2 catty

L

'lli'ifiiii‘

mar»: STAY": we”
:3 Gama

: I...
owe
ML).

fix"

Vi ‘

£33.

r
1"
J

3 ”SHE SHRQ‘C‘E‘iifiifi 31156:} sz'iam
1: I
a“ 93w; Eagtm 5% M. 3.

H164!!! f .
C
gfiga

,t
on
\

I
i

 

THESIS

   
     
   

  

LIB): r 1?
“Wu Sun

This is to oertifg that the
thesis entitled

Investigntion.of Che Structure and Shear
strength Characteristics of o.CI¢y 8011

presented by
Roger David Goughnour

has been accepted towards fulfillment
of the requirements for

M—degree laminating

Ww

Major professor

pm Egg-{27, / ié/

0-169

REE-739:? USE B?!”

INVESTIGATION OF THE STRUCTURE AND SHEAR
STRENGTH CHARACTERISTICS OF.A
CLAY SOIL

by
Roger David Goughnour

AN ABSTRACT

Submitted to the College of Engineering
Michigan State university of Agriculture and
Applied Science in partial fulfillment of
the requirements for the degree of

MASTER OF SCIENCE
Department of Civil and Sanitary Engineering
September 1961

Approved:

 

ROGER DAVID GOUGHNOUR ABSTRACT

This thesis is an investigation of the strength and
structural differences occurring between flocculated and
remolded clays.

The orientation of clay particles of flocculated and
remolded samples were studied by x-ray diffraction. A
method of freeze-drying was used in the preparation of the
samples. The shear strength was studied by means of triaxial
tests.

The x-ray diffraction studies showed a random particle
orientation in the flocculated clays. Clays which were
highly consolidated showed a high degree of particle orien-
tation. Remolded samples showed orientation over small areas,
but the direction of orientation changes from one place to
another.

The triaxial test showed that the cohesion of the
remolded clay acts much as a viscous liquid, while the

cohesion of the flocculated clay acts as a brittle bond,

INVESTIGATION OF THE STRUCTURE AND SHEAR
STRENGTH CHARACTERISTICS OF‘A
CLAY SOIL

by

Roger David Goughnour

A THESIS

Submitted to the College of Engineering
Michigan State university of.Agricu1ture and
Applied Science in partial fulfillment of
the requirements for the degree of

MASTER OF SCIENCE

Department of Civil and Sanitary Engineering

September 1961

ACKNOWLEDGMENTS

The author is happy to acknowledge the help of those
who aided in the conduct of the research and the preparation
of this paper.

Dr. T. H. Wu of the Civil Engineering Department, Michigan
State university, supervised the research and reviewed the'
paper. His guidance and comments are greatly appreciated.

Dr. M. M. Mortland of the Soil Science Department, Michigan
State university, gave valuable guidance in the x-ray studies.
Dr. 0. B. Andersland of the Civil Engineering Department,
Michigan State university assisted in making the final draft
of this paper.

Gratitude is also extended to the National Science
Foundation for their support of this project and to the
Consumers Power Company for providing the funds for the

author's fellowship.

ii

TABLE OF CONTENTS

ACKNOWLEDGMENTS . . . . .
LIST OF FIGURES . . . . .
LIST OF TABLES . . . . .
Chapter

I. DEVELOPMENT OF CURRENT KNOWLEDGE

II. OBJECTIVE OF PRESENT STUDY

III. SOIL STUDIED . . .
IV. X-RAY STUDIES . . .
V. TRIAXIAL TESTS . . .

VI. CONSLUSIONS . . .
APPENDICES--Figures . . .
Tables . . .

BIBLIOGRAPHY .. c .i .. .. .

Page
ii

iv

CDNO‘

1L!-
21
23
37
39

LIST OF FIGURES
Figure Page

1. Schematic Representations of Particle
Orientation . . . . . . . 23

2. Interparticle Forces . . . . . . 2h
3. Changes in Particle Orientation . . . 2n
n. Flat Plate X-Ray Camera . . . . . 25
5. .Arrangement of the Reflecting Crystal . . 26
6. X-Ray Diffraction Patterns . . . . . 27

7. Goniometer Patterns . . . . . . 30
8. Consolidometers and Constant Temperature Box 31
9. Triaxial Cells . . . . . . . 31

10. Shear and Normal Stress, Constant Strain Rate 32

ll. Stress-Strain Relationships, Constant strain
Rate 0 o o o s o o o 33

12.. Friction and Cohesion, Creep Test . . . 3h
13. Friction-Cohesion Time Relationships . . 35

1h. Friction and Cohesion at Failure, Creep Test 36

iv

LIST OF TABLES
Table Page
1. Soil Composition and Characteristics . . 37
2. Summary of Triaxial Tests . . . . 38

I. DEVELOPMENT OF CURRENT KNOWLEDGE

Soil Structure

The importance of the structure of cohesive soils on
the strength properties has been recognized by soil engineers
for more than thirty years. until about 1925, cohesion was
thought to be caused by some hypothetical amorphous substance.
In 1925, Karl Terzaghi discredited this theory with a paper
on the structure and bonds in cohesive soils. He stated that
cohesive soil has a structure similar to a honeycomb, with
large amounts of water being held within the voids and the
cohesion was caused by attractive forces between adjacent
soil particles.

In 1926, Goldschmidt performed a series of experiments
with mixtures of clay minerals and various liquids. He found
that with non-polar liquids, the clay lost its plasticity
and behaved only as a frictional material. He concluded that
the plasticity of clays was dependent on a combination of
both the mineral and the liquid. He proposed that the crx§tal-
line minerals were surrounded by a film of adsorbed water
molecules and that the water molecules stick to each other
and to the minerals because of their dipole moment. He intro-

duced the term "cardhouse" structure to describe the arrange-

ment of the flaky minerals in highly sensitive clays. This
structure is similar to that presented in Figures la and 1b.
The surplus water was considered by Goldschmidt as being
enclosed in the space between the mineral flakes. The less
the sensitivity of the clay, the more dense the arrangement
of the particles.

Observations of the structure of sensitivélclays have
been made with the petrographic microscope (Mitchell, 1956)
and the electron microscope (Rosenqvist, 1959). These obserw

vations tend to substantiate the cardhouse structure.

Forces Acting Between Soil Particles

The forces acting between soil particles may be divided
into attractive and repulsive forces. The attractive forces
include van der Waals' forces, Coulombic forces, water dipole
linkages, and hydrogen and ionic bonds. The repulsive forces
are primarily due to repulsion between like electrical
charges.

Because of isomorphous substitution* occurring within
the mineral lattice. the individual clay crystals carry a
net negative electrical charge. This net negative charge is
balanced by positive ions near the surface of the minerals.
Water may also be attached to this negatively charged surface
due to the dipolar nature of the water molecules. Figures 2a

and 2b show possible conditions under which attractivegforces

 

v—‘fi

*Isomorphous substition occurs when a metallic ion in the
lattice of the perfect clay crystal is replaced in the same
posithon with an ion of similar properties but lower valence.

may exist because of cations and water molecules (Lambe, 1958).

The edge of clay plates are believed to be positively
charged (Lambe, 1958). Thus, a surface to edge relation will
give an attractive force.

The main reason for repulsion of clay particles is due
to the like negative charge on the surface of the particles.
Also, when two particles with cations are brought together,
a repulsion will take place as shown in Figure 2c.

When the attractive forces of'a clay suspension exceed
the repulsive forces, the clay flocculates. The positive
edge attracted to the negative surface results in a random
particle orientation. This is demonstrated by the cardhouse
structure of Lambe shown in Figure 1b..As a result of this,
the flocculated soil has a higher void ratio at a given con-
solidation pressure, than a non-flocculated soil. It is also
possible to obtain a parallel orientation in a flocculated
clay by the addition of an electrolyte. By adding a salt to
the clay-water system, a high concentration of anions will
gather at the edges of the particles. This reduces the posia
tive edge charge as well as the repulsive force and permits

parallel orientation as shown in Figure la.

Effect Of Remolding

 

Nearly all sedimentary clays lose strength when worked
or remolded. Mitchell (1956) has shown the greater the orien-

tation improvement (changing from a random to a parallel

arrangement) with remolding, the greater the loss in strength.
The magnitude of this loss is indicated by "sensitivity" which
is defined as the ratio of the strength of the undisturbed to
the completely remolded soil. When a flocculated soil is remold-
ed, the edge to surface contact is broken and a loss of strength
results. At the same time, the particle orientation is improved

as shown in Figure 1c.

Shear Strength

The shear strength of a soil is influenced to a very
large extent by the attractive and repulsive forces, which
in turn depend upon the particle spacing, particle orientae;
tion and the structure.

The importance of the structure,iaccording to Lambe
(1958) is illustrated in Figures 3a and 3b. They show two
adjacent clay particles with the same average spacing A .
The sum of the contact pressure and the net attractive force
between two particles varys with the power function of the
spacing*. Thus, tilting the two particles from.a parallel
position to the position shown in Figure 3b, results in a
greater increase in the attractive force on the right side
than is lost on the left side. Therefore, the shear stress
required to slide the particles relative to each other is
greater in Figure 3b than in 3a. By the same reasoning, if

the particles are flocculated in a perpendicular position

 

*See Lambe (1953 or 1958) for a discussion of quantitative
interparticle forces.

as shown in Figure So, they pussessa.higher resistance to
displacement than the parallel Particles of Figure 3d. In
short, for a given average particle spacing, the more nearly
parallel the adjacent particles, the lower the shearing resis-
tance.

The shear strength of soils is usually considered to be
made up of cohesion and friction. Cohesion is the shearing
resistance between two particles which exists independently
of external forces. In the flocculant clays shown in Figures
1b and Sc, cohesion exists at the edge to plate contacts.
Lambe (1958) suggested that cohesion is mainly electrostatic
attraction between the negative plate charge and the positive
edge charge.

The friction may be due to both electrical forces and
physical interference between particles during shear displace-
ments. The maximum shear resistance due to friction is a
direct function of the force acting normal to the plane of
shear. A.clean sand is an example of a material with only

the frictional type of strength.

II. OBJECTIVE OF PRESENT STUDY

The primary objective of this study was to evaluate
the strength and structural differences occurring between
flocculated and remolded clays. The orientation of clay
particles of flocculated and remolded samples was studied

by x-ray diffraction. The shear strength was studied by

means of triaxial tests.

III. SOIL STUDIED

The clay used in this investigation is a glacial lake
clay from a site about 15 miles south of Sault Ste. Marie,
Michigan. It is composed of about 60% clay and h0% silt.
X-ray diffraction studies on the clay fraction showed
approximately equal percentages of illite, vermiculite
and chlorite. The soil has a flocculant structure with
a sensitivity of 8 in the natural state. It was found
that the clay could be fIOCCulated in the laboratory, and
that it would retain its flocculent state upon consolida-
tion. Other characteristics of this soil are summarized

in table I.

IV. X-RAY STUDIES

Use Of The Flat Plate Camera

 

.A diagram of the flat plate x-ray camera is shown in
Figure h. Use is made of Braggs law to understand the mech-

anics of the flat plate camera. Braggs law may be stated as:

 

nfl:2dsine or d=

It states that the distance (d) between parallel planes, for
a first order reflection (n=1), is equal to the wave length
(A) divided by two times the sine of the angle 9. Figure 5
illustrates the use of this principle. It can be seen that
the reflection of the x-ray beam depends on a plane of a
crystal being at the angle 9 to the x-ray beam..When a powder
specimen is used, the planes are arranged in an entirely
random manner. There should be enough particles turned at
the angle 9 to the incident primary beam of x-rays to produce
a continuous reflection. The reflection from a beam passing
through a powder specimen forms a series of circular cones.
On a photographic plate it is shown as a series of concentric
rings with uniform intensity. Each ring corresponds to one set
of planes of spacing d.

If instead of a powder, a clay with preferred orientation

in one direction (Figure 1c) is used, reflection occurs only

on the plane parallel to the orientation of the clay particle.
On the photographic plane, the reflected rays are concentrated
into points or short arcs instead of being distributed evenly
along a circle. Thus, by observing the length of the arc of the
reflected beam, it is possible to obtain an estimate of the
particle orientation of the clay.

Copper radiation (A=:l.539 A?) was used to produce the

x-rays in this investigation.

Use Of The Goniometer

 

The goniometer works on much the same principle as the
flat plate camera. The only difference is that instead of a
photographic plate, a counter is used in a preselected plane.
The specimen is stationary while the counter is rotated about
the specimen. A recorder records both the angle 29 and the
intensity of the reflected x-ray beam. If the crystals are
completely parallel, the goniometer will produce sharp peaks
at the angles corresponding to the interplanar distance
spacings..As the degree of orientation decreases, the peaks
will broaden and become less intense. Thus, it is possible
to obtain an idea of the orientation of the clay along the
plane selected.

The x-ray beam of the goniometer will penetrate through
only a few particle layers of the specimen. This makes it
necessary to obtain a sample with a surface which is both smooth
and undisturbed. This condition is difficult to obtain and
it limits the use of the goniometer for a quantitative analysis

of particle orientation.

10

Preparation Of Specimens

The flat plate camera requires very thin specimens and
an exposure of several hours. If a clay is left in a moist
condition, it will dry out during the x-ray exposure. It is
alsoznnpossible to obtain a thin specimen from a moist sample.
A.method of freeze drying was used to eliminate these diffia
culties. Specimens of approximately 2 square centimeters were
immersed in liquid air, which results in rapid freezing of the
clay. The specimens were then placed in a vacuum desicator
under a vacuum. The desicator was kept at temperatures well
below freezing until the specimens were dry. Rosenqvist (1959)
found that the freezing in liquid air involves a very slight
volume increase (3% for saturated samples), whereas the drying
took place without any noticeable change in dimensions. For
this reason, the mineral arrangement is assumed to be essen-
tially the same before and after freeze drying. The dried
samples were broken mechanically and thin specimens were obtained

for analysis.

Results With The Flat Plate Camera
The soil used is composed of illite, chlorite, vermiculite
and quartz. The chlorite and vermiculite have first order
lattice spacings of lu.A? and second order spacings of 7.A°.
The illite has a first order spacing of IO.A°. The quartz has
a spacing of 3.3 A?. These are the most significant spacings.
Figure 6a is a powder pattern of the soil. The 1h, 10

and 7 A?spacings are present though not too intense. It should

11

be noted that each ring is uniformly intense throughout.

Figure 6b shows the pattern from a specimen obtained
by placing a suspension of the soil in an evaporating dish
and drying it out to a thin film. The particles in this thin
film were very well oriented and were used as a base<>f com-
parisonvuitiother samples. The 1h, 10 and 7.A° spaCings appear
as arcs which are parallel to the particle orientation. The
arcs have much greater intensity than the rings of the powder
pattern. The 3.3.A? quartz spacing still appears as a complete
ring. This is because quartz is made up of irregular grains
which do not orient. In Figure 1, the quartz is represented by
the large, non-plate shaped particles.

Figures 60 and 6d represent patterns from a soil that was
flocculated and then consolidated to a pressure of 0.35 kg/cm‘.
These two figures represent the two extremes of orientation
that can be obtained in this soil. Figure 6c is much.the same
as that obtained for the powder pattern. This would indicate
that there is practically no orientation. Figure 6d represents
some orientation, but net nearly as much as given in Figure 6b.
It is probable that as pressure is applied to flocculated clay,
there is local breakdown in the cardhouse structure. Thus,
Figure 6c shows the clay still retaining the flocculated struc-
ture, while Figure 6d shows some breakdown of the structure.

Figure 6e was obtained from a soil consolidated to the
high pressure of no kg/cmZ. This pattern is similar to that
given in Figure 6b. It was also possible to find areas of
slightly less orientation. This would tend to support the

assumption illustrated in Figure 1c, in which an overall

12

orientation exists with local areas of slightly less orienta-
tion.

X—ray diffraction studies were also run on a remolded
sample which was hydrostatically consolidated to a pressure
of 2.25 kg/cm”. Some orientation could be found in local areas
of this sample. The orientation shifted from one direction to
another, with no one plane of preferred orientaion. Thus, as
the sample was remolded and hydrostatically consolidated, the
the clay particles were pushed closer together and local areas

of orientaion developed.

Results With The Goniometer

 

Figure 7 represents the x-ray diffraction pattern obtained
from the goniometer. The sample for Figure 7a was obtained as
follows. A suspension of the clay was placed on a porous plate.
A.vacuum was applied to the plate and the excess water was
drawn off, leaving a film of highly oriented clay on the plate.
The diffraction pattern shows distinct peaks at 1h, 10, 7
and 3.3 A? spacings.

Figure 7b is the pattern obtained on the horizontal
plane of a soil consolidated to a pressure of no kg/cmf.

The distinct peaks of Figure 7a are not present, but instead
there is a continuous intensity from about 15.A° to 6.A°.
Thus, orientation is present, but not to the degree given in
Figure 7a. The quartz peak of 3.3.A°is still present.

Figure 7c is the pattern obtained on the horizontal
plane of a flocculated clay. There are no distinct peaks at

10 and 1k A° spacings which indicates that very little orien-

13

tation is present.

The difference between the goniometer pattern of the
highly oriented and the non-oriented samples is obvious.
However, the fact that the diffraction surfaces are not
perfectly smooth and undisturbed makes more eXact.work

impossible.

V. TRLAXLAL TESTS

The triaxial tests were made on remolded and laboratory
flocculated soil samples. Three consolidometers were specially
constructed to prepare the flocculated samples. The COHSOlidOP
meters are made of two lucite tubes mounted on a six-inch
diameter brass ring. The total height is 18 inches..A constant
temperature box was built around the consolidometers to main-
tain a temperature of 26’C.i2° during sedimentation and consoli-
dation.

About six inches of a dilute suspension of the soil
(w=500% to 800%) was placed in the consolidometers and allowed
to flocculate and settle. After a week, the clear liquid above
the flocculated soil was drawn off and another six inches of
the dilute suspension was added. Care was taken in adding the
dilute suspension not to disturb the flocculated soil. This
process was continued until the flocculated soil almost filled
the consolidometers..A porous stone was then carefully lowered
to the top of the clay and consolidation loading was begun.

The first load increment was 0.002 kg/cmF. This load was doubled
each week (90%+ consolidation) until a final load of 0.35 kg/cmz
was reached. The load was then removed and the consolidated
soil allowed to rebound for three weeks. The soil cakes were

then extruded from the brass rings and six 1.5 inch diameter

in

15

by 3.0 inch triaxial specimens were prepared from each cake
by hand trimming.

The soil cakes, when removed from the consolidometers,
had a moisture content of about 60.0%. This is above the
liquid limit of 55% and soil trimming was done very carefully
so as not to disturb the sensitive structure which developes
in the soil during sedimentation and consolidation. When this
flocculantsotiil structure was disturbed, the remolded clay
behaved like a viscous liquid.

Triaxial tests<n1the remolded samples were performed
at Michigan State university by Mr..A. G. Douglas*. The results
’ of these tests are included in the discussion of the results.

To obtain the remolded specimens, a batch of clay was
prepared with an initial water content of approximately 38%
and stored in the moist room. Specimens were molded from this

material when needed.

Shear Strength Parameters

 

The shear stress7§ at failure on any plane may be repre-
sented as (Terzaghi 1936 and flyorslev 1937):
7}: Car" ff’“); “Miler
in which.c is the normal stress on the plane and u is the
pore water pressure. The subscript f denotes the values
obtained at failure of the specimen. The parameters C¢.and

0e are known as Hvorslev's parameters and represent the

 

*Graduate.Assistant, Michigan State university, (on leave
from New South Wales, Australia

16

"true cohesion" and the "true angle of friction" respectively.
The true cohesion is dependent only on the void ratio (moisture
content) at failure in saturated soils.

The measurement of Hvorselv's parameters by triaxial tests
requires the results from at least two specimens with equal
void ratios but different effective stresses at failure. This
was achieved using the procedure outlined by Schmertman (1960).
Three specimens were hydrostatically consolidated under a given
pressure,6} . This pressure was then reduced on two of the
specimens, one by 0.25 and the other by 0.50 kg/cmz, resulting
in one normally consolidated and two slightly overconsolidated
specimens. Since a small reduction of pressure causes neglig-
ible swelling, the three specimens have almost identical
void ratios. By using the same rate of strain on all specimens,
and measuring the pore pressure during the tests, it was
possible to determine the cohesion and friction at any strain
during the tests. Schmertman's method was used on all of the
triaxial tests of this report. Figure 9 illustrates the

triaxial equipment used in these tests.

Constant Strain Rate Tests

 

The three specimens used for this test were first con-
solidated in the triaxial cells to a pressure of 2.00 kg/cmz.
.After complete consolidation, the pressure was reduced on
two of the samples, one by 0.25 kg/cmz and the other by
0.50 kg/cm2.Apressure of 1.00 kg/cm‘2 was added to both
the chamber pressure and the pore pressure. This added

pressure does not change the effective stress condition,

17

but dissolves most of the air bubbles that may be present.
and gives more accurate readings. After 2h hours, the constant
strain rate was started. The pore pressure and the applied
load were recorded continuously during the tests. The strain
rate was h% per hour computed for the initial specimen length
of three inches. The deviator stress was computed by the method
recommended by Lambe (1951). The decrease in area due to the
consolidation was small and was not computed.

Friction and cohesion were obtained as shown in Figure 10.
The results on the remolded clay are shown in Figure 10a and
the flocculated clay in Figure 10b. For clearness, data from
only two specimens are plotted. Figure 11 shows the deviator
stress and the pore pressure as a function of the strain. Also
shown in the Figure are the friction and the cohesion as a

function of the strain.

Creep Tests

 

The specimens for the creep tests were prepared for
testing in the same manner as the specimens for the constant-
rate-of—strain tests. In the creep tests, each deviator stress
increment is imposed on the specimen and maintained constant
without drainage until the pore pressure comes to equilibrum.
Then another increment is added to the deviator stress. This
process is continued until the sample fails. In order to compute
friction and cohesion at any time during the test, it is
necessary to obtain almost identical strain-time relationships

To achieve this, the deviator stress was varied on two of the

18

samples by 0.25 and 0.50 kg/omz to correspond to the differ-
ences in chamber pressures.

Figure 12 shows the determination of Hvorslevls para-
meters at different times for one load increment. Figure 12a
represents the remolded clay and Figure 12b represents the
flocculated clay. Figure 13 shows the change in friction and
cohesion with time for both the remolded and flocculated
clays. Figure l3.is typical for all loading increments except
near the failure of the flocculated clay, On failure of the

flocculated clay, the cohesion drops to zero while the angle‘

of friction approaches 23.5°(Figure lu).

Discussion Of Results

 

Figures 10 and 11 illustrate the two different types of
failure occurring in flocculated and remolded clays. Figure 11
shows that for the remolded clay, the deviator stress builds
up quickly at first and continues to increase at a constant
rate:at high strains. The pore pressure rises to a certain
value and then remains almost constant. For the flocculated
clays, the deviator stress builds up to a maximum value at
low strains (2%) and then remains constant. The pore pressure
continues to increase and levels off at about 10% strain.
These results are shown in Figure 10a.where the stress state
of the remolded clay reaches the failure envelope quickly
as the deviator stress increases. In Figure 10b, the stress
state of the flocculated clay builds up until the deviator

stress reaches a maximum value. The average normal stress

19

then drops off with increasing pore pressure and reaches
the failure envelope as the pore pressure becomes constant.

These results can be explained by considering the struc-

ture of the clay. The cardhouse structure of a flocculant
clay makes it act much as a brittle material. The strength
builds up rapidly, until the structure is broken. During
this build up, the strength is composed only of cohesion.
.As the structure breaks down, the cohesion goes to zero
and the specimen act's much as a frictional material. This
would indicate that the cohesion is a result of the same
interparticle forces which cause fiocCulation.

The fact that there is a high initial cohesion in the
remolded clay would indicate that interparticle forces are
also effective in these specimens..After the initial strength
which is due to cohesion, the added strength is a result of
friction.

It is noted that at large strains, cohesion is present
in the remolded clays, while not measureable in the flocculated
clays. This may be due to the difference in the final moisture
contents.(about 39% for the flocculated clay and 28% for the
remolded clay)..A consideration of the soil water system
predicts that the interparticle forces have a diminishing
effect with increasing moisture. As the moisture content
increases, the distance between particles also increases,
which reduces the net attractive forces.

The creep tests gave results similar to the constant-

20

rate-of—strain tests. In Figures 12 and 13 it can be seen
that most of the strength of the flocculated clay is due to
cohesion. Most of this cohesion is still present after 24
hours, while there is little increase in friction. The
flocculated clay fails at low strains (2%). As failure
takes place and the cardhouse structure breaks down, the
cohesion goes to zero while the angle of friction builds

up u> about 23.5°. In the remolded samples, the cohesion
goes to zero within 2h hours after each load increment
while the friction builds up to resist the load. The cohesion
of the remolded clay may be compared to a viscous liquid.
There is a strong resistance to each load increment, but

this resistance goes to zero with time.

VI. CONCLUSIONS

X-Ray Studies

 

The flocculated clays, when consolidated to low pressures,
showed no orientation to slight orientation. This was attributed
to local break down of the fltructure as load was applied.

It was also shown that a clay soil attains a high degree of
orientation when it is consolidated to high pressures. The
remolded samples showed areas of local orientation with a
shifting from one plane to another. The results of this study
are in agreement with the schematic representation of Lambe

shown in Figure 1.

Triaxial Tests

 

Both the constant-rate-of-strain tests and the creep
tests yield results which are in agreement with the cardhouse
structure theory.

In the remolded clay, the cohesion acts much as a
viscous liquid. There is an initial high cohesion which
becomes almost constant for a constant rate of strain and
drops to zero for a constant load. The friction starts to
build up immediately.

The flocculated clay, with the cardhouse structure,

acts as a brittle material, There is a high strength at

21

22

low strains due to cohesion. As the structure is broken
with increased load, the cohesion drops off and the clay
takes on the strength properties of the remolded clay.

There is very little friction until this structure is

broken.

23

 

 

 

a. Salt Flocculation

 

 

 

 

 

 

b. Non-Salt Flocculation

 

 

 

 

[Jr
flrg

f—Ej f—gl
[2:221:31 -.-—'_
:22” I:
:::::n =3
[5:217
12:31 /i3
1:310:22: I:::L
l"-"":'I

c. Consolidated

FIGURE 1 (after Lambe 1953, 1958)
SCHEMATIC REPRESENTATIONS OF PARTICLE ORIENTATION

 

 

 

 

~.'~' “0

”UM.

 

 

 

_ — — — - " " '-'-"""4:""=‘3
“W 3:3 GEE-D
—""‘ CZZD<IZEQ

E~ +-—- +-
a. Cation Linkage b. Water Dipole Linkage

 

 

--.:-.-:1

E

c. Cation Repulsion

I®®
l®@

 

FIGURE 2 INTERPARTICLE FORCES

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

[—w _J i r 43
A
a . b .
4 ' j
in
f l~~~~ _.\I_.__---..-_--« I 7

 

 

c. d.

(after Lambe, 1958)
FIGURE 3 CHANGES IN PARTICLE ORIENTATIONS

 

 

25

 

 

$525 balsa mafia 53o : ammo:

Cosfiooam

upmfim ofinowcmopozm

1’

 

 

 

Emmm Ammux cam

woflozcfim was;

xom umoa

 

 

 

 

 

 

 

 

 

m—Imm

 

 

 

 

 

mash AwmnX:

 

 

 

26

 

 

.‘IO ~III1'. ‘1‘. ,. ...l‘|‘

nil... till... I ..r,:... .III'. '1‘.

 

4<Hw>mo @ZHHUmammm mmB m0 Hzmzmoz<mm¢. m mmDon

“Gm amp

I)...

Edam kmmux cwmz..

nt.\.opmfim vasomomODan
Goefiooaw web
to mamas mausosscmm

  

 

¢WMW\_\\\\\\ .nwmw

\\\\\
\\

 

 

.m\

Seem zmmax ompoofimo

 

 

 

0"L—jr

”vuw'.-'_— -vlbfl-O~-. _,‘.”

 

 

27

 

 

 

a. Powder Pattern

 

-—-——. -- -__- .-._...

 

i

L, J a
b. Oriented Film 2

.I

FIGURE 6 X-RAY DIFFRACTION PATTERNS i

 

 

28

 

 

 

c. Flocculated (no orientation)

 

d. Flocculated (some orientation)

FIGURE 6 (CONTINUED)

 

__ --v‘~

 

29

 

 

'—--*-~->~--a “»-~.~ .«Or 7.-

e. Consolidated (no kg/cmz)

FIGURE 6 (CONTINUED)

..-...~... u. -fi...—....._..

 

5.". “do. «av—nu..-” .4”. “- m--— ~

-u... ---‘~"~ ‘-

 

30

 

 

tfiQMWUn594GEXMWWAS

noonoeoe/asazwnes

(00 N 75/? Rem/Ive

 

 

 

 

 

 

 

 

 

 

 

 

 

 

80- s a. Vacuum
I Oriented
60-- 1
40.
a0- k\\fl//\~vfl’_afl.D/A.J
JL 1 . I J
032 I2!“ Ix'a I Ig‘r 629
3.3 4.0 5.0 7.0 /0 M MUN/73‘
80a
b. Consolidated

604
404

20-

0 J I l I l

32 24 /6 0

26(D£6R££5)

6b— c. Flocculated

60-
4o-

0 I l 1 l

32 24 /6 8 O 26

 

FIGURE 7 GONIOMETER PATTERNS

 

 

.31

 

 

 

 

FIGURE 8 CONSOLIDOMETERS AND CONSTANT
TEMPERATURE BOX

 

 

 

 

FIGURE 9 TRIAXIAL CELLS

 

 

 

 

a..-“

   
 

 

 

   

 

 

=225°E
€' (”freeze j
/.0' - a?
.SP£(.2 .5P[(.3 {050(g?:é§ .
0? :Z.00 0C:2.00 (3" :5
f‘.’ :- .V? .1- -. '
S 0.} / {a 02 P 00 I, E: 0/0(_¢c"8
$.54 Ill/{328519 01:28.2 ”vu/ ((e' a;
a: ' (w.
”I” .m. —-- '€=.003
”b ____._... d =45
lb) C’
|‘\ 5! )CI'.(:"‘A 5f£.(l'.._?“\fl [(36328
b
\IN | 1 T l j
0 '5 L5) 55‘ .20 227
%(Qi*d:)/%%%az
a. Remolded Clay
JPEC . 305 5.0a. 306 ;_
[OT 5‘ :2'00 0?: 2.00 5
N 0:? =fl30 0:3;2.00 €=‘/mfc 3::(33’005
\§ W; = 38.5% w;= 39,07; 8 1
°\ - ¢b=l3° I
k /€-'O3O (a: /3 ‘
05- I
/ . ab.» 3.00 1
I5 ,1 €-.O/O[Ce:.35 ‘
I 6‘ 5” C - 3 05 jflff. 305
\ __..
1 [ | ‘ _'
V” O 5 /50 A5' 2&0 255

was) Wm.
b. Flocculated Clay

FIGURE 10
SHEAR AND NORMAL STRESSES, CONSTANT STRAIN RATE

 

 

 

 

33

 

 

mag ZHgHw HZflwaOU .meIwZOHHdJ—mm ZthwlmmmMHm : mmDOHm

Q. on. too.

Ir b h

sane
essenseooHe .e

N Tatum... we. we
QO .anxm. .

fire:

\0. NQ. Q
_ _

 

 

.
.

 

         
 

 

Imfi

IQV

 

RES» VSXQ
Q. mo. No. As. me. n

P

 

Q

Q

 

 

 

)
0
X a w
.J. /
ab \ r \f.Q\J
W a
uQN.yuQN.
I
stow/Sm
o\. moo. SO. no. No. O
p P _ _ . b
has. MNNQé .8
\ Stash“, _
_ a
-ms W
S30 M
emEoEom .m 3
sh \ \ x -3 a.
.. .. l w

 

 

 

 

3’4

 

 

 

 

 

 

 

 

 

b. Flocculated Clay

FIGURE 12 FRICTION AND COHESION, CREEP TEST

 

- /) fl =/70
/-0 21/ [(91 —,02
~ fl =/2
kg 266cc?“
¢¢=85o
* 2m[C9-0.//
0.5-
7.3
35 ,
\lrv
/ T l I
0 0.5 no /.'5 2.r0 2'5'
TYGSHfCa) ’99443
a. Remolded Clay
2 h[¢c:2 9.700

/.0- was: 3/5 e We. 3/3 5 "

NE 02:32; :22; 27.330;
R 03 :2.25- 1.2.00 [(3:: 2.0
it w =38.4 W= 38.6 33

0.51 ‘
1?’ W

'L‘

‘tv

0 0.5 fa /.'5 2'.0 2.T5

 

 

35

 

 

 

 

 

 

 

 

930-
t
ixzo\ .
l
s
\
h
hWm f
8' \ 1
\ o i
\\ \\
\\p ‘
O\ \ \\
0 3'“ l I ¥ l r
0 Ad .20 (1 A9 .20
f/NE -//&o/P.S
L)
5%»
a ——————— ~
§fi / /
Q a”
‘\ lo. I
\J J
% .- _ __ .. ... - o
x ,o/ ’ ’
,5
/ f v
y fi Kr...“
VI,
00 l l I | I
’0 20 o M 20
77M£ - flail/P5
- - - RAWOMID
--f200fllAfZD

FIGURE 13 FRICTION-COHESION TIME RELATIONSHIP

 

 

 

36

 

 

 
 
 

’\ /.0_
E
K o>=23£0
s C: 0
’\ Pé‘t. 3/5'
lDos- 5
l ‘vSREC 3A3
'\
S

 

 

l I l l

C) 0h5' AC? A57 .249
free) We:

 

FIGURE 1h FRICTION AND COHESION AT FAILURE, CREEP TEST

 

-—......~-——

 

 

TABLE 1

Composition

Clay Fractionn

Silt -

Sensitivity

Liquid Limit

Plastic Limit

SOIL

Cation Exchange Capacity

COMPOSITION

AND CHARACTERISTICS

(Milliequivalents/100 grams - -

Specific Surface (Square meters/gram) - - - - - - -

Total Potassium Content

 

60%
u0%

55%

23%

m
100
2.07%

nApproximately equal percentages of illite, vermiculite and

chlorite.

’1'?

 

TABLE 11 SUMMARY OF TRIAXIAL TESTS
Spec. Initial Final
N0. w w 02 0:3 of up Type of Test
30h 59.6% 38.5% 2.00 1.75 0.90 3.23 Const. Rate of
305 59.5% 38.5% 2.00 1.30 0.8u 3.00 Strain
306 58.5% 39.0% 2.00 2.00 1.35 3.50
307 59.3% 38.1% 2.50 2.00 1.35 2.10 Creep
309 60.0% 35.3% 2.25 1.75 1.05 1.56
310 59.6% 30.1% 7.00 6.00 2.87 b.85 Const. Rate of
311 60.8% 30.3% 7.00 6350 3.05 5.50 Strain
312 58.2% 30.4% 7.00 7.00 2.95 6.15
313 59.h% 38.3% 2.25 2,00 0.90 2.20 Creep
31E 58.7% 38.h% 2.25 1.75 0.85 2.20
315 59.8% 38.u% 2.25 2.25 1.15 2.50

38

l.

2.

9.

IO.

11.

G.

R.

J.

T.

T.

J.

J.

K.

BIBLIOGRAPHY

L. Clark, (19h0) Applied X-Rays, New York and London:
International Series in Physics, McGraw-Hill Company,
Inc., P. 81ff.

E. Grim, (1953) Cla Mineralo McGraw-Hill Series in
Geology, New‘YorR: McGraw-HIII Book Co., Inc.
M. Hvorslev, (1937) "Uber die Festigkeitseigenschaften

Gestorter Bindinger Boden". Ingeniorvidenskabelige
Skrifter, No. A115, P.159.

M. Lambe, (1960) "A.Mechanistic Picture of Shear Strength
in Clays", ASCE Research Conference on the Shear
Strength of Cohesive Soils, Boulder, Colorado.

 

M. Lambe, (1958) "The Engineering Behavior of Compacted
Clay", Journal of the SoilMechanics and Foundations
Division, Proc. Amer. Soc. Civil Engineers, V01. 8h,
No. 8M2.

M. Lambe, (1953) "The Structure of Inorganic Soils",
Proc. Amer. Soc. Civil Engineers, V79, No. 315.

M. Lambe (1951) Soil Testing for Engineersl New York:
John Wiley and Sons, Inc.,

 

K. Mitchell, (1956) "The Fabric of Natural Clays and
Its Relation to Engineering Pr0perties", Proc. Hwy. Res.
Board, Vol. 35,

Th. Rosenqvist, (1959) "Physico—Chemical Properties
of Soils: Soil-Hater Systems, Proc. Amer. Sci. Civil
Engineers, Vol.85, 8M2.

H. Schmertman and J. O. Osterberg, (1960) "An Experi-
mental Study of the Developement of Cohesion and
Friction with.Axia1 Strain in saturated Cohesive
Soils", Proc. Conf. on Shear Strength of Cohesive
Soils, Amer. Soc. Civil Engineers.

Terzaghi, (1936) "The Shearing Resistance of Saturated

Soils and the Angle between the Planes of Shear?,
Intern. Conf. on Soil Mechanecs, Vol. 1, p5h.

39

Price 1 u.

Ml ICHIGAIN STATE UNIVERSITY LIBRARIEI S

 

1 11111 1111 11111111111