors. The

main factors that determine the soil's potential to swell
are related to the nature of the soil particles,
specifically, the type and amount of clay present. As
discussed below, the remarkable ability of some clays to
attract dipolar water molecules and various cations to
their surfaces is the principal cause of their expansive
A second set of factors, relating to conditions of soil
placement and the field environment, are extremely
important in determining the extent to which the soil will
realize its swelling potential. These include: the initial
dry density (either in-situ or as compacted); initial water
content; soil structure and fabric; surcharge and/or
confining pressure; the availability of water; the type of
electrolyte in the water and its concentration; thickness
of the soil stratum; time permitted for swelling to occur;
and temperature. Thus it appears that while two soils may
have the same potential to swell, the actual amount of
swelling realized may be quite different, owing to a)
either placed at or possessing different in-situ physical
conditions, b) placed at different chemical compositions,
or c) subjected to different climatic conditions. These
factors are discussed further in the following sections.

2.2.1 Clay Mineralogy
Parcher and Liu (1965) note that the mineralogical
composition of a clay soil is of primary importance in
determining its potential for swelling. Other intrinsic
properties and the ambient environmental conditions tend to
control the magnitude and rate of swelling. The properties
of clays in general depend on their small grain size. In
most clay samples, the particles have dimensions entirely
or partially within the colloidal range (usually defined as
less than 1 micron (10'3 mm) in diameter). This means that
they remain suspended in water for long periods, may be
flocculated by electrolytes, and that they absorb ions out
of solution. These absorbed ions may be replaced by others
when the concentration of the solution changes. Also, by
virtue of their small size, clay particles have the ability
to absorb water and organic materials and so to become
plastic; i.e., remain coherent and capable of being
remolded when wet. These properties, while present to some
degree in all clays, differ markedly from one clay mineral
to another.
Chemically, the clay minerals are described as hydrous
aluminum silicates, although many clays contain other
metals as well, particularly magnesium and iron. Molecular

formulas of the principal clays may be written as H4Al2Si209
(kaolin) and HAlSi206 (montmorillonite). Actual
compositions show variations in the Si/Al ratio, a variable
quantity of water, and usually considerable amounts of
magnesium, iron, calcium, and the alkali metals. There
seems to be little correspondence between chemical
composition and properties of the clays. Two clays with
similar proportions of the various elements may show great
differences in plasticity and capacity for ion exchange,
while another pair with very different compositions may
show strikingly similar properties (Krauskopf, 1967). Octahedral and Tetrahedral Layers
Clay minerals (with a few rare exceptions) are
phyllosilicates with continuous sheet structures. Clays
have a characteristic structure composed of alternating
layers or sheets of two kinds. Successive unit layers of
the clay structure are stacked one on top of another and
the differences in geometry give rise to clays with
different properties.
One sheet consists of the ions Al3*, 0=, and OH'; the
negative ions form octahedra around Al3* and their relative
numbers are adjusted to satisfy the valence of the entire
structure. The 0=, and OH' are shared between adjacent

octahedra so that the pattern is continuous in two
dimensions (Fig. 2-1). This pattern is referred to as the
"octrahedral" or "gibbsite" sheet of the clay structure.
(^) = Hydroxyls
O = Aluminums, magnesiums, etc.
Figure 2-1. Diagram of the octrahedral, or gibbsite
sheet of phyllosilicate structures. (The
diagram at upper left shows a single
octahedral sheet.) When Al3+ occupies the
centers of the octahedra, only two-thirds
of the possible sites are filled; when Mg++
occupies these positions, all sites are
filled (from Grim, 1953).
The second kind of sheet is composed of Si4+, 0=, and OH
ions; each Si4+ is in the center of a tetrahedron of oxygen

ions. The tetrahedra all face the same direction and the
oxygens at their base are linked to form hexagonal rings
(Fig. 2-2). This sheet is known as the "silica sheet" or
"tetrahedral sheet" of the clay structure.
(^) and ) = Oxygens
# = Silicons
Figure 2-2. Diagram of the tetrahedral sheet, or silica
sheet of phyllosilicate structures (single
tetrahedral unit at upper left) (after
Grim, 1953).
The complete clay structure consists of one of several
combinations of the octrahedral and tetrahedral sheets.
The simplest combination, the layer structure of kaolinite,
is shown in Fig. 2-3. A single octahedral sheet is linked
to a single tetrahedral sheet by sharing some of the oxygen

ions. The dual layer extends indefinitely in two
dimensions, and the clay crystal is built up of succession
of these layers one on top of another.
Figure 2-3. Diagram of the structure of kaolinite (from
Grim, 1953) .
Another way in which the octahedral and tetrahedral sheets
are combined is shown in Fig. 2-4. Here, in the mineral
montmorilIonite, an octahedral sheet is sandwiched between
two tetrahedral sheets forming a symmetrical composite
layer. Successive layers are more easily separated in the
montmorillonite structure than in the kaolinite structure.
The difference is that 0= ions in one layer always face 0=
ions of another, while adjacent layers in kaolinite have O'

ions in one facing OH" in the next. The attraction of 0=
for OH' is much greater than that of Cr for other 0= ions
(the H+ from the OH' is shared between the Cf of adjacent
layers of kaolinite forming a hydrogen bond and linking the
layers together).
A kaolinite-type clay is often said to have a two-layer
structure (one gibbsite sheet and one silica sheet),
whereas a montmorilIonite pattern is said to have a three-
layer structure. Other combinations of the unit sheets are
possible but these two are the most common.
Montmorillonite is capable of holding potassium ions, K*,
whose relatively large size enables it to fit snugly
between layers of the clay structure. Clays which have a
deficiency of positive charge, for instance due to
substitution in the octahedral layers, tend to hold K*
especially tightly. These clays, with successive layers
held together by K" ions, have properties different than
montmorillonite and are given the name illite.
Another complexity of the clay structure is shown by clay
minerals whose crystals consist of alternating layers of
different kinds. These are termed mixed-layer clays. The
interlayering may follow a regular pattern so many layers

of one kind followed by so many of the other or it may be
completely random.
Exchangeable cations
O Oxygens @ Hydroxyls % Aluminum, iron, magnesium
O and Silicon, occasionally aluminum
Figure 2-4. Diagram of the structure of montmorillonite
(from Grim, 1953) . Isomorphous Substitution
Isomorphous substitution, or the substitution of one ion
for another in the crystal structure, is extensive in

montmorillonite clays, but not in kaolinite. This is
another important difference in the two kinds of structure.
The common substitution in the tetrahedral sheets of
montmorillonite is Al3* for Si4+, the amount of substitution
being limited to about 15%. In the octahedral sheet, a
much greater variety of substitution is possible, and the
amount may range from very small to 100%. The most common
ion substitutions are Ca++, Mg++, and Fe3*, and rarer ones
are Zn*+, Ni++, Li*, and Cr3+.
The substitution of Al3+ for Si4+ and of Mg++ for Al3+ leaves
a deficiency of positive charge in the montmorillonite
layers. The deficiency may be compensated in a variety of
ways: by replacement of 0= by OH'; by introduction of excess
cations into the octahedral layer; and by adsorption of
water molecules and other cations onto the surface of
individual layers. Some of the compensation is always
accomplished by the last method, so that montmorillonite is
characterized by the constant presence of adsorbed ions.
These ions are held loosely to the clay structure and are
readily replaced by others, explaining why the
montmorillonite clays typically have a high capacity for
ion exchange.
20 Clay Particle Attraction
Clay minerals occur as tiny platelets having two types of
exposed surfaces edges and faces. The edges are more
irregular in shape, exhibit a smaller surface area, and
possess both positive and negative charges which are
primarily due to broken bonds. The faces are generally
flat and account for the majority of the surface area of
the platelet. In the case of montmorillonite and illite,
the faces possess an electron-rich surface due to the
presence of oxygen atoms in the tetrahedral layer. The
montmorillonite clays are also characterized by the
substitution of Mg++ for Al3+ in the octahedral layer. This
substitution results in a net negative charge imbalance
which may be satisfied by cations situated on interlayer
positions (faces) and to a lesser extent on the platelet
edges. The magnitude of the electrostatic charges and the
resulting attractive forces is intensified by the extremely
small size of the clay mineral platelets (Snethen et al.,
1977) .
Clays attract and hold water molecules through the process
of hydrogen bonding of the water molecules to the clay
mineral surface and dipole-dipole attraction of the water
molecules for one another. The hydrogen bonds may develop

by either the oxygens attracting and bonding with the
positive side of the dipolar water molecule or the
hydroxyls attracting and bonding with the negative side of
the water molecule. The hydrogen bonding of water
molecules to the clay mineral surface provides the basic
building blocks for the double layer water which is the
driving force in soil expansion (Snethen et al., 1977).
Another contributing mechanism to the volume expansion of
clay minerals is known as cation hydration. Instead of the
clay particles being surrounded by water molecules to
balance the positive charge deficiency, cations such as
calcium, magnesium, sodium or potassium are attracted to
the clay surface. Hydration of these cations, which
results from the charge of the cation not being fully
neutralized, provides a considerable force for the
attraction of water. Thus the influence of cation
hydration involves both an attraction force for water
molecules and a physical increase in size (ionic radii)
following hydration. Relation of Structure to Properties
Minerals of the kaolinite group, because layers in the
structure are held tightly by the opposition of Cf and OH'
and because little substitution of other cations for Al3+

and Si4+ is possible, show less capacity for adsorbing ions
and water than do other clays. Montmorillonite, with
layers more easily separable and with extensive
substitution, has an abundance of adsorbed ions and shows
the capacity to adsorb water to an extreme degree.
Montmorillonite therefore is much more plastic than
kaolinite. The degree of plasticity depends somewhat on
the kind of adsorbed ions, being greatest for sodium and
hydrogen montmorillonites and least for calcium
montmorillonites. The extensive use of montmorillonite
clays in drilling muds and in filters to remove
undesireable ions and organic coloring materials from water
and other liquids depends likewise on their adsorptive
capacity (Krauskopf, 1967). Illite clays, with structural
layers partly bound together by nonexchangeable cations,
show intermediate plasticity and ion-exchange capacity.
Most clays in nature are mixtures of two or more clay
minerals and their properties are accordingly intermediate
between the extremes. The properties of a clay mineral may
be greatly influence by a minor constituent; in particular,
small percentage of montmorillonite in a clay can radically
change its plasticity. This ability of certain clays to
modify the properties of mixtures makes it difficult to

predict the behaviour of a given sample from its
composition alone (Krauskopf, 1967).
2.2.2 Inital Moisture Content
Chen (1988) presented results from a series of tests to
determine the relationship between the initial moisture
content of an expansive soil and its volume change and
swelling pressure. Soil samples were compacted at constant
density but varying initial moisture contents and subjected
to free-swell/reload tests. As expected, the specimens
with low initial moisture content swell the most, and the
slope of the e-log p curve decreases as the initial
moisture content increases. However, the swelling pressure
required for zero volume change remains practically
The variation of moisture content versus volume change is
shown in Fig. 2-5. For Chen's test soil, the results
indicate that with moisture content slightly higher than
optimum moisture content, the volume change should be
negligible. This is the basis for the commonly accepted
procedure of prewetting the foundation soil to control
swelling characteristics. However, where high moisture
content soils will experience less heave, footings founded
on an expansive soil will experience the same swelling

pressure, and the same amount of dead load pressure will be
required to insure zero volume change (Chen, 1988).
Moisture Content
Figure 2-5. Effect of varying moisture content on
volume changes from constant density
samples (after Chen, 1988).
2.2.3 Initial Density-
Laboratory studies by Chen (1988) and others have
demonstrated that both volume change and swelling pressure
increase with increasing dry density. Furthermore, the

; initial density, whether undisturbed or remolded, is the
| only element that influences the pressure exerted by a
swelling soil. Chen suggests that swelling pressure can be
used as a yardstick for measuring expansive soil as it
reflects only the swelling characteristics of the soil and
will not be changed by placement or environmental
conditions. The swelling pressure of a clay is independent
of the surcharge pressure, initial moisture content, degree
j of saturation, and the thickness of the stratum (Chen,
| 1988) .
| The linear relationship between dry density and volume
change is shown in Fig. 2-6, which is adapted from Chen's
| experimental data. Figure 2-7 is a semi-log plot showing
the effect of varying density on the swelling pressure of
an expansive soil. This curve can be expressed as:
log y = ax-b
where: y = swelling pressure
x = dry density, and
a and b = constants depending on soil property,
and "a" is the slope of the curve (from Chen, 1988).

Volume Change
Dry Density (pcf)
Figure 2-6. Effect of varying density on volume change
for constant moisture content samples
(reproduced from Chen, 1988).

Figure 2-7. Effect of varying density on swelling
pressure for constant moisture content
samples (reproduced from Chen, 1988).
2.2.4 Surcharge Pressure
It is well recognized that the magnitude of swelling can be
controlled if sufficient restraining pressure is applied to
the expansive clay. Many studies in the literature support
the fact that for conditions of equal initial dry density

and moisture content, the swelling potential is reduced by-
increasing the surcharge or confining pressure. In this
regard, results from Chen (1988) are presented in Figure 2-
8. The shape of the relationship between volume change and
surcharge pressure is a hyperbolic curve and the
intersection of the curve and the abcissa indicates the
pressure required for zero volume change. This pressure by
definition is the soil swelling pressure (Chen, 1988).
Surcharge Pressure (psf)
Figure 2-8. Effect of varying surcharge pressure for
constant density and moisture content
samples (from Chen, 1988).

In the one-dimensional swell test (ASTM D 4546), the
surcharge pressure applied to the soil specimen in the
consolidometer simulates the dead load pressure exerted on
the footings or pier foundation. This surcharge load is
essential to control foundation movement. Chen (1988)
points out a spread footing foundation can generally be
used if the soil swelling pressure is less than about 239
kPa (5,000 psf) Thus, in principle, volume change in the
soil will be constrained upon soil wetting as long as a
minimum surcharge of 239 kPa (5,000 psf) is maintained.
The required dead load pressure may be drastically reduced
if a certain amount of differential uplift can be tolerated
by the structure.
For highly swelling soils a pier foundation will be
required to contentrate the dead load on a smaller area.
Dead load pressures in excess of 958 kPa (20,000 psf) are
not difficult to achieve in a small-diameter pier; however,
the swelling effect along the shaft of the pier must also
be considered. The pressure exerted along the pier shaft
can be many times greater than the pressure exerted on the
bottom of the pier. Dead load pressure alone is generally
not sufficient to prevent uplifting and should be assisted

by anchorage of the pier in a zone not affected by moisture
changes (Chen, 1988).
2.2.5 Soil Structure and Fabric
In fine-grained cohesive soils, the soil structure is taken
to' mean both the geometric arrangement of mineral grains
(fabric) as well as the interparticle forces which may act
between them. All the clay structures in nature result from
a combination of the geologic environment at deposition,
the subsequent geologic and engineering stress history, and
the nature of the clay mineral. The soil structure
strongly affects the engineering behaviour of a particular
Parcher and Liu (1965) point out that in naturally
sedimented soils, the arrangement of the soil grains will
be affected by the nature and sizes of the particles and by
their depositional environment. During compaction, the
arrangement of grains in the soil structure will be
influenced by the amount of compactive effort, the method
used to compact the soil, and the moisture content during
compaction. These same factors also produce measurable
effects on the swelling that may be realized by the soil
under a given set of conditions.

Since the clay minerals posses a sheet-like structure, it
follows that any mechanism that would tend to enhance the
preferential alignment of the clay particles would have an
effect on the soil's swelling potential. Soils with
particles preferentially aligned, swell more in directions
normal to the clay platelets than in directions parallel to
the platelets (Gromko, 1974) Static compaction tends to
produce more swell than kneading or dynamic compaction.
The soil structure also also affects the permeability of a
soil deposit. Chen (1988) points out that the permeability
of a soil deposit determines the rate of progress of water
into the soil, be it by gravitational flow or by diffusion.
The rate of water flow into the expansive soil then
directly affects the rate at which heave will occur.
2.2.6 Availability of Water
As is noted by Seed et al. (1962) and Chen (1988),
irrespective of the soil's swelling potential, no volume
change will occur if the moisture content remains
unchanged. Volume expansion, both in the vertical and
horizontal directions, will develop with a change in
moisture content. Generally the more water made available
to an expansive soil, the greater probability that it will
realize its full swelling potential, for a given set of

confining pressure, initial dry density, and initial
moisture conditions.
It is important to note that complete saturation is not
necessary to accomplish swelling. Detrimental swelling can
occur from only slight changes in moisture content, on the
order of only 1 or 2 percent. Some examples include clay
specimens which swell in the consolidometer with slight
increase in humidity, and the severe cracking of a floor
slab when the soil moisture content is increased slightly
due to local wetting (Chen, 1988).
The transmission of water to the foundation soil can occur
in a variety of ways. The most common method is by
gravity. The seepage of surface water from precipitation,
snow melt, and landscaping are a few common examples.
Shrinkage cracks which develop from surface dessication
provide channels for the downward flow of water into the
deep soils.
The flow of moisture can occur in all directions including
upward when under artesian conditions. In well-
consolidated clays and shale bedrock, flow is generally
confined to bedding planes and/or continuous fractures and
fissures. Fine-grained soils are also susceptible to
significant amounts of water transfer due to capillary

force. The height of water rise into the capillary fringe
varies inversely with the radius of the capillary tube
(Chen, 1988).
Chen (1988) points out that the heaving of expansive soils
may take place without the presence of free water. In this
case, the process of vapor transfer through the soil plays
an important role. Water vapor in an area of higher
temperature will migrate toward the cooler area in an
attempt to lower the thermal energy gradient between the
two. Upon reaching the cooler area, such as beneath
buildings, roadways, and sidewalks, condensation can take
place and provide sufficient moisture to initiate swelling.
2.2.7 Type and Concentration of Electrolytes
As mentioned previously, the surface attraction
relationships which exist between clay minerals, between
clay minerals and water, and between clay minerals and
cations are a result of the shape and crystallographic
structure of the clay mineral. Clay minerals occur as tiny
platelets having two types of exposed surfaces edges and
faces. Isomorphous substitution in the smectite clays,
particularly montmorillonite, results in a net negative
charge imbalance which may be satisfied by cations situated
on interlayer positions (faces) and to a lesser extent on

the platelet edges. The magnitude of these electrostatic
charges and the resulting attractive forces is intensified
because of the extremely small size of the clay mineral
Cations such as Ca*+, Mg++, Na+, and K+, are common in pore-
water solutions and are readily adsorbed onto the surface
of individual clay layers in the effort to render it
neutral from a charge deficiency sense. The depositional
history and subsequent weathering environment of a
particular soil deposit will largely determine which ions
are present. These cations are exchangeable; i.e., they
can be replaced by one or more cations that can satisfy the
valence deficiency of the clay crystal.
The exchangeability of cations depends primarily on their
valence and ionic radius. Higher valence cations easily
replace lower valence cations; for ions of identical
valence, the larger ions can more readily replace smaller
ions. For example, Ca++ will tend to replace Mg'"", K+,
and/or Na+; whereas K+ will more easily replace Na+ in the
absence of other more active ions. In general, the greater
the replacement ability of cations, the stronger the ionic
bonds that develop between the cations and clay crystals,
and the more stable the clay structure becomes. Thus, it

is less easy for water to enter the double layer and cause
volume expansion. This is the basic concept behind the
chemical stabilization of expansive soils.
2.2.8 Thickness of the Soil Stratum
Chen (1988) conducted a series of tests to explore the
effect of soil stratum thickness on the amount of volume
change and swelling pressure. The thickness of the test
samples ranged from 13 mm to 38 mm (0.5 to 1.5 inches) and
the samples were compacted to uniform moisture content and
density, and allowed sufficient time to complete saturation
of the thickest sample.
The results of Chen's tests are shown in Figure 2-9. The
magnitude of the volume change is proportional to the
volume thickness of the sample; however, the percentage of
volume increase remains relatively constant. Additionally,
the shape of the reloading curve in these tests remained
almost identical for various sample thicknesses and the
measured swelling pressure was virtually constant.

0 0.25 0.5 0.75 1 1.25 1.5 1.75
Sample Thickness (in)
Figure 2-9. Effect of varying sample thickness on
volume changes for constant density and
moisture content samples (after Chen,
As Chen points out, dead-load pressure exerted on a footing
can only control volume change of the near-surface soils.
With depth, the pressure exerted on the footing is
distributed over a larger area and is not effective in
counteracting the swelling pressure of the expanding soil.
Volume Increase

Thus deep-seated swelling is controlled only by the weight
of the overburden soil and not by dead-load pressure
exerted on the foundation system.
2.2.9 Time Available for Swelling to Occur
The time required for the soil to reach its maximum
swelling potential may vary considerably depending on the
initial dry density, permeability, soil thickness, and the
availability of water. This has implications for both soil
deposits and test specimens. Remolded soil samples
generally obtain 95 percent of the total available swell
within 24 hours; however, an undisturbed high-density clay
shale may require several days or even a week before
complete saturation can be achieved (Chen, 1988). The
standard 24-hour swell period, as is practiced in many
geotechnical laboratories, may not be sufficient to allow
full swelling to be realized in all cases.
As a practical matter, there is ample time available for an
expansive soil deposit beneath a constructed building to
adsorb moisture on the way to realizing its full swelling
potential. The length of time it takes will depend on the
predominant moisture transfer process: faster if it is
mostly through seepage of surface water through fractures
and/or shrinkage cracks; slower if dominated by capillary

forces or thermal gradients. Since moisture transfer is a
relatively slow process, it is not suprising that distress
of a building often takes place several years after
occupancy (Chen, 1988).
2.2.10 Swelling Fatigue
Chen (1988) notes that pavements founded on expansive clays
which have undergone seasonal movement due to wetting and
drying have the tendency to reach a point of stabilization
after a number of years. Laboratory clay samples show
signs of fatigue when they are subjected to full swelling
in a consolidometer, allowed to dessicate to their initial
moisture content, then flooded with water and permitted to
swell again. The amount of swelling is observed to
decrease with each successive wetting and drying cycle
(Chen, 1988).
Al-Homoud et al. (1995) studied the cyclic swelling
behaviour of six expansive soils in the city of Irbid,
northern Jordan. Remolded specimens were compacted into
the consolidation ring at their natural water content and
allowed to swell under a vertical confining pressure of 6.9
kPa (1 psi). The test sample was then dried to its initial
moisture content before being wetted and allowed to swell

again. After each cycle, the swell potential and swell
pressure were measured.
The results of Al-Homoud et al. confirm the earlier
observations of Chen. In all cases, the swell percent is
maximum for the initial cycle but decreases steadily to
arrive at an almost constant value after the fourth or
fifth cycle. Also, the time required to achieve maximum
primary swell is remarkably less in the fifth cycle as
compared to the initial cycle (480 1,080 min as opposed
to >1,440 min). The authors attributed these phenomena to
the continuous rearrangement of clay particles during
cyclic wetting and drying as observed through scanning
electron micrographs.
2.3 Thermal Treatment Effects
The literature review uncovered several studies regarding
the application of heat energy to soils and its effects on
resulting soil properties. Results of this research are
described in the following sections.
2.3.1 Strength
The consolidated-undrained triaxial tests of Wohlbier and
Henning (1969) demonstrated significant shear strength

increases in kaolinite specimens when heated above 400C.
Kaolin IA from Zettlitz, former USSR, was selected for its
optimal properties of drying and heating, and previously
well characterized thermal reactions. Specimens were
prepared at an initial water content of 48 to 50 percent
using a vacuum extrusion technique. After careful drying
for 28 days, the specimens were heated for a period of 48
hours to temperatures up to 600"C using a heating rate of
60'C per hour. The treated samples were kept in a
dessicator for 14 days, following which, both dry and
water-saturated specimens were tested for shear strength.
The triaxial tests were carried out at different confining
pressures to define the Mohr failure envelope for every
pretreatment temperature.
Results of Wohlbier and Henning's work are shown in figures
2-10 and 2-11 for the dry and water-saturated samples,
respectively. It is apparent from the results that little
change in shear strength occurs until the specimens are
heated over 400'C. Above this temperature, the kaolinite
was reported to loose all plasticity, resist swelling when
exposed to water, and assume more and more the character of
a brittle rock.

0 10 20 30 40 50 60 70
a (kg/cm2)
Figure 2-10. Effect of heat treatment on dry strength of
kaolinite (after Wohlbier and Henning,
0 10 20 30 40 50 60 70
o' (kg/cm2)
Figure 2-11. Effect of heat treatment on the rewet
strength of kaolinite (after Wohlbier and
Henning, 1969).

From previous publications in ceramic research, Wohlbier
and Henning expected the transformation of kaolin to
metakaolin to occur around 440C to 450C. This
transformation occurs from the loss of structurally held
OH" ions in the form of water. The attendant change in
crystalline structure essentially results in formation of a
new clay mineral. The differential thermal analysis on the
Zettlitz kaolin indicated a strong endothermic reaction
ranging from about 450C to 600'C, which may have been
shifted due to the relatively high velocity of heating the
samples. Wohlbier and Henning confirmed the structural
transformation of the kaolinite by comparison of X-ray
diffraction patterns between temperatures of 400"C and
The work of Wohlbier and Henning also covered the extent of
change in other soil characteristics as a result of thermal
pretreatment. The results of their investigations are
shown in Table 2.1. It is evident from these data that
significant changes in specific gravity, liquid limit, and
plastic limit occurred when the specimens were heated to
temperatures greater than 400C. Up to 400'C, a gradual
reduction in the plasticity index occurs whereas at 500C
and above, the soil loses all plasticity. These changes in
soil properties are directly related to alteration of the

clay mineral crystalline structure as the kaolinite is
heated above 400C.
Table 2.1 Results of Heat Treatment on Zettlitz Kaolin IA
(after Wohlbier and Henning, 1969).
Heat Treatment Specific Gravity Gs (g/cm3) Void Ratio of Dry Samples e Liquid Limit % Plastic Limit % Plasticity Index %
Untreated 2.629 - 52.9 34.6 18.3
48 hr 105C 2.629 0.800 53.0 34.0 19.0
48 hr 200C 2.622 0.814 52.5 33.6 18.9
48 hr 300C 2.622 0.819 52.0 35.1 16.9
48 hr 400C 2.620 0.815 52.2 37.8 14.4
48 hr 500C 2.456 0.860 - - -
48 hr 600C 2.485 0.848 - - *
Joshi et al. (1994) investigated the effect of heat
treatment on the strength of kaolinite, bentonite,
kaolinite-bentonite mixtures, and natural clay sediments
from the Western Beaufort Sea. Compressive strengths in
this study were determined through constant-strain,
unconfined compression tests. The authors compared the
strength of specimens that were oven dried at 100"C to
those thermally treated at 300, 400, 500, 600, and 700"C.
In all of the clays tested, unconfined compressive strength
increased gradually with an increasing temperature, with
more pronounced effects in the bentonite than kaolinite.

Joshi et al. also performed thermogravimetric analyses of
the clays and clay mixtures. This type of test measures
weight loss with temperature and is used to determine the
temperatures at which the different types of water are
released through the application of heat energy. At lower
temperatures, free water in the pores and that adsorbed on
to the clay particles is dissipated. At higher
temperatures, structural water is released in a prominent
endothermic reaction and accompanied by a marked reduction
in weight. Joshi et al. refer to this as "dehydroxylation"
and found the dehydroxylation temperature to range between
300C and 500'C in the kaolinite specimens and between
500'C and 700C in bentonite. Samples thermally treated to
temperatures higher than dehydroxylation temperatures
gained significant increases in strength and resistance to
disintegration when exposed to water. The release of
structurally held water during dehydroxylation is an
irreversible reaction which limits the clay's ability to
readsorb moisture.
2.3.2 Pulverization and Stabilization
One commonly used technique in the construction industry is
the stabilization of highly plastic clays through the
addition of chemical stabilizers such as lime and cement.

The workability of the clay becomes an important factor as
the process involves adequate pulverization of the soil,
optimum mixing of the soil with the stabilizer (with water
if necessary) and compaction. Chandrasekharan et al.
(1969) studied the influence of thermal treatment on the
pulverization characteristics of two typical tropical soils
(lateritic and black cotton soils), and the effect of heat
and aggregate size on cement stabilization of a pulverized
black cotton soil. During the course of testing, the
authors noted a rapid reduction in the plasticity index for
black cotton soil in the temperature range from 300"C to
400'C, while no significant change was observed in the
lateritic soil when heated to a maximum of 300*C. The
change in consistency of the black cotton soil was mostly
due to a decrease in the liquid limit, and was attributed
to the change in moisture adsorption characteristics of the
predominant clay mineral montmorillonite. It should be
noted that the lateritic soil was only heated to 300*C and
this may not have been enough to effect a change in the
kaolinite which the authors believed to be the dominant
clay mineral present. The findings of Wohlbier and Henning
(1969) would seem to support this conclusion.
The work of Chandrasekharan et al. is notable in
demonstrating that pulverization of high plasticity soils

is enhanced though heat treatment alone, and can be further
aided by the addition of sodium chloride prior to heating.
2.3.3 Fusion of Soils
Fusion of the soil particles will eventually result if
sufficient heat energy is applied to raise the soil
temperature up to and beyond its melting point. Post and
Paduana (1969) point out that most soils consist of
mixtures of minerals comprised of several different oxides
and therefore will exhibit incongruent melt ranges rather
than a single melting point. Most soils exhibit a melt
range that varies between 1,250C (2,282F) and 1,750C
(3,182F), with fat clay soils tending to melt at about
1,700C (3,092F) and the feldspar-rich soils tending to
melt around 1,300C (2,372F). Post and Paduana (1969)
show that the melt range of mineral aggregates can be
lowered through the addition of fluxing agents such as soda
In general, effective thermal stabilization of soils may be
accomplished by firing the dried soil to a temperature of
about 600C (1,112F), a process known as bricking, or by
heating the soil above incipient fusion temperatures. It
is only necessary to reach the inital melt range of the
soil to effectively cement the soil mass. Soil bricking is

caused by the loss of
structural water by the soil minerals
with the formation of
new minerals.

3.1 General
As mentioned earlier, the fundamental objective of this
study was to investigate the capability of thermal
treatment to reduce (and potentially eliminate) the
swelling properties of a naturally occurring expansive clay
soil. The focus of this study was on "proof of principle"
as determined from a series of controlled laboratory
experiments. It was not a formal objective of this study
to determine whether thermal treatment would be a viable
alternative to current soil remediation practices, or even
be feasible from an economic standpoint. However, having
shown the capability of thermal treatment to reduce
swelling properties in expansive clays, it was further
desired to correlate this reduction with representative
thermal treatment parameters. Doing so would allow follow-
on work to address the application of thermal treatment
techniques, treatment system economics, and possible field

3.2 Test Approach
In order to provide an adequate challenge for the thermal
treatment technique, it was desired to select a naturally
occurring soil with a known, and preferably high swelling
potential. Discussions with several local geotechnical
engineering firms identified potential sampling sites near
the foothills of the southwest Denver, Colorado,
metropolitan area. The suburban developments in this area
have encountered widespread problems with expansive soils
resulting in substantial damage to residential and roadway
foundations (Gill et al., 1996; Nichols, 1991). For ease
of access, a newly developing subdivision located on the
north side of The Meadows Golf Course was chosen for
sampling. All of the soil used in the test program was
obtained from the same location and sampling depth.
A relatively simple experimental program was designed to
measure the results of thermal treatment on direct and
indirect parameters of expansive soils. A total of 10
samples were heated from 150C to 800C (302F to 1,472F)
for durations ranging from 24 to 240 hours. Following
thermal pretreatment, the soil samples were tested for
swelling potential, swelling pressure, Atterberg limits,
and specific gravity. Additional properties of the
pretreated soil were provided from unpublished term

projects completed at the University of Colorado at Denver
during the Spring of 1997. These included moisture-density
relationships under Standard Proctor compaction conditions
and particle size analyses. Since the data sets were
collected under different quality control conditions, they
were analyzed separately; the unpublished data sources were
used for trend analysis only and were not incorporated into
the formal statistical comparison analysis performed under
this study. A summary of pretreatment conditions for
individual test samples and the post-treatment soil
properties measured are shown in tables 3.1 and 3.2.
Table 3.1. Thermal Pretreatment Conditions for Test
Thermal Pretreatment Conditions
Test No. Temperature (C) Heating Duration (hrs)
T-l (room temp.) constant
HT-1 150 24
HT-3 150 120
HT-4 150 240
HT-6 250 24
HT-7 250 72
HT-8 250 120
HT-9 250 240
HT-5 400 24
HT-10 600 24
HT-11 800 24
Note: Test No. HT-2 was not performed

Table 3.2 Post-Treatment Soil Properties Measured in
the Laboratory Test Program
Post-Treatment Soil Parameter Test Reference
Swell Potential ASTM D 4546, Method B
Swelling Pressure ASTM D 4546, Method B
Liquid Limit (LL) ASTM D 4318
Plastic Limit (PL) ASTM D 4318
Specific Gravity (Gs) ASTM D 854
Optimum Moisture Content (OMC) (1) ASTM D 698-78 (Std. Proctor)
Maximum Dry Density1* ASTM D 698-78 (Std. Proctor)
Particle Size Gradation: Mechanical(1> Hydrometer(1> ASTM D 421 and ASTM D 422
(1) Data supplied from unpublished student term project at the
University of Colorado at Denver.
Throughout the testing program it was attempted to limit
the influence of other variables that could affect the
soil's expansion behaviour. For example, it is well
established that the initial dry density and moisture
content of remolded expansive soil specimens have a
significant impact on the swelling potential and swelling
pressure (Chen, 1988). Thus it is was important to limit
moisture/density variations in the specimens as they were
prepared for swell-pressure testing. This was accomplished

by closely monitoring the initial moisture content of the
soil and by developing a precision method of compacting the
soil specimens. Likewise, once the soil was prepared for
remolding, it was not allowed to dry out in order to
minimize the influence of wetting/drying cycles. As noted
earlier, Chen (1988) and Al-Homoud et al. (1995) have
demonstrated the tendency for an expansive soil to decrease
in swelling potential with increasing numbers of
wetting/drying cycles.
3.3 Sampling Location and Geologic Conditions
The soil used in this study was collected from within the
Fairway Vista Subdivision, located off Coal Mine Avenue,
just east of Simms Street, in the city of Lakewood,
Colorado (see Figure 3-1). This site is underlain by the
Pierre Shale and lies within the Jefferson County
Designated Dipping Bedrock Area (DDBA).
Schultz et al. (1980) describe the depositional environment
of the Pierre Shale and equivalent rocks of Late Cretaceous
age as a north-south trending trough nearly 1,600 km (1,000
mi) across that extended from the Canadian arctic at least
as far south as New Mexico and at times to the Gulf of
Mexico. During Pierre time several major marine
transgressions and regressions occurred, leaving behind a

wedge-shaped sedimentary deposit that thins from more than
1,000 m in the west to several hundred meters in the east.
Figure 3-1
Location map of sample collection site for
expansive soil used in this study.

The source areas for Pierre Shale sediments were the
mountain ranges to the west of the marine basin. The
sediments range from very fine grained shales and
marlstones in the east, far from the source areas, to silty
shale and nearshore marine siltstone and sandstone in the
middle and western parts of the basin. Schultz et al.
(1980) report that mixed-layer clay is the dominant type of
clay in the Pierre Shale of the northern Great Plains
Region. The layers comprise 20 to 60 percent illite, about
35 percent beidelite, and the remainder montmorillonite.
Bentonite beds are locally present in the Pierre Shale and
commonly consist of more than 90 percent smectite in which
montmorillonite is interlayered with a smaller amount of
The Jefferson County DDBA is part of a wide zone of heaving
bedrock that extends along the Front Range Piedmont of
Colorado. Geologically, the DDBA is characterized by
steeply dipping sedimentary bedding, with dip angles
greater than 30 degrees from the horizontal, and zones of
expansive (swelling) claystone bedrock. A high rate of
damage to roads, utilities, and structures has occurred
where the steeply dipping beds of expansive claystone are
found in close proximity to the ground surface. The
combination of bedding angle and swelling characteristic

produces a particularly destructive linear style of ground
deformation in this area (Noe and Dodson, 1995).
Numerous single-family residences were under construction
in the Fairway Vista Subdivision when soil samples were
collected during the Summer of 1996. Approximately 50 kg
(100 lb) of olive-gray shale was collected from Lot #25
which is located along the southern boundary of the
subdivision. The back of this lot faces south and
overlooks The Meadows Golf Course.
Soil samples were pulled from the side wall of a foundation
cut approximately 1.2 m (4.0 ft) beneath the top of the
excavation. Information provided by the consulting
geotechnical engineer indicated that the lot was located in
fill material (Tim Spencer, GTG Geotech., personal
communication). The natural moisture content of the fill
was measured at 19.7% with an in-situ dry density of 16.24
kN/m3 (103.4 pcf). A one-dimensional swell test performed
on a soil boring from Lot #25 indicated a swell potential
of 10.2% against a simulated vertical pressure of 23.94 kPa
(500 psf).
3.4 Soil Properties
Several index property tests were performed in order to
characterize and classify the soil material. Moisture-

density relationships were obtained under standard Proctor
test conditions. Results of these tests are summarized in
Table 3.3 and figures 3-2 and 3-3.
Table 3.3. Results of Index Property Tests on
Soil Used for Thermal Treatment Studies
Soil Property Value
Maximum Dry Density 15.25 kN/irf (97.10 pcf)
Optimum Moisture Content 25.1 %
Liquid Limit (LL) 58 %
Plastic Limit (PL) 27 %
Plasticity Index (PI) 31 %
Specific Gravity (Gs) 2.79
Percent Passing No. 8 Sieve 99.8
Percent Passing No. 200 Sieve 95
Percent Finer than 5 |i 45.6
Percent Finer than 2 (I 34.2
USCS Soil Classification CH
Activity PI / (% finer than 2 |l) 0.91

17.00 t
Figure 3-2. Moisture-density curve for untreated soil
used in the thermal treatment studies.
As illustrated in Figure 3-2, the maximum dry density
obtained from the Standard Proctor moisture-density curve
was approximately 1 kN/m3 (6 pcf) lower than the in-situ
dry density measured at the Lot #25 sampling site.
Geotechnical investigations of other nearby lots in the
subdivision indicated in-situ dry densities ranging from
16.2 to 17.3 kN/m3 (103 to 110 pcf) with moisture contents
between 18 and 22% (personal communication, Tim Spencer,

GTG Geotech). To ensure a highly swelling condition for
heat-treated samples, a target value of 17.28 kN/m3 (110
pcf) dry density at 18.5% moisture content was selected for
use in the remolded laboratory swell tests.
Particle Size Distribution
Figure 3-3. Typical gradation curve for untreated soil
used in the thermal treatment studies.
Clay mineral identification techniques such as X-ray
diffraction and differential thermal analysis were not
available for this study. However, an indication of the
clay mineral content of the test soil was provided by the

Atterberg limits and the characteristic positions of common
clay minerals as plotted on Casagrande's plasticity chart.
The position of the test soil in relation to other clay
minerals is illustrated in Fig. 3-4. This position agrees
reasonably well with the regional clay content of the
Pierre Shale as reported by Schultz et al. (1980).
Liquid Limit
Figure 3-4. Relation of test soil to the positions of
other clay minerals on Casagrande's
plasticity chart (after Holtz and Kovacs,

3.5 Sample Preparation and Testing Procedures
Sample preparation and testing in this study were performed
in substantial accordance with applicable ASTM protocols.
The following sections describe the procedures used in
certain portions of the testing program which were
developed to satisfy unique test objectives.
3.5.1 General
The clay soil was prepared for thermal pretreatment by
first allowing it to air dry for several days. This
improved the workability of the soil and allowed it to be
mechanically size reduced in a steel pan using a rubber
mallet and small sledge hammer. Following size reduction,
the soil was screened through a No. 4 sieve and passing
material was mixed thoroughly in a clean steel tray. The
soil moisture content was then returned to its in-situ
state (approximately 20%) so that natural conditions would
be simulated prior to the thermal pretreatment process.
Approximately 5 kg (10 lb) of processed soil was prepared
at a time. Samples were collected for determination of
initial moisture contents. Then deionized water was added
to the soil to raise the moisture content up to in-situ
levels. Using a spray bottle, the water was added in small
amounts and the soil was thoroughly mixed by hand after

each incremental addition. The batch was then split into
individual plastic bags and set aside to cure for a minimum
of 24 hrs before extracting the first samples for thermal
pretreatment. All soil prepared for thermal pretreatment
was processed in the same manner.
3.5.2 Thermal Pretreatment
A commercial laboratory oven was used to heat soil
specimens up to temperatures of 250"C (482F). This model
of oven was a Stabil-Therm manufactured by Blue M
Electric Company of Blue Island, Illinois (see Fig. 3-5).
The oven has a working range up to 300C; however, a faulty
limit breaker prevented the oven from holding a constant
temperature above about 260"C. Consequently, another
furnace had to be used to heat the soil to greater
Sample batches for thermal treatment in the laboratory oven
were measured to approximately 1,000 g of moist soil and
heated in large metal bowls. The metal bowl and soil were
placed in a preheated oven and cooked for the desired time
interval. At the end of the heat treatment cycle, the
bowls were removed from the oven and set aside to cool for
several minutes. The weight of the bowl and dry soil were

Figure 3-5. Photograph of Stabil-Therm oven used for
pretreatment of soil samples to 250C.

measured to verify the beginning moisture content of the
At temperatures above 250C (482F), a muffle furnace was
used for heating the soil samples. This furnace was
located at the U.S. Bureau of Reclamation, Earth Sciences,
Engineering, and Materials Laboratory located at the Denver
Federal Center in Lakewood, CO. The muffle furnace, Model
No. F1930-1, was manufactured by Thermo Electric
Manufacturing Company of Dubuque, Iowa. The upper
operating limit on this furnace is 1,649'C (3,000*F).
A similar batch process was used to heat the soil samples
in the muffle furnace with the exception that the heating
duration was limited to 24 hours. Also, due to the
interior dimensions of the muffle furnace, only about 600 g
of soil could be heated at a time. Three samples of moist
soil were placed in ceramic bowls and weighed prior to
inserting them into the muffle furnace. Each bowl was
capable of holding approximately 200 g of moist soil. The
three bowls were placed into the muffle furnace at room
temperature and brought up to the desired heating
temperature at a rate of about 20C (68"F) per minute. The
soil samples were heated for 24 hours at which time the
furnace was turned off. The samples were allowed to cool

down in the furnace (with the furnace door closed) until
the following day. The weight of the dry soil and ceramic
bowls was measured and the samples were then transported to
the CU-Denver soils laboratory for subsequent sample
preparation. The thermally pretreated soil being removed
from the muffle furnace following a heating cycle is shown
in Fig. 3-6.
3.5.3 One-Dimensional Swell Test Specimen Preparation
The initial step in preparing specimens for the one-
dimensional swell test was to raise the moisture content of
the thermally pretreated soil up to 18.5%, which was the
target moisture content selected for the remolded soil
specimens. Based on the dry weight of the treated soil
batch, an appropriate amount of water was added using the
same hand spray/mixing technique described previously. The
soil samples were then placed into plastic bags and allowed
to cure for a minimum of 24 hours.
In order to accurately and repeatedly produce swell test
specimens at a consistent dry density, a load cell was used
to compact the soil into a steel confining ring. The type
of load cell used for this purpose was a Material Test

Figure 3-6. Photograph of Thermo Electric Mfg. Co.
muffle furnace used to pretreat soil
samples above 250C.

System (MTS) Model 810, manufactured by Materials Systems
Corporation of Minneapolis, Minnesota. The hydraulic
actuator on this frame has a capacity of 22 kips-force with
a piston travel of 15.24 cm (6.00 in). The configuration
of the load cell, set up for compacting the swell test
specimens, is shown in Fig. 3-7.
Soil specimens were pressed into the confining ring in
three lifts. The mass of moist soil in each lift was
weighed carefully in order to prepare the specimen to the
desired dry density. The top of each lift was scored
between presses. A dial gage was attached to the load
frame to record the height of the hydraulic ram and
accurately press each lift in the soil specimen to the
required height. Various wood blocks and steel disks were
set onto the load cell and leveled to raise the
consolidation ring to the proper height. The compaction
force required to press each lift into the consolidation
ring was recorded initially to assure an even compactive
effort was being applied. After the first two samples, the
compaction effort was found to be highly consistent between
lifts so this practice was subsequently discontinued.

Figure 3-7. Photograph of load cell used to prepare
pretreated soil specimens for swell
68 Oedometer Cell
A standard fixed ring oedometer cell was used to determine
the swelling potential and swelling pressure of all
specimens in this study. The cell is fitted onto a steel
loading frame (consolidometer) which is equipped with a
mechanical lever arm to apply pressure to the specimen.
Weights are placed on a stand atached at one end of the
arm. The mechanical advantage of this system is 11:1.
Dial gages on the consolidometer are used to measure the
one-dimensional expansion/contraction of the test specimen
as weight is applied or removed. Resolution of the dial
gage is 1CT4 in. Figures 3-8 through 3-13 show the
oedometer in various stages of assembly and the complete
cell in position on the consolidometer platform ready for
Specimen dimensions are 63.5 mm (2.50 in) in diameter by
19.05 mm (0.75 in) in height. The specimen was remolded
into a stainless steel confining ring that has a sharp
cutting edge on one end. Porous stones were situated
directly on the top and bottom of the soil specimen.
Filter paper was not used due to its high compressibility.
The specimen and ring were placed in a clear lucite
saturation cell and held in place by means of a stainless

Figure 3-8. Disassembled standard fixed-ring oedometer
cell. Parts shown clockwise from top left:
(1) lucite cell wall attached to base; (2)
bottom porous stone; (3) confining ring;
(4) locking ring with thumbscrews; (5) top
cap with attached porous stone.

Figure 3-9. Stainless-steel confining ring placed on
bottom porous stone inside the oedometer

Figure 3-10. Photograph of locking ring installed over
the steel confining ring and secured with
thumbscrews. The locking ring holds down
the confining ring against the bottom
porous stone and base of the oedometer

Figure 3-11. Top view of the oedometer cell, fully
assembled with top porous stone and cap in
position on top of the specimen.

Figure 3-12. Close-up view of the consolidometer
platform. The recess in the base of the
standard oedometer cell fits snugly onto
the raised aluminum disk at the center of
the platform.

Figure 3-13. Photograph of standard oedometer cell
placed on consolidometer platform, ready
for testing.

steel locking ring. A one-piece stainless steel loading
cap and porous stone is placed directly on top of the
specimen. This cap fits into an adjustible pin on the
consolidometer when the oedometer cell is fitted into
place. Procedure
Measurements of swell potential and swelling pressure of
soils in this study were performed in substantial
conformance to ASTM D 4546, "Standard Test Methods for One-
Dimensional Swell or Settlement Potential of Cohesive
Soils". This standard provides for three alternative
methods to be used; conditions common to each method are
that the soil specimen is restrained laterally and loaded
axially in a consolidometer with free access to water.
Method B was chosen for use in this study as it widely used
by geotechnical laboratories in the Denver metropolitan
area. This method measures the percent heave against a
predetermined vertical pressure. The magnitude of this
pressure is usually chosen to simulate the in-situ vertical
overburden pressure and/or structural loading. In order to
be consistent with prior testing at the soil sampling
location, a simulated overburden pressure of 23.94 kPa (500
psf) was selected (Tim Spencer, GTG Geotech, personal

communication). Each soil specimen was inundated with
water and allowed to swell for 24 hours. At this time, the
specimen was loaded in increments as in a standard
consolidation test, until it was compressed to the initial
void ratio obtained just prior to flooding. The pressure
at which the specimen returned to its initial void ratio
was determined to be the swelling pressure. The
swell/compression sequence vs. applied pressure for this
method is graphically shown in Fig. 3-14.
Figure 3-14. Measurement of swell potential and swelling
pressure using ASTM D 4546, Method B.

The abbreviated procedure used for the swell tests
performed in this study is as follows (refer to figures 3-8
through 3-13 for identification of individual equipment
Pedometer Cell Assembly:
Note: It is necessary to assemble the oedometer cell
quickly to minimize change in specimen water content and
volume due to evaporation. The oedometer assembly can be
covered with a loose fitting plastic membrane, moist paper
towel, or aluminum foil for this purpose. The wrapping is
cut away and discarded at the time of specimen inundation.
1. The lower oven-dry porous stone and specimen in
confining ring are placed in the clear lucite cell. It
is necessary to keep the porous stones oven dry to avoid
the specimen absorbing any water from the porous stones
and swelling before the start of the test. Note: be
sure that the cell is properly seated on the oedometer
base and that the O-ring is adequately lubricated.
2. The specimen and ring are locked in place with the
retaining ring and clamping screws.

3 The stainless steel loading cap and attached oven dry-
porous stone is placed on top of the specimen.
4. Place the assembled oedometer cell in the loading frame
as shown in Fig. 3-13. Be sure that adequate clearance
is available for the oedometer cell to fit beneath the
steel reaction pin and that the cell is properly seated
on the consolidometer platform.
Consolidometer Adustment
1. The lever arm on the consolidometer is balanced and
leveled so that the specimen loading cap just comes into
contact with the reaction pin.
2. Adjust the dial gage by means of the guide rods on each
side of the consolidometer assembly. The frame
supporting the gage is leveled and lowered into place so
that the dial gage contacts the measurement surface.
Set the pointer on the dial gage to a convenient
starting point. Be sure that the working range of the
gage is adequate for the amount of expansion/compression
expected from the specimen.
3 Apply a small seating load to the specimen by adjusting
the small sliding weights on the lever arm. This should
remove all excess "slack" from the system.

Swell Test:
1. Apply a vertical pressure within 5 min of applying the
seating pressure. As mentioned earlier, this pressure
is usually chosen to simulate an overburden pressure
and/or structural load of interest. The vertical
pressure used for this study was 23.94 kPa (500 psf).
2. Read the deformation within 5 min of placing the
vertical pressure. The specimen is inundated
immediately after the deformation is read.
3. Upon specimen inundation, the deformation is recorded at
intervals of 0.1, 0.25, 0.5, 1.0, 2.0, 4.0, 8.0, 15.0,
and 30.0 min and 1, 2, 4, 8, and 24 hrs.
4. After completion of the 24-hr swell period, the specimen
is recompressed to its initial void ratio/height. In
practice, the specimen is compressed to less than its
initial void ratio because the exact magnitude of
vertical pressure required to reach the inital void
ratio is unknown. Apply incremental vertical pressures
of approximately 48, 96, 192, 384, 768, 1,536 kPa
(1,000, 2,000, 4,000, 8,000, 16,000, and 32,000 psf)
with each pressure maintained constant for a 24-hr
period. The specimen deformation is recorded at the
same time intervals as in the swell phase of this test.

3.5.4 Equipment Calibration
As pressure is applied to the soil specimen, some
deformation of the test equipment also occurs. In the case
of the oedometer cell assembly, this deformation occurs
through compression of the porous stones and closure of
minute gaps between the porous stones and the specimen,
cell base, and loading cap. Small gaps are also present
initially between the oedometer loading cap and reaction
pin, and the oedometer base and consolidometer platform.
To account for these effects, the oedometer and
consolidometer were calibrated under anticipated test
conditions using a steel disc having approximately the same
dimensions as the test specimens. The calibration
procedure is provided in ASTM D 4546 and ASTM D 2435 "One-
Dimensional Consolidation Properties of Soils".
The steel disk measured 62.2 mm (2.45 in) in diameter and
19.0 mm (0.75 in) in height. Porous stones were soaked and
the oedometer cell was filled with deionized water. Filter
paper was not used in the equipment calibration due to its
high compressibility. The steel disk was loaded to a
maximum of 3.4 MPa (72,000 psf) in increments of
approximately 24 to 383 kPa (500 to 8,000 psf). The amount
of axial compression was measured for each load increment
using a standard dial gage with a sensitivity of 10~4 in.

Under the maximum applied pressure, 0.2057 mm (0.0081 in)
of axial compression was measured. A plot of the
calibration curve for the consolidometer is shown in Fig.
Oedometer Calibration
Axial Pressure (kPa)
Figure 3-15. Calibration of consolidometer and standard
oedometer cell; axial compression plotted
against applied axial pressure.

4.1 General
The laboratory program as described in previous sections
encompassed several commonly used measures for
characterizing soils in general, as well as several
direct tests to measure potential soil volume and
pressure changes. Test parameters included the percent
swell potential, swelling pressure, liquid limit,
plastic limit, and specific gravity of soil solids.
Soil compaction data are presented here from an
unpublished student report which was completed during
the Spring of 1997 at the University of Colorado at
Denver. This combination of parameters was selected
because they provide a relatively complete
characerization of effects caused by the thermal
treatment process, and their relationship to expansive
soils have been well documented. In addition, these
data parameters are routinely used in design and
construction and are measured with standard geotechnical
testing procedures.

A total of 11 samples were tested. Prior to testing, 10
of the soil samples were thermally pretreated to
temperatures ranging from 150C (302F) to 800C
(1,472F) for durations of 24, 72, 120, and 240 hours.
One specimen, T-l, was held at laboratory room
temperature (approximately 24C or 75F) prior to
testing and was used to simulate the untreated soil
condition. The test parameters mentioned above were
determined for the untreated expansive soil and its
derivatives from the thermal treatment. Swell-pressure
tests on all of the samples were conducted on a Bishop-
type consolidometer, and all test samples were prepared
from remolded soil.
4.2 Experimental Results
The outcome of experimental work from this study are
shown in the test results matrix, Table 4.1. The
information presented in the table includes the sample
identification number; the applicable test standards
used; thermal pretreatment conditions for each sample;
conditions and results of one-dimensional swell tests;
Atterberg limits; and post-treatment specific gravity of
soil solids. The results of the compaction testing are
presented in a separate section later in this chapter.