\%ST. OF TECHIVOo
8 1963
FEB
LIBR A R'
THE EFFECTS OF SALT ON
THE SHEAR STRENGTH OF
BOSTON BLUE CLAY
by
William A. Bailey
Submitted in Partial Fulfillment
of the Requirements for the
Degree of Bachelor of Science
at the
MASSACHUSETTS INSTITUTE OF TECHNOLOGY
June, 1961
Signature of Author:
Departme
Certified by:
. . . .....
0 .r
of Civil Engineering, Ju
.,
..* 0 ., 0
.,
.. * . . .
...
,g*-.
V
1961
*
Thesis Supervisor
Accepted by:
....
..
..........
.
.
Head of the Department
..
...............
00 00
ABSTRACT
EFFECTS OF SALT ON THE SHEAR STRENGTH
OF BOSTON BLUE CLAY
Isotropically consolidated, undrained triaxial tests with
pore pressure measurements were run on three types of Boston Blue Clay
(MIT 1139): 1) samples consolidated from a slurry prepared with "water"
(less than 3 gm NaCl/liter) pore fluid; 2) samples consolidated from a
slurry prepared with "salt" water (35 gm Nal/liter) pore fluid; and
3) four "leached" samples of salty clay, consolidated to 3.0 kg/cm2 a
leached of the majority of their salt prior to shear.
The data showed:
1. Water samples had a lower Wf/ ec ratio but the same
S / rc ratio as the salt samples.
2.
Water samples have a slightly lower (20) effective
stress envelope at maximum principal stress difference
than salt samples for normally consolidated specimens.
Water samples have a slightly lower effective stress
envelope at maximum principal effective stress ratio
than salt samples for normally consolidated specimens.
4. Leaching the salt from the normally consolidated salt
samples at constant rc produced the following effects:
c
a) decreased the shear strength and the strain at
failure (maximum principal stress difference),
3.
b) increased Au and Af at maximum obliquity; at maximum principal stress difference Au is virtually unaffected and A changes only very slightly due to a
combination of reduced strain at failure and reduced
shear strength,
c) reduced the effective stress envelope at (0" - Q )
(*) from 26.50 (salt samples) to as low as 190 (L-4).
-ii-
The data indicate that the lower strength at the larger strains
near maximum obliquity of the leached samples is due to higher pore pressures generated in the sample by remolding in the failure zone. These
higher pore pressures are caused by increased particle reorientation
which could be explained by the relative inability of particles to remake
broken contacts during shear because of the increased double-layer
repulsion in samples with lowered salt concentration and high water
content at a given consolidation pressure.
-iii-
ACKINOWLEDGEMENTS
The author wishes to express his gratitude to Mr. Richard
Harkness for hours of conversation: critical, informative, and helpful;
and to his thesis supervisor, Professor Ladd, for advice, commentary,
and unending patience.
-iv-
TABLE OF COTIENTS
Page
TITLE PAGE
ABSTRACT
ACKNOWLEDGEMENTS
TABLE OF CONTENTS
SYMBOLS and NOMENCLATURE
LIST of FIGURES and TABLES
CHAPTER
I:
INTRODUCTIOI
Section 1.1
1.2
1.3
1.4
CHAPTER
II:
Section 2.1
2.2
2.3
2.4
2.5
2.6
CHAPTER
III:
Section 3.1
3.2
3.3
3.4
3.5
3.6
3.7
A General State ment
Expected Effec'ts of Salt Content on
Shear SLrcngrth
A Review of Somc Previous Investigations
Scope of the Present Investigation
PROCEDURES
Summary
Sample Preparation
Triaxial Testing
Arrangements for Leaching out Salt
Water Contents After Failure
Void patios After Oven-Drying
TEST RESULTS
Classification Tests
Stress-Strain Curves
Shear Strenrgth
Effective Stre;;s Parameters
Hvorslev Parameters
Water Contents After Failure
Void Ratios After Oven-Drying
vii
CHAPTER
IV:
DISCUSSIONS OF RESULTS
Section 4.1
4.2
4.3
4.4
4.5
4.6
4.7
4.8
CHAPTER
V:
BIBIZOGRAPHY
Effects of Salt on Atterberg Limits
Effects of Salt on Stress-Strain Curves
Effects of Salt on Shear Strength
Effects of Salt on Effective Stress
Parameters
Effects of Leaching Out Salt After
Consolidation
Effects of Aging
Use of Hvorslev Parameters to Adjust
Undrained Shear Strengths
Distribution of Pore Water After Failure
and Void Ratios After Oven-Drying
COCUSIONS
64
64
64
65
66
67
70
72
74
80
83
SYMABOIB and NOMCATURE
=
obliquity of the Mohr-envelope at maximum principal stress
difference on the plane of maximum obliquity.
= obliquity of the Mohr-envelope at maximum principal effective
stress ratio on the plane of maximum obliquity.
=
obliquity of the Mohr-envelope at maximum principal stress
difference on the plane of maximum shear stress.
=
obliquity of the Mohr-envelope at maximum principal effective
stress ratio on the plane of maximum shear stress.
=
axial strain ='change in length/ initial length
w
=
water content = weight of pore water/weight of soil grains.
vI
=
volume of pore fluid/weight of soil grains.
v
=
liquid limit.
£
wp
plastic limit.
P.I.
plasticity index.
CIU
OCR
NC
triaxial test: consolidated isotropically, sheared undrained with
pore pressure measurements.
=over-consolidation ratio = maximum past consolidation pressure/
consolidation pressure just prior to shear.
=
=
C
normally consolidated
over-consolidated.
=
normal effective stress.
shear stress.
q
=
p
(
3)*
(3).
S
u
T c=
=
undrained shear strength = .1 ( '
2
1
consolidation pressure.
vii
-)3 max
SUBSCRIPTS:
S
=
salt samples.
W
=
water samples.
L
=
leached samples.
f
=
"at failure" or "on the failure plane".
ff
"on the failure plane at failure".
viii
FIGURES and TABLES
Figure
Number
1.1
Title
Data on Asrum Clay (fr:
Pahge
BJERRUM and ROSENQVIST
(1956))
3.1
13
Grain Size Distribution (MIT 1139)
Fig. 3.2 thru 3.26 plot:
pore pressure vs. C; ( d
("
26
- a3) v3.e;excess
) vs.E. ; A vs.4 for
the tests listed below:
3.2
S-1 ( ac = 2.0)
27
3.3
S-4 ( c = 3.0)
28
3.4
3-9 (&c = 3.0)
29
3.5
S-2 (&
= 4.0)
30
3.6
S-3 ( &c = 6.0)
31
3.7
s-6 (4 -c = 0.5; oCn-6)
32
3.8
s-8 (
33
3.9
3-5 ('e
3.10
W-4 (Cc = 2.0)
35
3.11
W-2 (* c = 3.0)
36
3.12
W-8 (i-e = 3.0)
37
3.13
w-6 (4&c
g-c = 1.0; OcR:12)
= 1.0; OCR=12)
= 3.0; consolidated, rebounded to zero,
and reconsolidated)
= h-o)
3.14
W-5
3.15
W-9
3.16
W-3
3.17
W-1(&c = 6.0)
( c
( c
&c
34
-
4.03)
6.0)
38
39
11o
Figure
Number
Title
Page
W-10 (e
=
1.0; cR=6)
43
3.19
W-11 (&
=
0.5; oc.n=12)
44
3.20
W-12 (
0.25;
45
3.21
L-1 (
3.22
c
6C
C=24)
3.0)
46
L-2 (
= 3-0)
47
3.23
L-3 ( c
=
3.0)
48
3.24
L-4 (
= 3-0)
49
3.25
S-10 (e
3.26
W-7 (
3.27
S
U
=
= 3.0; aged 46 days)
50
3.0; aged 46 days)
51
vs. w';
c vs. w' (all normally consolidated
C
samples)
52
3.28
g
vs. p;
Su V .
(salt samples)
53
3.29
qma.
vs. p;
Su vs.
c (water samples)
54
3.30
Effective Stress Paths for N.C. Salt Samples
3.31
Effective Stress Pt'h3 for HT.C.
3.32
Effective Stress Pathi: Comparison of Leached
Samples with Salt and Water Samples at the Same
Consolidation Presssure
57
3.33
Effective Stress Patahs for O.C.
3.34
Effective Stress Paths for 0.C. Water Samples
59
3.35
Hvorslev Parameters for Salt Samples
60
3.36
Hvorslev Parameters for Water Samples
61
3.37
Pore Water Distribution after Failure
62
3.38
w' after Failure and Oven-Drying
63
s
Water Samples
Salt Samples
55
56
58
Figure
Number
4.1
Tile
Page
Leached Samples: Effective Stress Envelope at
Maximum Principal Stress Difference
4.2
Effective Stress laths for Aged and Non-Aged
Salt Samples
4.3
Effective Stress Paths for Aged and Non-Aged
Water Samples
5.1
1.2
2.1
79
Comparison of a Leached Sample (L-3) With a
Salt Sample (S-10)
Table
Number
1.1
77
82
Title
Pare
Effects of Salt Leaching on the Shear Strength
of Horten Clay (fr: SKEMPTON and NORTHEY
(1952))
6
Properties of Asrum Clay (fr:
ROSENQVIST (1956))
8
BJERRUM and
Age, Leaching Time, and Salt Content at Failure
of Leached Samples
17
3.1
Properties of MIT 1139 (Boston Blue Clay)
19
3.2
Sunmiary of Triaxial Test Data (Salt and Leached)
24
3.3
Summary of Triaxial Test Data (Water)
25
3.4
Maximum Obliquities for Salt, Water, and Leached
Samples
21
4.1
Effects of Salt on Atterberg Limits (fr:
(1956))
GREEN
64
4.2
Effects of Aging on Water Content
68
4-3
Effects of Water Content on
73
4.4
Su of Salt Samples Adjusted for Water Content
73
4.5
'Tf
74
of Leached Samples Computed From Hvorslev
Parameters
xi
OF BOSTON BLUE CIAY
EFFECTS OF SALT ON THE CTFAR STRENGT
CHAPTER
I:
TITROIUCTION
A General Statement
1.1
In recent years it has become generally recognized that the
engineering properties of clay-water systems can be changed appreciably
by altering the environment contributed to the system by the pore
fluid.
These environmental changes do not necessarily need to be
drastic,'but rather are often within the variations that might typically
be encountered in the field.
The engineering behavior of clay-water systems is determined
by the structure of the clay; that is,
by the arrangement of particles,
or fabric, and the forces that act between particles.
LAMBE (1958)
contends that the structure of a clay is determined by the electrical
forces associated with the charged clay platelets.
Changes in the
characteristics of the pore fluid that affect these electrical interparticle forces will change the structure of the clay and, consequently,
changes in the engineering behavior of the soil should be expected.
1.2
Expected Effects of Salt Concentration on Shear Strength:
LAMBE (1960) has offered an equation for force equilibrium
across a unit area in a saturated soil-water system:
W =
-am + ua w +(N-A)
(Eq.
1.1)
The terms in this equation are defined as follows:
G" =the average total stress across the unit area,
u
=
the average stress where there is "mineral-mineral"
contact,
=
the pressure in the bulk pore fluid,
a m
the fraction of the unit area where there is
"mineral-mineral" contact,
a = the fraction of the unit area where there is
w water-mineral contact or water-water contact,
(i-I) =
net electrical effect (repulsive when +) acting
across the unit area.
The "effective stress", 0', is deduced from this equation by taking
a
V
1.0:
+ (-X)
u=
'-u=Ya
m
Thus, there are two components of effective stress:
and 2)
forces.
(Eq. 1.2)
1) contact stresses,
stresses due to an electrical contribution to interparticle
LADD (1961) has separated
am from (:-X) on the basis of
m where interparticle distance is
proximity of the particles; (ram acts
less than 200 A. Thus, there are "short-range" (contact) stresses, Tam'
and "long-range" stresses, (R-X).
If the effective stress, d,
is maintained constant, the contact
stresses, 4a , must be reduced if the electrical component, (R-X), is
increased and vice versa.
The effective stress,
da
, has proved to be
auseful shear strength parameter for saturated soil-water systems.
It
is not an ideal parameter since one would expect that the relationship
between shear strength and effective stress would change if the distribution of a given amount of effective stress between contact stresses
and long-range stresses was changed; that is:
the "coefficient of
friction" associated with contact stresses would not be expected to
have the same value as the "coefficient of friction" associated with
long-range stresses.
Soils in which the (T-X) component of effective
stress is thought to be comparatively large, such as a sodium montmorillonite (see LADD (1961)), show friction angles in terms of effective
stresses that are less than friction angles in other soil-water systems
where the (s-X) component is believed to be less important.
Therefore,
a transfer of effective stress from contact stress to long-range stress
might be expected to reduce the shear strength of the soil-water system.
Also of interest is the effect of changing, after consolidation and
prior to shear, Ta /(-X) in a soil to be sheared at constant volume.
If C am /(R-X) for a soil of a given fabric (particle arrangement) were
changed prior to shear and the fabric did not change, it might be
expected that the behavior of the soil in the vicinity of a newly
broken contact would be different from a soil of identical fabric and
consolidation pressure in which T a /'(N-X) had not been changed prior
m
to shear. In particular: if (7-A) were increased prior to shear, this
dispersive force would tend to hamper the reforming of broken contacts
during shear and some normal stress in the area of the broken contact
would have to be transferred to adjacent contacts, to (B-X), or into
pore pressure, u. To transfer the normal stress to (B-A) would require
some additional consolidation and. higher than normal pore pressures
would result during undrained shear.
This could appreciably reduce the
undrained shear strength of the soil by reducing the effective stress.
Thus, we are herein concerned about the effect of salt concentration
on'
/
A).
m
The attractive component (A) of the net electrical effect, (R-X),
Ca
is believed due primarily to van der Waals-London forces.
They derive
from attractive forces associated with the charge on sub-atomic particles
in the clay mineral lattice and are not much affected by changes in the
pore fluid characteristics.
The repulsive component (n) of (1-I) derives largely from the net
electrical charge on clay minerals and is very sensitive to the characteristics of the pore fluid (Lambe (1958)).
The electric charge on clay
minerals results from the substitution of atoms into the crystal lattice
during the formation of the clay crystal that have a valence inappropriate
to produce a neutral crystal.
This phenomenon
is called isomorphous
substitution and in clay minerals it always results in a net negative
charge; positively charged clay minerals have not been observed.
The
negative charge (which varies from particle to particle depending on
the quantity and type of isomorphous substitution) is neutralized by a
number of cations that are distributed over the faces of the clay
If the clay particle is placed in a
particles when the clay is dry.
dispersive medium (such as water) the cations will cluster around the
clay platelet.
The volume of fluid around the clay platelet in which the
concentration of cations exceeds the concentration of cations in the bulk
pore fluid is called the diffuse double layer.
The distribution of
charge in the double layer, and consequently the electric potential, is
described by the Poisson-Boltznann Law (for distances greater than
3
several cation diameters from the particle surface) and is similar to
the distribution of electrons around a charged, infinite capacitor
plate.
A repulsive force results when two like-charged doible layers
overlap.
By using the Poisson-Poltzmann Law, one can predict that the
double layer repulsive force will decrease as the concentration of
cations in the bulk pore fluid incracs (see IADD (1961),
. IIT-3).
Consequently, one would expect the not electrical effect, (-X), to
decrease if the cation concentration were increased, and vise versa.
It is recognized that chanrges in cation concentration may have a
direct effect on ra
for a given effective stress as well as an effect
m
on (I-X).
effect.
Unfortunately, it is not possible to estimate the actual
If one would assume that
C a
is transmitted over extremely
small interparticle distances ((6 W) it would seem that changes in
cation concentration would have little effect on
"a
m
since cations
would be largely excluded from the contact area because of the proximity
of the particles.
Direct effects of cation concentration on contact
stresses will henceforth be ignored.
Consequently, the expected efrect of changing salt concentration are:
1) Reduced shearing resistance for a given effective stress
if the (R-X) component of F is increased,
2) Higher pore pressure during shear if (n-X) is increased at a
constant & and no change in fabric results.
C
These effects would be expected assuming that the clay particle
acts as an infinite flat plate.
In addition to the negative charge on
clay particles due to isomorphous substitution, there is evidence of
positive charges on the edges of clay particles (see LAIME (1958))
resulting from unsatisfied primary valence bonds. Edge effects would
be expected to appreciably influence the behavior of soils if the edge
area contributes appreciably to the total surface area.
Some data on
kaolin (WISSA (1961)) show that changes in salt concentration affect
soil behavior quite the opposite of what would be expected from consideration of double layer forces only.
1.3
A Review of Some Previous Investigations:
In recent years, the leaching of salt from the pore fluid of
marine clays has been shown a factor contributing to their sensitivity.SKEMTON and 11ORTEY (1952) have noted that the sensitivity of quick 2
clays is due to a low remolded shear strength.
remolded shear strength to high liquidity index
They have related low
and have concluded:
"the liquidity index of an extra-sensitive clay must be very considerably
greater than that of a medium or low sensitivity clay existing at the
same overburden pressure."
SKEMPTON and NORTHEY and, later, BJERRU1
(1954)
agree with an earlier conclusion by ROSENQVIST (1953) that the liquidity
index of these quick clays results from removal of salt from their pore
fluid subsequent to consolidation.
Removing salt from these illitic
clays reduces the liquid limit and, hence, the plasticity index.
There
was no evidence of fabric collapse during leaching; the in situ water
content was little affected by removal of salt after consolidation.
ROSENQVIST (1953)
reported having performed tests on a "typical
quick clay" in 1946. This clay had an undisturbed shear strength of
2
0.12 kg/cm and negligible remolded strength. After the salt content of
the pore fluid had been increased to 30 gm NaCl/liter, the shear strength
(remolded) increased to 0.09 kg/cm 2 .
The salt was then removed by
2
dialysis and the shear strength measured as 0.10 kg/cm . Upon remolding,
the shear strength was again found too small to measure.
In these
experiments no attempt was made to determine whether or not the water
content had changed during leaching.
undisturbed shear strength
remolded shear strength at the same water content
1.
Sensitivity =
2.
"quick" has been defined as meaning a sensitivity in excess of 16.
3.
the maximum value of principal
"shear strength" hereafter is
stress difference in an undrained test.
4.
liquidity index = L.I. = w w P
w -w
1 p
A
Tnese tests clearly illustrate that the sensitivity of the clay
had been greatly increased by removal of the salt in the pore fluid
after consolidation.
Tne effects of salt removal on the undisturbed
shear strength are not very conclusive.
S1MITON and NOR--TH
(1952)
performed tests on a Shellhaven Clay,
undisturbed, that had a relatively high natural salt content (26 gm/l).
By leaching out salt until the concentration in the pore fluid was
reduced to 12 gm/liter, the sensitivity was increased from 10 to 18.
It was also noted that the undisturbed shear strength was reduced
from 3.2 to 2.8 psi.
The water contents for the leached samples were
found to be the same as those of the non-leached samples but no confining
pressures had been applied to the soil samples during leaching.
quently, it
Conse-
could not be concluded that the fabric would not have
collapsed had a confining pressure been applied during leaching.
More
tests were run on a batch of Horten Clay samples subjected to consolidating forces throughout the leaching process.
The results of the tests
are presented in Table 1.1.
TABLE 1.1
Effects of Salt Leaching on the Shear Strength of Horten Clay
Su (usi l
leached
1.29
w
31.6
leached
1.29
31.2
- no t
not
L.I.
undisturbed
remoided
St
aX_ I
saltgn
28.6
1>.b
1.23
0.145
0.13
1-1
-
24.5
1.7
1.7
0.144
0.03)
3.7
2.2
from:
*Sensitivity
S
XT and NTH0EY112
(1952)
These data illustrate that the effect of salt leaching on this
particular clay had been to in rease the sensitivity by decreasing the
remolded shear strength; the undisturbed shear strength had not been
affected.
The water content remained nearly constant during leaching
despite the confining pressures and the increase in liquidity index was
affected by a reduction of the liquid limit.
LAME (1953)
has proposed that creating dispersive double layer
forces while the fabric of a clay is hield constant, as 'would result
from leaching salt from an illitic
clay, should greatly reduce the
remolded shear strength and reduce the undisturbed shear strength to
some limited extent depending on how the contact stresses in the undisturbed soil were affected by the change in salt concentration.
LA13E
gives some data for undisturbed Boston Blue Clay leached with a disperundisturbed shear strength
sant (Na6P4 01 3 ) that support his theory:
was reduced by the leaching 30-50; remolded shear strength was reduced
from 41 pounds per square foot to zero by leaching, indicating an
increase in sensitivity from about 10 to infinity.
The work sunarized above illustrates that sensitivity can be
increased by removal of salt after consolidation.
strength are inconclusive.
BJEREUM (1954),
The data on shear
after a systematic inves-
tigation of Norwegian marine clays, has related low S /
plasticity index.
ratio to low
Since salt leaching was known to reduce the plasticity
index of illitic clays, Bjerrum concluded that salt leaching should also
reduce the undisturbed shear stren;th.
BJERUM (1954) presents some
data for tests on an artificially sedimented Norwegian marine clay that
show reduction of undisturbed shear strength from 0.31 kg/cm 2 to o.06
kg/cm2 after the salt concentration in the pore fluid had been reduced
from 39 gm/l to 1 gm/l.
Sensitivity was increased from 6.4 to 90.
data, contrary Lo those of SKEMPTON and NORTHEY (1952),
These
show a remarkable
reduction of shear strength accompanying a reduction of salt content.
More complete and conclusive data regarding the effect of salt
content on shear strength has been presented by BJERRUM and R0SENQVIST
(1956).
An extensive series of tests was performed on four groups of
samples of a typical illitic Norwegian marine clay of glacial origin.
Its properties are sumnarized in Table 1.2.
Three groups of the samples
were prepared with salt-water pore fluid and the fourth group was prepared
with fresh-water pore fluid.
The salt samples were produced in the
laboratory by sedimentation; several months were expended in the process.
The fresh water samples were consolidated from a slurry at an initial
water content of 150
(L.I.=13.3) to avoid segregation of silt particles.
After sedimentation and K -consolidation, the three groups of salt
samples were treated as follows: one group was leached while under
consolidation pressures so that the salt concentration of the pore
fluid was reduced; a second group of salt samples was leached as in
the first groun until the salt concentration was reduced, then reconsolidated to higher pressures; the third group of salt samples was
not subjectcd to any leaching at all.
The data obtained from these tests are rep;o-duced in Figure 1.1.
The following conclusions are indicated:
1)
the undisturbed shear strength was reduced approximately
40~, by decreasing the salt concentration of the pore fluid
from 35 gm NaCl/liter to 5 sm NaCl/liter,
2) the leaching out of salt has not caused any appreciable
change in the water content,
3)
a ntastable structure has resulted from the leaching as
indicated by the large decrease in water content and increase
in shear strength of the two samples reconsolidated to higher
pressures after leaching,
4)
the fresh water clay had a lower water content and a considerably higher shear strength than the unleached salt clay for
the same 4c
BJE~PUM and ROSETQVIST explained the reduction of shear strength
caused by leaching salt from the
-srum
Clay by reference to the following
equation in terms of Hvorslev Parameters (the choice of Hvorslev
Parameters is arbitrary; an equally valid equation can be had by replacing the Hvorslev Parameters by effective stress parameters,
)/2
cr Cos < r + &c(KO+ A (1 - KO))sin
1 + (2A f - 1)sin
*r
r
TABLE 1.2
Properties of Asrum Clay
Natural Clay
natural water content
liquid limit
plastic limit
plasticity index
Salt Clay
L7r
28r')
18.8%
9.2%
41.5o
20.0e
21.5%o
#
and c):
1
TABLE 1.2 (Cont'd)
Properties of Asrum Clay
Natural Clay
Salt Clay
salt concentration (NaCl)
base exchange capacity
0.05 gm/l
13.5
35 gm/l
clay fraction (-2A )
502
0.18
50%
0.43
activity
-
-
For a given consolidation pressure and assuming that K does not change
because of leaching, there are three variables in the equation that can
account for the reduction of the principal stress differences caused by
leaching:
1) A reduction of Hvorslev cohesion, cr'
2) A reduction of Hvorslev friction, +r;
3) An increase in A .
BJERRUM4 and ROSENQVIST believed that a reduction in cr would not account
for the 40% reduction in shear strength because the cohesion in
Norwegian marine clays is not thought to contribute nearly 40% of the
strength at maximum principal strength difference.
Unconfined compres-
sion tests, which permitted observation of the inclination of the
failure planes, indicated that 4r was the same for leached samples and
salt samples.
It was concluded that an increase in A
the reduction in shear strength.
The A
must account for
was believed to be higher for
the leached samples because their fabric is highly compressible. Shear
stresses cause a partial collapse of the fabric resulting in high pore
pressures.
This analysis implies that the strength parameters at maximum
principal stress difference for this clay, in terms of effective stress,
are not affected by leaching out salt.
The smaller Mohr's circles at
failure for the leached samples will be shifted to the left by their
larger Af until they are tangent to the normal Mohr envelope and exhibit
the normal values of
4
and c
(or i and c ).
To account for the 40%
reduction of shear strength observed by BJERRUM and ROSENQVIST, Af at
maximum principal stress difference during an undrained test would have
to have a value of 3 or 4. Unfortunately, pore pressures during undrained shear of an undisturbed quick clay had never been measured.
Attempts had failed because the metastable structure broke down during
reconsolidation in the triaxial cell and the resulting soil did not
behave like a quick clay.
Some data is presented in Section 4.5.2
which suggests that the effective stress parameters are indeed affected
by removal of salt but that the pore pressure at failure is little affected;
the increase in A f is not due to abnormally high pore pressure but,
rather, to reduced shear strength.
The clay-water systems that have been discussed thus far have
consisted of glacial marine clays with primarily illite as the clay
mineral.
A different reaction to salt might be expected in a kaolin
because of the affect of the positively charged edges of the particles.
WISSA (1961) has recently completed experiments on a homoionic, sedimented,
sodium kaolinite in which NaCl was diffused into some of the samples
after consolidation.
His data lead to the following conclusions:
1) no change in water content is caused by the introduction of
salt into the pore fluid after consolidation,
2)
the undrained shear strength is not affected by the introduction of salt after consolidation,
3) the principal stress difference at maximum principal effective
stress ratio is reduced and the effective stress parameter,
)
at maximum principal stress difference is reduced by the introduction of salt to the pore fluid after consolidation.
These data show that the effect of salt on strength behavior is not the
same for all clays.
The effect on the effective stress path after
maximum principal stress difference is reached caused by a removal of
salt in illitic clays is accomplished in kaolinite by the addition of
salt.
In kaolinite, where the effects of positive edge charge may
outweigh the effects of double layer forces, the effect of salt on
shear strength may be nil or even the opposite of the effect on an
illitic clay.
Data presented by LADD (1961) for Vicksburg Backswamp Clay (25%
illite, 25% montmorillonite, w 65%, w -25%) show increased shear
p
strength and increased maximum obliquity ( %/ #3) for samples in
which the content of pore fluid salt (CaCl 2) was increased after consolidation.
Similar results were obtained for samples in which the salt
had been added before consolidation.
These results are in agreement with
expectations based on double layer considerations.
1.4
Scope of the Present Inrestigation:
In view of the interpretation by BJERRUM and ROSENQVIST (1956)
of the behavior of the leached Asrum clay and of the expected effects of
salt leaching discussed in Section 1.3 and 12 respectively, it appeared
desirable to measure the pore pressures during shear in a quick clay to
determine if A, alone would account for any reduction of the maximum
principal stress difference caused by removal of the salt after consolidation.
Also, measurement of pore pressures would permit an assessment
of the changes in effective stress parameters at maximum principal
effective stress ratio (complete failure of the soil skeleton) caused
by the presence of salt.
Consequently, this experiment has two main purposes:
1) to investigate the effects of consolidation in salt water
(35 gm NaCl/liter) on the strength characteristics of a
local, illitic, silty-clay (IT 1139) (Boston Blue Clay).
2)
to investigate the ef'fect on the strength characteristics
of the same silty-c ly of leaching salt from the pore fluid
after consolidation.
1.11mim,_
, , .
_____
LIQUkl>
-f
LIMC7
3.9
34.0
P , S$Tv T C, 1 1 '
I
q*r~
LIQW:)ITY
4
7 "Y
1-
-a6 A C
su
$
C)
.6~5 Gcj.
______
$fwt Cz~ ~o.
s
I
u (VvUL&).
9
.r
I
I
±
10
0j
±
A0
10A
(b)
o
1.0
6,
- S fikC,
o
E
1.0
~c~Z
s
+1-~
2.0
2.0
3.0
C-KGC.?-1
(CO)
Su FoR5A9-et
~
e~
z
~P)
--1-_
4-
-L'CNE
0
I.A.
CC.
a.0
CHAPTER II: PROCETURES
2.1 Sumar
It
was hoped that a clay with reasonably high sensitivity
could be produced in the triaxial cell by removing the salt from the
pore fluid by diffusion after the soil had been isotropically consolidated in the cell.
The leaching had to be done in the triaxial cell
in order to avoid the disturbance to the fabric that would inevitably
occur if the soil were leached before consolidation in the cell.
To perform the experiment, three types of samples were needed:
1) "water samples" in which the salt content of the pore
fluid was relatively low (less than 5 gm NaCl/liter)
during consolidation and shear,
2)
"salt samples" in which the salt content of the pore
fluid was 35 gmI NaCl/liter during consolidation and
shear,
3) "leached samples" in which the salt content of the pore
fluid was 35 gm NaCl/liter during consolidation and
reduced to much lower values prior to shear.
Triaxial testing was performed on the samples to obtain the
following data:
1) effective stress parameters at maximum principal
effective stress ratio and at maximum principal stress
difference for the water, salt, and leached samples,
2)
Hvorslev parameters at maximum principal effective stress
ratio and at maximum principal stress difference for
both the salt samples and the water samples,
3)
effective stress paths for four leached samples; ideally,
each of approximately the same age and having a different salt content.
Water contents and shrink;age limits after failure were determined
as possible indicators that greater fabric collapse had occurred in
the leached samples during shear.
2.2
Sample Preparation:
The results of previous investigations sumnmarized in Section
1.3 indicate that an appropriate soil had to have at least the following
two characteristics if any considerable degree of sensitivity was to
be obtained by leaching salt from the pore fluid after consolidation:
1) it must flocculate readily and produce a highly porous
fabric when consolidated in salt water,
the liquid limit must be affected by salt content so
that the liquidity index would be increased if salt were
removed from a conoolidated sample.
The soil samples used in this experiment have proved to satisfy these
2)
requirements only to a limited degree.
not been produced in the triaxial cell.
A quick clay has definitely
Nevertheless,
the sensitivity
has undoubtedly been increased considerably and the results of the
tests on leached samples are believed to illustrate adequately the
clays.
effects of leaching salt from consolidated samples of illitic
Both the salt samples and the water samples were prepared from
the same batch of soil.
A slurry was prepared by mixing the soil and
its natural pore fluid with enough demineralized water to bring the
water content to 421L. The soil had not been dried prior to this mixU)g.
From this batch sufficient soil was removed to make the salt samples.
The water content was increased to 724 (L.I.=3.2) and sufficient salt
(reagent NaCl) was added to bring the salt content of the pore fluid
to 35 gm NaCl/liter. This slurry was thoroughly mixed and then deposited in an evacuated consolidometer and consolidated under YO conditions to 1.5 kg/cm
( C1 ) in a single increment of pressure.
The
samples, when removed from the consolidometer, were stored in transformer
oil.
The material for the water samples was the remainder of the original slurry at 42% water content.
This soil was mixed with sufficient
demineralized water to bring its water content to 51% (L.I.=2.7) where
it appeared about as stiff as the salt slurry.
The slurry was then de-
posited under vacuum in the same consolidometer used for the salt samples and consolidated under KO conditions to 1.5 kg/cm2 (
I ).
The
samples were removed immediately after primary consolidation was completed and stored in transformer oil.
2.3
Triaxial Testing:
Strain-controlled, consolidated-undrained triaxial tests with
pore pressure measurements were run on all samples.
English triaxial
equipment (see BISHOP and HEMEL (1957)) was used with the null system
developed by the Norwegian Geotechnical Institute for measuring pore
pressures (see ANDR1ESEN, et.al. (1957)).
Pore pressures during shear
were measured through a porous stone at the bottom of the sample.
Samples were trimmed to 10 cm2 area and 8 cm length and set up
with eight 1/4" wide filter paper drains and latex membranes (two
prophylactics with silicone grease between). Except for the four leached
samples, the tests were set up with bottom drainage only.
All samples were back-pressured (usually 3 kg/cm ) to increase
the pore pressure response (B= 4u/ 3 ) and sheared with constant pore
pressure and varying cell pressure.
For most of the normally consolidated samples, the nominal strainrate was adjusted to 0.00036 inch/min until the sample has reached maximum principal stress difference (about 2% strain) and then increased
to 0.0008 inch/min for the remainder of the test.
A couple of tests
were completed at a constant strain-rate of 0.00036 inch/min.
The time
to failure (maximum principal stress difference) varied from l- to 6
hours depending on the strain and the deflection of the proving ring.
In total, 25 consolidated-undrained tests were run that are reported herein: 1) 4 leached samples, 2) 9 salt samples, and 3) 12
water samples.
2.4 Arrangements for Leaching7 out Salt:
Four salt samples were set up in triaxial cells to be leached
(actually diffused) of the majority of salt in their pore fluid. The
leaching was done by permeating demineralized water through the filter
paper drains around the periphery of the samples. Arrangements were
provided for drainage at both top and bottom surfaces of the samples.
Salt from the pore fluid diffused into the filter paper drains where the
concentration of salt was less.
The rate at which water permeates thru
the filter paper is a variable affecting the rate of diffusion.
rate of permeation decreases appreciably with time.
The
Sample (L-4) passed
11 cc of water during the first 39 hours of leaching but only 10 cc
during the last 105 hours (12 days later).
2.4 cc during
16
Sample (L-1) passed only
the first day of leaching and this rate decreased to 0.9 cc after 50
days. It can be seen in Table 2.1 that (L-4) produced a lower final
salt content than did (L-1) although the former was leached for 4
much shorter time.
The seepage head on all samples during leaching was
2
about six feet of water or 0.2 kg/cm . This seepage force caused a
condition of anisotropic consolidation on the samples which was inadvertantly ignored when the samples were sheared,
2
consolidated to 3.0 kg/cm prior to leaching.
Table 2.1 shows the age,
All four samples were
leaching time and salt content of the
pore fluid at failure for the four leached samples.
TABLE 2.1
aged (days)
leached (days)
salt (gm/l)
(L-1)
56
48
3.1
(L-2)
42
30
4.7
(L-3)
46
14
4.o
17
14
2.6
(L-4)
Salt concentrations in the pore fluid after failure were determined by measuring the resistance of the pore fluid with a resistance
bridge as described by WISSA (1961).
2.5
Water Contents at Failure:
After completion of the triaxial tests, the samples were
weighed and measured.
Water contents were determined for the top,
middle, and bottom thirds of each sample.
Water contents of the
samples containing salt were corrected for salt content so that the
water contents quoted would be a true indication of void ratio at
failure.
2.6
Void Ratios After Oven-Drying:
Shrinkage limits of the top, middle and bottom thirds of the
triaxial samples were determined by immersion in mercury.
These
shrinkage limits are actually the water content required to fully
saturate the void space of the o-en-dried sample.
Such water contents
were computed after accounting for the salt content in the soil
pattie.
Volume corrections for the salt were not made, however, since
it was felt that the salt was still in solution when the fabric of
the pattie was set up and thus the salt crystals would have had little
effect on the volume of the dry pattie.
CHAPTER
III:
TEST RESULTS
Classification Tests:
A silty-clay (MIT 1139) from a local clay pit was used in
this experiment. Some properties of the soil are given in Table 3.1.
3.1
A grain-size distribution curve is shown in Fig. 3.1.
TABLE 3.1
Properties of MIT 1139 (Boston Blue Clay)
water" clay
30.0
liquid limit
plastic limit
17.5
plasticity index
12.5
40Q,
clay fraction ( 4 2,4)
activity
0.31
salt concentration (NaCl) 2-3 gm/liter
specific gravity
2.77
3.2
"salt" clay
34.7
17.7
17.0
40%o
0.42
35 gm/liter
2.77
Stress-Strain Curves:
Plots of principal stress difference, pore pressure, and A
versus strain are presented in Fig. 3.2 through Fig. 3.26. These data
have not been corrected for effects of piston friction and filter strips.
Pertinent information from the stress-strain curves is tabulated
in Tables 3.2 and 3.3. In these tables, the values
of ( -1 - C3
quoted have been reduced 0.10 kg/cm 2 to account for the strength of
filter-paper drains.
In most of the triaxial tests on normally consolidated samples,
the strain-rate was increased -'rom 0.00036 in/min to 0.0003 in/min
soon after the maximum principal stress difference had been reached.
The increase in strain-rate produced a noticable discontinuity in the
(( 1 - T3) vs. strain plot.
A few tests (S-9, Fig. 3.4) run at
0.00036 in/min failure are consequently smooth in the region of peak
( Gl - C3 ) and have been used as "models" in arbitrarily smoothing out
the data where the strain-rate had been increased.
The values of
maximum stress difference tabulated in Tables 3.2 and 3.3 are the
19
maximum values prior to any chan;e of strain-rate. All quoted strainrates are, of course, a.nly nominal and were affected by rate of proving
ring deflection.
3.3
Shear Strength:
Undrained shear strength (SU) is plotted versus water content on Fig. 3.27. In this plot, the water content used is the average
of the three parts of the sample at failure. The values of S have been
u
taken from Tables 3.2 and 3.3 and have been reduced from measured
values by 0.05 kg/cm2 to account for filter paper drains.
The shear strength is plotted versus consolidation pressure
(
)
in Fig. 3.28b for the salt samples and in Fig. 3.29b for the water
samples.
The S vs. 0- plots have definite cohesion intercepts and
u
c
provide the following relationships:
S
u
= 0.15 + 0.28 a-
(Salt and Water Samples)
c
The data used for these plots are from Tables 3.2 and 3.3 and are
reduced 0.05 kg/cm2 ((r
.
3)/2) as a filter paper drain correction.
3.4
Effective Stress Parameters:
At Maximum Principal Stress Difference:
Mohr failure circles for the plane of maximum shear stress
are plotted in Fig. 3.2 8 a for the salt samples and in Fig. 3.29a for
the water samples.
These plots provide the following relationships
between shear stress and normal stress on the failure plane at failure:
Salt Samples:
= ifftan
ff =
Water Samples:
.50
ff =fftan
4
(*=
26-50)
ff
(
= 24.50)
4
T ff = 0. 5%f
Data from Tables 3.2 and 3.3 have been used in these plots.
Mohr failure circles for the leached samples are shown in Fig. 4.1
on page 77.
I
At Maximum Principal Effective 3tress Ratio:
Effective stress paths on the plane of maximum sher St ress
are plotted for normally consolidated samples in Fig. 3.30 and Fig. 3.31
for the salt samples and the water samples, respectively.
The data
used in these plots were obtained directly from the stress-strain
arves and have not been corrected for the strength contribution of
filter-paper drains.
Consequently, the values of shear stress shown
are somewhat larger than on Figs. 3.28 and 3.29 and in Tables 3.2 and
3.3.
As a result, the small cohesion intercept of 0.05 kg/cm2 results.
From Fig. 3.30, 3.31, and 3.32, the value of the friction angle on
the plane of maximum obliquity can be deduced.
Table 3.4 compares the
maximum obliquity angles for the three types of samples.
TABLE 3.4
Sample
Salt
Water
Leached
______u
270
250
250
30.50
280
280
The cohesion intercept has been ignored in the determination of these
effective stress parameters.
In Fig. 3.32 the stress paths for the four leached samples are
plotted in comparison with stress paths for salt and water sanples of
the same consolidation pressure.
all reach maximum obliquity on the
It is seen that the leached samples
-angle corresponding to the water
samples.
In Fig. 3.33 and Fig. 3.34, stress paths are plotted for the
over-consolidated salt samples and water samples, respectively.
3.5
Hvorslev Parameters:
Hvorslev Parameters at maximum obliquity are presented for
salt samples and for water samples in Fig. 3.35 and 3.36, respectively.
The data are plotted by the method of BISHOP and HENKEL (1957).
graph is a plot of the equation:
(Cr
2f
r
CrCO or
e*e
(1
e
e
-
sin +r
+f
sin4
r
)
The
e (1
e
-
sin
*
r)
where the subscript "f" refers to values measured at failure and
F
e
is
the value of ic for normally consolidated samples from Fig. 3.27 or the
water content of the failed sample.
By plotting the data as shown, the
slope of a straight line through the data is:
the intercept on the ordinate is:
Cr
S
OC
cr
=
-e
+
crCos $ r/
Offtan + r
sin 4 r/(l - sin i );
(l - sin 4
and
); where:
(Hvorslev's shear stress eq.)
shear stress on the failure plane at failure
a "cohesion" component of '
by the parameter
related to void ratio
'
e
tan
b
r
a "friction" component of
^C
'ff;
water content were kept constant
ff
/ ~0Cr
if
f
the
th
The Hvorslev Parameters for the salt clay were obtained from
Fig. 3.35 and from Fig. 3.36 for the water clay.
Graphs are presented
of the data both at maximumn principal stress difference and at maximum
principal effective stress ratio.
In computing the parameters, the
data from Tables 3.2 and 3.3 (corrected for filter strips by subtractir
the cohesion intercept of Fig.
3.30 and 3.31) were used.
3.6 Water Contents at Fati lure:
Each of the stress-strain curves tabulates data for the
water contents measured after failure for the top, middle, andi bottom
thirds of the samples.
In Fig. 3.37, the difference between the
average water content and the water content of the center piece is
plotted.
The water content of the center piece is probably a better
indication of the water content in the failure zone.
3.7
Void Ratios After F%.ilure and Oven-Drying:
Since shear stresses are believed to cause particle :erienta-tion, void ratios after failure and drying were measured to indicate
possible more fabric collapse and particle reorientation in the leached
samples than in the salt samples.
Fig. 3.38 plots oven-dried void ratio
against consolidation pressure for all the samples tested in undrained
shear.
TRBLF- 32
0
z
~'1w~S~
'5Wwlse
J
(~)
RFTE-R
a:
Ry 0C O F,
5
QFCm
'i
TIs
FT
0~
I-
LI~
&3~
'a'
0
,25.~
Is-i
(3)
C,
I SLT
(z)
(a;/~ 3 \) ~1F~(
AT:
AT:V
(( :(ys)
cs-3
0/Nu
1-1;70F*7T71
*
A
(5~3
q9
1*~~"~~~
~
O.6Z
Omi COO
~
4 2AZ
.25
o
42
1103
3.;Zg
I2.1..il
3.1
5.90
s-3
98
6.o
2.1
2.
3Ac9
4.131
2. q7j I.Ot
&o3.~
1A.69
3JAI
I .Eo
I.OL
1,5.
5 -A ge.
97
Z~
J~S,0
162~
3
Pz) G 2. -P
1 .06 3.3
M?.
'---Z
Z.18
.18
Z-5
3
^30
-0.0!7 6 .C
3.S
c9 iq 0
2
.ZA
-?Z ' .
4.1,
~.I
01s g .381
S,
1.90
__
1.8 8
ji.~
I
~
j2~.6
~3.3o 1 .
______
_________
_____________
L -I18S
'2q. O
t~q
1
2-23
.o
8~0 R-3
--9
S-1O~9 2
o.91 0E1
3.~
___
__
___
___
____
____
______
2.o ,3.o
13.1
14j
-~
56
t
~
1.9
-~
__
__
__
-~
__
____
j
~
Y~ ~.68
__
__
48
t~~dj
/
-r
~.&
0.C~~c' 3&. ~
~
~
.-I1
1 Co
q5 .9
-
10.901
NCT~fr~
TT
' t4e.AR
j
__
__
___
__
____
2
0.9 1
tcA G
.S
__
-
___
1_____
I
I
____
Ie
.q2~q.
1A3
C~E~ 3.~I
__
2____
is l
IL.09 ~
____I
__
__
__
____
__
____
-f
__
____
__133___
I.Srq
o).p 1
3
1.03i
28a8.63
1 .zo 0.
4'0
o351 ~
15
UW-)
I___
I..84
__
1 .(,3
C0.52
j3.~1Z~
tA$$~e,~
PPES5'.;It;-
l
C.\,, >
3.03z
0).Sl
~0.0
I.1
___
1.. 51
C A.o
1 A
OF
1.
EoA!
2.~
I aA
~
UE~$
30
12. C
Ij1
rq'I
L...r~zzz
4
-A
'A2
1
1
11 ~3 18.o ~
0.88 1.01
' __
~~66
'~
C0.e3 13.?VI
12.091'8.0
LIWPEUt
I~
TALBU3.
ALL
S1RES $tt> IN
'
-I
Ow
a:
3
0
~to.
w-1
c)
(Plv4(
26.9
?o. ~
(WSL
le
20.6 6.0q
~ .4
A
MT:
RT(04
(e)
(~'/X~(3)-
MAY
TI
(o/)
2.5,
--I-
(~)
'G'3J
3.52
2.11
T3
-4.35
r)O~
7'
1
1
'Z . So
If-32
5.ZS
o.
&,
6
1.3L ,3.0
A.I., .1
0
(.%)
25.1
2 1. qs
3.Z 0
21 .'?9 I3.'1 S
7T.125.2
(6
j ;52P6
c
E
25_
12j.9o
-01-- - ..I 1
_.(o_
3 19E)
9
I
Alce
j2.
6
.
SG
46q
20
?~o
.(o
8.3
.-331
I'7AZ
IAZ~.
1-7. .5&o
1
3.5. 3~1
21.A
fc
31VI
11
L1
E .67-.^
a. 7. (or-)
)4
--
0..
I.A
2.38
a
.Z)l
;Z.
4.3
3.0 ,6.c
.
0.661
1.
R. 61
6.'
R-2
1,11
2.qB
OB
A )
2.512c
3.o, ~.o,
3.0
l.o
'..
I.A
2-o
4,c~
.,
J2?.8
1-
1.1
i1
____
30.
3.o
3.$o~
3.So
.10
W-1_2_
2.30
9
o
C)
I As
q.13 oo22,c> 35, 1.,
W-8 5 .77 1730
'T-9
4
CC
Vf-7
2.55 1 .39 8 O.cut
'1.52
11 4iL1
0~.
.
53-3 1.o'i
2 ~qq
~Aq
3.oz 1-0.03.2
1.0p
5.?
2
i-09
2
~i
2. 3q
Z
10
2
1oi
2
IA'?
?21$
O.6 13.?'' I-OAtZ
o01
1. .3'?
1A2g
4)
0A~ 3.o(z 7-O.Z3j 0.
$0
0.26 SAq~
.__
ct,%jt Ci EC.A'J$E
0;7
0
44R
=~
c~No~~~N
S
~oo
LEA14JZ
IA
=0.xz
IN~
W~~4T
?~$.jr?~Z
z
5UbR~~~
hiq
Q.58
GRAIN
SIZE
DISTRIBUTION
MIT 1139
MIT
SAND
CLASSIFICATION
MEDIUM
COARSE
SILT
CLAY
MEDIUMI FINE
COARSE
FINE
COARSEI MEDIUM
FINE
100
90
80
w70
-t-
-
-
-
--
>60
Im
-
-
-
-
--
-
- -
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-_
-
7w50
z
-_
z
_-_
--
_
-__
-
z
-L
-
-_-
-
w
~a2
10
0
1.0
0.01
DIAMETER IN MM
0.1
0.001
FlcuRE 3.1
0.0001
0
.- O
-0
-~
c~
/IT
6
m1
o.A
0.2
0
.~
0oc 0
o
0oo
4
(V
U-S-
TEST NO .
uA&.O .o%&
Cra.o9%&
i~2
i15
Lig'
'-0.6
TOP
nom
25.8 25.6
16." 11.9
FIGURE
3.2
oL
2
3
5
k 6
STEN
9
~4o(
1o
C .III
---
3.o~
0
a
~-
0-
o
0-
1I3
0
C
/
4:
.1
/
Ki
0
_
~
~
-
k
6
6---
~-A
rJ
5:
0
L4
f/tb
A1.?
0
I!
II
I
Th.ST NO. CIU-S4
Rb-
'<"tv
3.0 l%-l7/ t.3.
IF
MIWIO.
'TP
eoTTO1
CA
Rc7u Fg- 3.3
0'-0
3
4
$"TRRAN
t4
0/0
9
a C)
II
~~0
0
o0-0
a-
0-o
.4'
2
TEST No. Clu-ScP
=
3
KQ.0
T-o?
0.6
3.c) 101/ctl z
U-:;:
0
Wr___
Ui,4
252
WI LJ~jt
/1~3jI. 0
FGURE 3., 1..
0'-
0
3
4
5
6
17
,
10
~
(S/
Q
C'
-~0
3.C1
A,
p/
0
(j'
0
P
&-
0
A
A
L
A
00O
O
-0
-- C
2.5
N-
1.5
TEsr No. CIU-S-2
vr 4.o
TOP
w4.o
".
imotewT7e1l
o.5~
RcuRE 3.5
OL
a
3
$
STRaiN
-
*-/o
1.2
0
0~0
/
/0
'lv-
/0,0
~I.0
0.E6
O.A
~
.0
-A
A
,-
-o-0
-
0
--
0-0
0
0
-
T,
0
OA
0/
...
0
0'
/4/A
~0
0
2.0
ThsTr NO. CItJ-S-3
6 -*r.o /r<M1. E 8
U cc,-;;
"TOP
toC
L
I
23.0
1
22.
A~1t6.aK~~i.
hGURE
0
mkjoorcz
3.6.
3.
0
0'
3J.1
/
(
OAo
4i
o.ao
0
cxoo
0
10.0
-o.zo
1.0
0
0
0
0~~
aA
0
1.o
0 .6
TEsT No. CI U- S- 6
**1
~0
ac&= o.5 % k=3.o
0
oc=
-Top
neo.
a
eom.
-224.5 24A.6 24-.5
0.2
X%%- 16A4 1').5 16.6
,-c~bk
FIGURE- 3.2
-021
-
'o
1
2
3
4
5
6
ST RAliN ~ l
?
9
10
---
11
K%4
-o
C
-
-
-
0
0
C
0
I
0.110
I
K a
0
0.20
%a-
ol
o
-
Q0
0
C)
0
0 p
7<
.3
0/
IA
.TEST NO. C1US8
uj'I?c 123 L 3,
I
a
~
R=1.IV
164
0A~~
FIGURE 3.8
-4
o
o
-1
2
STRR~N
___
3
'-RIo
5
6
9
10
C>-
0---0
3.0~
~0
0
0.?
4:
rz
A
A
0
0
0
p.,
0
~c1
'U
Th-STNO. CIU-S-5.
0'
/
0,.,
0
TcMP
M10E v orrcm
242' 242
0.6/
LJ$L1~.CV?.6'.SEA
RiGuRE.
JA-
-0?2'-
1.
.2
,
5
STrRlwP,.N
9
30
0
C
0
7
C
o
~
0
-.-----
-
0
I
14
. 0
00
0~
L /
0
~b4W
lo
A
/
TEsr No. C U-\- 4
o.
-fop
e
0.21
0
i'.o 16
FIG U RE
2
6
3
5T RA N
-~* o
7
3.10
9
Eor
1'1.Q)
3.0
-o
0
8
0
~
0-
I
s
C
~~
/
a.o
7
I
0
'3/
/
to
0
-- o00
goO
0
t
0
0000,
00
T-6
bko
//4
1/ina
jwa-A
~a
A'
TE..ST NO. CIU-W--2
3.0I&
k%
0.8,
2J.[2.3 22.0jt2:.-j
CL.4
i 0.21
FiGURE
0 L-
0
1
5
ro
,5"TR R V4 ,- 0/0
±7
3.11
$ 5 10
11
M.0
lea
.4 p-
0-0
19
0
119
a
l
0
T/s z-.O -
I.A
0
(-)
(S
<3
Q.E~
7~zj
0.6
40
I!
WSL
iD.E$ )
I
cx ~
FGURE.
0
2
3
A
.4
5
5
ST R RIw
6
6
10
7
7
6
9
.
~LO
10
.il
11
C-
C~
,
4
OOO~O-O.-.o...~
~0 ~ .ch~o~-C
0*~~~
'p
cC
~
/
0
,/
1~)-
~>
iA
7
A
0
0
fi
.-
0
2.4
0~*
rt~'
/
0
~-
0
0
--- o~o.-o~0 ~ ~
0
I
0
I
0
I
S-4r
-~
~0
IJ
/
I,
11.1
z
iF-QT
I-Fop
N.CU-VJ-b
122:0,uj
-t
SL-
U
oa
I-.
0'
3
A
S
0/
6
7
e
9
IC
III
----
0
-c
0
*-
-o
;;
0
0
0~0
4f
0
0
/
/
0
a
p
a
0-~~~ ~
I
0
0
-c
0 ~
.- O
~
0
M1.0
0~~
op
0
A~
2.5
00
('7
4
0
0~~~ 0
-00~
~
0
1.c
/L
1.,5
TEsT
C1UV&~
NO.
I/-
3-
=%
Aeo
L 2i±~L~?I
2
~
L~r~L~k ~ie~ LkZ~'
FtURE.
0'-0
3
4
314,
1()
11
3.O'o
'o
03-
00
V
-
e~
-
A
~
A-
R.
4
A
S
//
/,0~~
0
/
~
A
~O
t/4
1.5
A
I
o
71ru-rT N~o. COU-V>-9
A
O
/cf M'
II~,
4,-.
-:-1
a.o
C)
I
c
a
3 A5
6
ST RN N '/
4o
7
8.
9
ID
3.o
0-0
/C
~~'
11
.~
I
0
/
A
3.0
2.0
6
ft
17'
.
TesT
No.
CIU-W- 3
TOP
20.5 20.4 I0.6
Urlibo15816'
C-'
0
1
2
3
4
5
STPAN
6
0/~
6
9
10
1
3.0
0~
~
\\
~ -
~
-
-10
2.00
0
-A
.
0...i--
i~2-
-
A-
-4
A
0
.oo--o0
*-Q~~,o
0
*
-
0
30
I
J
iv
U
TEST Nc. CU
L2 12.1n
1
2o.?
o.24 E6.9
1
FiGURpE 3.1
0
.1
2
3
4
5
STR'AIN
6
7
*/o~
8
9
1
11
-1-0
0-oO0-
0
1.00
1.2.
00
0~
OC
00
0
STP l\tl
FIGURE -3.i
3. 01
u
0-0-0
/0
0
0
0
I~2
S
0 0 - 0
A_
1.01
-
0.6
o.A
A
A
'4
"A
C
0
3.4
0
0 0-0
-
0-0
O
*
*
Th-sT No. CU21
0.8
66. 503.
OCR
12
0.60
-w.
225
LL
ib
3
4
5
2 "?2.6
1k6.? Tr.O
I1?.1
8
hGURE
9
3.19
-o
0
V
00")
6
00
0.4
0
A
IA
It,
It,
A
0
I
~
(
1.0
-
o
-0
0~~
/
0.0
0.8
/
0.5
0
0.3
/
TEST No. CU-W-12
0.
i
OCR=24
0.1
'TOP IkLDE ear.
Ls
A... A ..
~
-0.2
-- a
~A-~A
0
1
2
3
4
5
STRAIN
6
6.
1.4162
6
9
10
U- */o 320o
FIGURE
1
01*
NO
i
RE
5:
010
0
0
3.0'STEZ
3.T0
/0~G.
0o
/
TO
Ut
20,1
L:''-
CAiI
I!
I
Ct
0
/
~.
2
3
$
~
?
VF1-
1
0
Ir1'
o-~-f'~-~- ~
C
o
0
i
~
c'-~
-~
~
0~~-~~
~-
R
A
~A
~
A
A
-~
,6 '
0
~
00
.0--
('j
/
A
'Tes7Nc. CItJ-L-3-
ii
AGt1-,a £-,)ys
11
TOPI Mt oot Forrcm
II
LAftr+ .
0. 4
0'
c
I
a
.~CIA>
24.5 2S.2
0
.~
0~~~
'*-
0 _--0
/0
4f
A
V
A'
II
1/
/
A
/~
0
A
1!
0**
''
TEST 4o-CIU-L-3
NGE
'Ii
I
0.6
I
46 D
sa~CT cosTsNT 4.o "%
JS 1 24.1
24.2. 25.
LJ 5 17.2.
i'.
18.2
Ft6U uR E 3.a3
CA
i
2
3
4
6
STRMA ~ */o
2l1
3l
0
0
K
R.I
0
.0
00 0
A
Th-ST
NO.QlJLI
KG/
C-c- 3.0
P&G ED15 t>RY 5
SRLfC c0o4mFX:
T09*v~
I.S
0p
w4j
0A
GM/L
o~~
A.~.0~.
bh0
3.4
tu
00
Cx.4
FIGU RE 3.2L4
01
C
6
1-2"
I3
d-
.49
6o
87
9
--- 4
-00
0.0
0-
7
/
0
0'
/
0
I.C)
~0
/0
2.c,
-
a
~
~
A
0
-~
~
0
/
4
10
I
. (.)
I.A
I
1~'
7
i:V'
(
I
&
4
.ip-
TE.sT
/
I.
Sc,
N4o.
C~U-S4O
w
zr-,
TC.= 3.0 1,'C
,
S o.L I/.C
6.
I'
Ii
. r 25W ?
-~242
0.2
-FieGURF0
a5
6 7
S- IRR
l
)0
3.Z!S
9
.1i~ C)
3.0
~
0-,o
cr
-0-
-0---
--
0-0
10
-
t~
A
2A
/0
-0
C
0.0
.
0
~~-4 -A
A,4
~
7
A7
~.
/
Q
TE-sT wo. CIU-V)?7-
<3
TOP
LJSL 2E0
c
3
A
Sm
.5
6
3.0C I'/
V,e
3. 0
it
_
1.
2.
20
_
....
I
-
-
rv~~~ni~i~~i Li~~ti~
il
I~ Il~J'4
t
L
-.
-ii'{-I --~
--
Ki -t -
-t -------
1-
t
-
-T_--
-
S---
l-T
T...
H
T; - -
F
-
--
--
-~--
-
4.iy
-
-T--
~
-
-- ---- -
--
-K -.
rp:
-- - - --
1--
l
N4- itj
I
1 lilt!T ; Iill F II
i I
$
V
Mill
4
-,I-
O'
0
I
Cj
I I
MCIO
1"
I
t
I
-T-1
Su
ScL:T
dIJ
+Co.ze
5u
<s
t-I
Yf
(At-;
S-4
-
(
0
*
(b
s
SALT SALEs,
.
MV
.osoc
N
IS
0
S e
(a)
C
FKGu.E 3.26
WP
I~
i , F
I
N
(2
c~9
W:G
W '7
(I
(b)
(Al
V)RTP-PR
I
^C,, =
6
5AMVt'L:
h-CE
40
tI:~j
(CO)
"'
K
________
a
0
______________________
'a
3
F1GU~<E. 3~?9
NORMRLLY CONSDLIRTED
*1
£
U'
U'
,~Aw
Co
'-4
tu-s-2
4
N
/'
,..~
.9
FPFE-CT
NFY-,
RT
I.,
'N
0
ltI tj
ful
4.
6
+4Y-
VG/
~GUR~E 3.3'L
CF-PEC-TN)F- STRESS FKTl4(:;
CIU-L-1
clu-
Lu ,g4a ,
A-r
c
RM
-P
t Vcn-' .0
ri
FIGUtFrz-
SP~~t~
T V~L
C,
co
C'U-SCIO
olu-s- $
OCU-56
,CCR, - E
<---1---E
c
6:,tc?
FIGURE
:33.3
~WGT~E5ES5 PATHS
______________TE.
TO
G-0A
"'cA-
0
II:I w
olu
-w-1o
IOC)CR=r
clu-W-1?
Or-RZ a zk
0)
e
FIGURE 3- 3q
I
0.A
0.3
C,
lb
('3
H\/op -:sU-:-
PA~PU-Ilf
RS-
Ir MX.PPN10
"5
100
0. a
SALT SAMPLS:
0.3
+X
%
0.4
o
0
0
0
0
-~
4D
0.3
5
~
0
ckc~
ATMA.
"'t
STRIZLSS
&zl-Ci
L
SA~vPL2.S:
rO6
14O0
.0.1
-D.1
0.Z
as3
OA
FIGU~
0.5
3s3
o0000
Z0
co
(~-i~)
0
q)
X7.sL~P~rz
0.?.
'
LI. I
37
WR*.?R. SAMPLES:
;
OA
4.
.(*A t
0.13
0.
.
0a)
ci
(taI)
0.6
(L- 3)-,*
4)
0.4
&
*
0
0
0
0.3B
09
NFriRl$&1v
7~z,
RT MRX. PK if-ICA Pf L
ArE.
r :z 1I.-a"; crz 0.1 :
0.1
0
OA
0.Z.0.3
.4
0~
0.
0.9
1.0
1.1
1.2
~GyR~3.36
0.9
I
0. e
0.1 j
0.6
-+
0.
oA
:3
0.3
U)
0
0.11
-0.1
2
>
C.SWRER Sa\t
N-
-
-
O.C.
''
*
.s
3
n
5
4
sW \
P0.R
A
FT ERFA
5.iVZ
LUR
E
-0.0C.
cD
-
(L:-1)
C -
(L-2)
-
e-
(L-
5)
(L-4)
FIGURE 3.31
0
Top p ve
A
A
A
---- - -
3
- , it./
19
0 A
*
d
t(?)
~ 6
AI7
\ t>)
LE PIEG
W iFr ER
16
15
LT
6
-
0
0*
16
A
0
L'I V)
A
A
-A
A
3.4
~crrcA ?~c~
_
_
_
Vy
-W R
5E
-
LA
FmuL
U
rap
O\te-)tsyC,
F~uR~ 3.3~
A
CHAPTER IV:
4.1
DISCUSSION OF RESULTS
Effects of Salt Content on Atterberg Limits:
The properties of the silty-clay (MIT 1139) used in these
experiments are given in Table 3.1 on page a9. It can be seen that the
liquid limit is only little affected by the addition of 35 gm/l of NaCl
This may be due to several factors:
to the pore fluid.
1) a small amount of actual illite despite the 40% clay-size
fraction,
2) the influence of the initial amount of electrolyte in the pore
fluid.
The second factor seems to be the most likely cause since the soil
produces differential thermal analysis quite similar to other soils in
which the percent illite is known.
GREEN (1956) has presented the data shown in Table 4.1 for a batch
of Boston Blue Clay homoionized with sodium.
It was washed thoroughly
during preparation and contained a negligible amount of initial salt.
The data illustrate that the first few grams/liter of NaCl has a very
Thus, the salt content in the
prominent effect on the Atterberg Limits.
water samples (2-3 gm/l) before additional salt was added could have
precluded a more pronounced change in liquid limit.
TABLE 4.1
P.I.
salt (gm/1)
w
w
1
3*
33
45
22
22
11
23
5
49
22.5
26.5
20
36
53
52
24
25
29
27
from:
*interpolated
4.2
GREEN (1956)
Effects of Salt on Stress-Strain Relationships:
The salt samples and the water samples, at the same consolida-
tion pressure, have similar stress-strain plots until the maximum
principal stress difference is reached.
about the same value.
The initial tangent moduli have
After maximum principal stress difference has been
reached, the stress-strain curves for the salt samples and the water
samples differ somewhat with the principal stress difference for the salt
samples tending to remain about constant for large strains while the
principal stress difference for the water samples tends to fall off.
This.
is quite noticeable for samples at low consolidation pressures (compare
Fig. 3.2 with Fig. 3.10, Fig. 3.4 with Fig. 3.11) but less apparent at
higher consolidation pressures (compare Fig. 3.6 with Fig. 3.16). Much
of the difficulty in comparing the curves at high consolidation pressures
derives from the effects of strain-rate change.
Considering only the higher water content of a salt sample for a
given consolidation pressure, it would seem that the salt sample should
show greater tendency to lose shearing resistance after initial contacts
had been broken than would the water samples which have lower water
contents.
Apparently, the salt makes it
easier for the broken initial
contacts to "remake" and carry the shear stress without loss of strength.
The contacts in the water sample cannot remake as readily as in the salt
samples and some shearing resistance is lost-after the initial contact
breaks.
4.3
Effects of Salt on Shear Strength:
Fig. 3.27 which plots water content versus shear strength and
water content versus consolidation pressure, indicates the following:
1) at the same consolidation pressure, the salt sample and the
water sample have shear strengths that are very nearly
equal;
2) at the same water content, the salt sample is about twice
as strong as the water sample;
3) at the same water content, the consolidation pressure of the
salt sample is about twice as large as the consolidation
pressure for the water sample.
It must be emphasized that these observations are pertinent only to the
soil-water systems peculiar to this experiment and must not be taken out
of context to illustrate the effects of salt on shear strength.
Obviously,
the difference between the water content at failure of salt samples and
water samples at the same effective stress is a valid enough reason for
differences in the shear strengths of these two types of samples before
any consideration is given to the possible effect of salt concentration.
If an evaluation of the effect of salt alone on shear strength is to be
derived from the data, the variable of void ratio must be eliminated.
Use of Hvorslev Parameters at maximum principal stress difference would
seem an appropriate means for adjusting the data for water contents at
failure.
The fact that the salt samples have the same Su/W
value as the
water samples yet have a somewhat higher void ratio at failure indicates
that the salt clay would surely be stronger than the water clay if
the two
. This follows, of course,
c
from a consideration of the meaning of Hvorslev Parameters. In Section
samples had the same void ratio at the same
4.7, an estimate of the strength- the salt clay would have at the water
content of the water clay will be made using the Hvorslev Parameters
quoted in Fig. 3.36.
4.4
Effects of Salt on Effective Stress Parameters:
4.4.1
-
-
at Maximum Principal Stress Difference:
Fig. 3.28a and Fig. 3.29a provide the following relationships between shear stress and normal effective stress on the failure
plane:
= 0.50 T-f
T
Tff = 0. 45 f
(salt samples)
(water samples)
Since the undrained shear strength is the same for the salt clay
and for the water clay at the same consolidation pressure, the sizes of
the Mohr's circles at failure are equal for the two samples and the
increased obliquity for the salt sample means that the Mohr's circle is
closer to the origin.
This indicates a higher Af for the salt samples
than for the water samples at maximum principal stress difference.
This
would seem to indicate that there is somewhat more collapse of the initial
fabric in the salt sample than in the water sample during the first few
precent strain; the higher water content of the salt sample would seem
to explain this behavior.
The effect of the salt is apparently to offset
the effect on shear strength that might be expected to accompany this
collapse of fabric.
The broken initial contacts can remake readily
because of the suppressed double layers and carry a considerable portion
of the shear stress carried by the original bond.
This behavior will
later be contrasted to the behavior of leached samples where similar
initial fabric collapses in the presence of dispersive double layer
forces.
4.4.2-
-
at Maximum Principal Effective Stress Ratio:
The ji-angles of Fig. 3.30 and Fig. 3.31 provide the
following effective stress parameters for the failure plane if one
attributes the cohesion intercept to filter strips and piston friction:
= 30.5
Vff = 0.59
(salt)
ff
?ff = 0.53 T
u = 280
(water)
It follows directly from the expected effects of salt discussed in
Section 1.2 that the angle of maximum obliquity for the salt samples
should be greater than that of the water samples.
At the large strains
at maximum principal effective stress ratio the original bonds along a
failure plane will mostly all have been destroyed so that the shear
strength will be dependent upon the strength of remade bonds and the
.value of (R - T) in the remolded soil on the failure plane.
Double layers
are suppressed by salt and conseguently we expect that a larger portion
of the effective normal stress is carried by contact stresses than is the
case in water samples.
If it
is conceded that
(N -
A) will develop less
shear resistance than Tam for a given value of normal effective stress,
then the increased c
4.5
U of
the salt clay is explained.
Effects of Leaching out Salt after Consolidation:
It has been noted in Section 1.3 that previous investigators
have shown that the leaching process after consolidation is one of
essentially constant volume; that is, there is presumably little change
in fabric.
The water contents at failure of all salt samples and all
2
leached samples that were consolidated to 3.0 kg/cm are summarized in
Table 4.2.
The data show that the aged samples have slightly lower water
contents than the two samples failed soon after consolidation; but the
water contents of the leached samples are not lower than that of the aged,
unleached sample. These "data support previous conclusions that no
appreciable volume changes accompany leaching.
TABLE
4.2
EFFETS OF AGING ON WATER CONTENT
Age at
Test
Failure
(days)
S-4
S-9
L-1
L-2
3
6
56
42
25.1
25.0
24.8
24.9
28.9
29.8
28.9
28.7
L-3
S-lo
46
46
24.6
24.6
29.4
29.7
L-4
17
24.5
29.2
The effect of salt on an illitic clay is to increase the liquid
limit as has been discussed and illustrated in Sections 1.3 and 4.1.
This
will increase the liquidity index and reduce the remolded shear strength
as shown by SKEMPTON and NORTEEY (1952) and discussed in Section 1.3. The
sensitivity of the clay will be increased. It was suspected that it
should also reduce the undrained shear strength.
The change of environment caused by removal,of salt will leave the
fabric somewhat out of equilibrium with the electrical forces of the
A disturbance to the structure by application of shear
stresses should cause marked reorientation of soil particles. It would
be expected that the stress-strain curves and the effective stress paths
double layer.
should be quite different for leached and salt samples.
4.5.1
Effects on Shear Strength:
If salt were leached from an illitic
clay without fabric
change and with constant effective stress, the normal stresses at original
points of contact should be reduced. The ability of the clay to reform
contacts should be impaired.
Both of these effects should contribute to
a reduction of shear strength.
The maximum shear strengths versus water content at failure for the
*
w. is the water content when the sample was set up in the triaxial
cell prior to isotropic consolidation.
four leached samples tested during this experiment are plotted on Fig.
Three samples have noticably reduced shear strength; the fourth
sample fits the salt-curve guite well. None of the points for leached
samples fit the S vs. wf curve for water samples very well although the
3.27.
data for (L-4) fits the water-line better than it fits the salt-line.
If
the leached samples had the same salt content as the water samples, they
might be expected to matbch the curve for water samples better.
The salt content after leaching is undoubtedly an important factor in
explaining these data
(see Table 2.2 for a summary of salt content in
the leached samples).
The strongest sample (L-2) has the highest salt
content; the weakest sample (L-4) has the lowest salt content.
BJERRUM (1954) and SKEMPTON (1953) have proposed that the ratio
Su/ic can be correlated with plasticity index. Since GREEN (1956) shows
that the effect of salt on plasticity is confined largely to salt concentrations of less than 5 gm/liter, the importance of accurate salt determinations is apparent.
The distribution of- the salt in the triaxial sample
should also be of considerable importance.
If the shrinkage on drying
may be accepted as an indicator, the amount of structural breakdown that
occurred in the top third of the leached samples was appreciably greater
relative to the rest of the sample than was found to be the case for any
water or salt.samples.
This may suggest that the zone of failure in the
leached samples may have been shifted toward the top of the samples
because the salt concentration was less there.
Consequently, the salt
contents at failure as reported in Table 2.2, which are averages for the
entire sample, may not be accurate enough indicators of the salt content
in the failure zone.
4.5.2
Effects on Effective Stress Paramenters:
At Maximum Principal Effective Stress Difference:
The Mohr's circles at failure, ("
- 3 max. for the four leached
samples are plotted on Fig. 4.1 along with the o.-envelopes for salt
samples and water samples.
The 4( -envelopes are the obliquity at
maximum shear stress on the plane of maxiu=,Shea- stress as determined
on Fig. 3.28a and Fig. 3.29a.
The obliquities at failure of the leached
samples vary from 170 to 210; less than the obliquity of either salt or
water samples at failure.
This shows that the reduced shear strength
of leached samples cannot be accounted for by A
stress parameter,
*,
alone; the effective
is definitely affected by leaching out salt.
At Maximum Principal Effective Stress Ratio:
Effective stress paths for the four leached samples are plotted on
Fig. 3.32 along with the effective stress paths for a salt sample and for
2
a water sample consolidated to 3.0 kg/cm . The maximum obliquity
envelopes for the salt samples and for the water samples are also shown.
It can be seen that maximum principal effective stress ratios for the
leached samples fall on the maximum obliquity envelope of the water samples.
This is as expected.
At the large strains corresponding to maximum
obliquity, almost all of the initial bonds have been destroyed and the
shearing resistance that remains will be dependent solely on environment.
Since the leached clay is electrically equivalent to the water clay, it
would be expected to behave similarly after initial structure has been
destroyed.
The high pore pressures that develop in the leached samples are in
marked contrast to the behavior of the normal salt'samples at the same
void ratio and consolidation pressure.
Prior to attainment of maximum
principal stress difference the pore pressures in leached samples and salt
samples are virtually equivalent; after about l% strain, the leached
samples begin to develop greater pore pressures than salt samples.
The
higher pore pressures in leached samples at high strains are due to their
dispersive double layer forces.
As the initial structure breaks down
during shearing stresses, ability to remake contacts is hampered by
expanded double layers in the leached clay.
Clay particles tend to
reorient into a more parallel array and transfer normal stress into pore
pressure.
The A-factor for the leached samples increases rapidly after
maximum principal stress difference is reached and is still increasing
at.maximum obliquity.
The A-factors for the two weaker leached samples
(Fig. 3.23 & 3.24) increase almost linearly with strain.
4.6 Effects of Aging (consolidation during aging is permitted):
Since the leached samples would be considerably more aged after
consolidation when sheared than any of the salt or water samples, it
would be possible that aging might increase the shear strength of the
To get some idea of the effect of aging, a salt sample
2
(S-10) and a water sample (W-7) were consolidated to 3.0 kg/cm , aged 46
leached samples.
days, and then tested in undrained shear.
A. salt solution (35 gm NaCl/
liter) was used in the cell with the salt sample to avoid loss of salt
from the pore water by diffusion through the membrane during aging.
The effective stress path for the aged salt sample is plotted in
Fig. 4.2 along with the effective stress paths for the two non-aged salt
samples used to determine the effective stress parameters. There is no
appreciable difference in these stress paths. The shear strength of the
aged sample is virtually equal to that of the stronger non-aged sample.
However, the lower water content of the aged sample causes the data at
maximum shear stress to plot nicely on the Su vs. r'c plot of Fig. 3.27.
The Af at maximum principal stress difference is slightly less in the
aged sample but not enough to affect the effective stress parameters;
in fact, as seen in Fig. 3.28a, the data falls right on the z-envelope.
4.3
The effects of aging on a water sample is shown on Fig.
by the
It is seen that this aged sample
effective stress path for test (W-7).
has a considerably higher undrained strength and a lower A
non-aged water samples also plotted on Fig. 4.3.
than the two
The aged sample also
developed an appreciably higher maximum obliquity for a few percent strain
but subsequently dropped back to the normal failure envelope.
in obliquity occurred almost instantly.
This change
On Fig. 4.3, the last data above
the envelope and the next data on the envelope were taken consecutively
at 6.7% and 7.0%strain.
This marked decrease in obliquity and in effective
stress is shown very well by the stress-strain plot, Fig. 3.26.
It is
believed that this behavior was caused by formation of a failure plane.
The increased strength of the aged water sample is due to reduced
water content.
This sample had a water content of 22.0%, somewhat less
than either of the two non-aged samples.
on the Su vs.
Thus, the aged sample plots well
Tc line in Fig. 3.27.
Another water sample, (w-6), (Fig. 3.13) also consolidated to 3.0
kg/cm but not aged developed a shear strength equal to that of the aged
This sample, because of a leaky membrane, was consolidated,
water sample.
rebounded to (
=
0, and reconsolidated. As a result, its water content
was lower than that for the normal samples consolidated to 3.0 kg/cm but
On Fig. 3.27 and Table 3.3, the data
equal to that of the aged sample.
for (W-6) compares well with the data for the aged sample.
On Fig. 3.28a, it was shown that aging has had no effect on the
effective stress parameters at maximum principal stress difference for the
salt sample.
For the aged water sample, however, Fig. 3.29a shows that
the effective stress parameters at maximum principal stress difference
have been increased a noticeable amount by aging.
The difference in the effect of aging on the salt sample and the
water sanmle makes it difficult to surmise the effect on the leached
samples since the salt content of the leached samples was changing during
,he aging period.
The data for the aged samples indicate that it is unlikely
that the effective stress parameters of the leached samples are greatly
affected by aging. However, Fig. 3.27 shows that aging may contribute
a bit to the shear strength but the magnitude of the effect is likely to be
no larger than would be expected from the slightly reduced water content;
aging most likely does not diminish the departure of the data for the
leached samples from the S vs.
u
4.7
(
c
line.
Use of Hvorslev Parameters to Adjust Undrained Strength:
Ideally, one would like to have salt and water samples at the
same void ratios and consolidation pressures if the effect of salt alone
on shear strength is to be estimated.
The
i
vs. w
line for the salt
samples does not correspond to the 6'C vs. wf line for the water samples
(Fig. 3.27); due, probably, to having been consolidated from slurries of
different initial water contents.
It was believed that Hvorslev Parameters
might provide a means for making the adjustment in measured shear strength
necessary to separate the effects of salt from the effects of voi
ratio.
A shear strength equation in terms of Hvorslev Parameters for the
salt samples is:
ff
ff =
rK + T
r
0 '16
e
ff tan .re.f
e + 0.25
The cohesion component can be adjusted by using a value of rc corresponding to the water content of a water sample at the consolidation
pressure of the salt sample.
The frictiOn component should also be
adjusted since changes in the water content may affect
*
even though
(
that
r should be independent of water content.There is some evidence
will increase if the water content decreases at constant consoli-
dation presoure.
TABLE 4.3
Effect of Water Content on T
ff
Test
fc
lf
S-9
S-10
3.0
3.0
3.03
3.24
3f
ff
1.15
1.22
f
1.67
1.77
e
25.0
24.6
2.85
3.30
In Table 4.3, the normal stress on the failure plane at failure is
seen to have increased 0.10 kg/cm 2 for a reduction in water content of
0.4%.
For a decrease in water content (2.5%) corresponding to the adjust-
ment in cohesion that must be made, an appreciable increase in G-
For lack of adequate means to evaluate the effect of water content
result.
on
may
Gff
the effect will be ignored.
Table 4.4 adjusts only the cohesive
component of shear strength.
TABLE 4.4
Su of Salt Samples Adjusted for Water Content
Test
cGT
c
lf
3f
f
ew
0.16~
(rtan~
ew
ff
r
e
'
ff
u
S-1
2.0
2.31
0.89
1.28
4.5
0.72
0.32
1.04 1.16
s-4
3.0
3.18
1.16
1.70
7.5
1.20
0.42
1.62 1.80
S-9
S-2
3.0
4.0
3.03
3.92
1.15
1.50
1.67
2.17
7.5
9.0
1.20
1.43
0.42
0.54
1.62 1.80
1.97 2.19
S-3
6.0
5.87
2.38
3.39
13.0
2.08
0.85
2.93 3.39
- is the equivalent consolidation pressure for leached samples read
ew
line for water samples.
from the Wf vs.
*
"c
The S
vs.
c relationship for this soil would be:
Su = 0.15 + 0.46r c
and if the effective stress parameter, $,
at maximum principal stress
difference is not to exceed
Iu
(30.50), Af must be on the order of 0.5,
which would not be too unreasonable.
Since the leached samples have about the same salt content as the
water samples, it was thought that the Hvorslev Parameters for the water
clay might be used to predict the shear strength of the leached samples
by using the consolidation curve for. water samples to determine ie for
the leached samples.
on Fig. 3.36.
The data obtained by this assumption are plotted
is seen that the Hvorslev Parameters for the water
It
samples are not generally appropriate for the leached samples although
the quoted parameters for water samples give the strength of (L-3) and
(L-4) with good agreement as indicated in Table 4.5.
TABLE 4.5
ff of Leached Samples Computed from Hvorslev Parameters
f
Test
e
L-1
L-2
L-3
L-4
4.8
__ff
1.30
1.23
1.40
1.45
2.04
2.03
2.15
1.84
0.16r
tan
%
ff
0.42
0.41
0.44
0.37
r
e
0.21
0.20
0.22
0.23
computed
measured
ff
'ff
0.63
0.61
0.78
0.66
0.73
0.63
0.88
0.61
Distribution of Pore Water After Failure and Void Ratios
After Oven-Drying:
It is considered by some researchers that the gradients in
water content between the failure zone and the ends of a triaxial sample
often noticed after an undrained test are an indication of structural
reorientation during shear.
The clay particles will tend to realign
parallel to shear forces and induce a tendency to consolidate in the
failure zone (normally consolidated samples; the reverse in over-consolidated
samples).
Pore water will migrate away from the failure zone due to
excess pore pressure there and result in the water content near the center
of the sample being lower than the water content near the ends of the
triaxial sample.
This concept is currently subject to considerable debate.
Recent
measurements at MIT have shown that gradients in water content exist in
triaxial samples that have been merely consolidated in the triaxial cell
but not sheared.
NASIM (1961), using Vicksburg Backswamp~Clay, has shown
that the gradients in water content after failure in normally consolidated
samples can be largely explained by the gradients that existed prior to
It follows, then, that water migration must occur in overconsolidated samples to account for the fact that the distribution of
failure.
water in an overconsolidated sample after failure is different from the
distribution of water in a normally consolidated sample after failure;
the evidence indicates that water migrates toward the failure plane in an
overconsolidated sample during shear.
The triaxial samples of MIT 1139 used in this experiment were cut,
perpendicular to the centerline, into three parts of approximately equal
In Fig. 3.37 is
size and the water content of each piece determined.
plotted the difference between the water content of the center piece and
the average water content of the entire sample versus the consolidation
pressure.
The data suggest the following:
1) the distribution of water in the normally consolidated samples
is not much affected by consolidation pressure,
2) the distribution of water after failure in the over-consolidated
samples differs from the normally consolidated samples,
3) the leached samples show a more pronounced reduction of water
content in the failure zone than do either the salt or the water
samples.
(1) and (2) are in general agreement with the findings of NASIM (1961) and
suggest either that water migrates away from the failure plane in normally
consolidated samples, into the failure plane of an overconsolidated sample,
or both. Unfortunately, the distribution of water in a triaxially consolidated, unsheared, sample of MIT 1139 was not determined.
(3) suggests that even if there is no water migration in the normally
consolidated salt and water samples, there must be migration in the leached
samples.
The weakest leached samples show the greatest variation of water
content between the middle piece and the average (Fig. 3.37).
LAMBE (1958) has suggested that the shrinkage upon drying of a clay
specimen may be used as an indication of particle orientation; at the same
water content, parallel particles should show a larger reduction of void
space upon drying than random platelets.
Shrinkage limits after failure and oven-drying are plotted in Fig.
3.38.
This figure indicates the following:
1) dry void ratio is little affected by consolidation pressure,
2) dry void ratio of the center piece (failure zone) is less than
the dry void ratio of the top and bottom pieces,
3) dry void ratio for water samples is less than the dry void
ratio for salt sanples but the difference is less than the
difference in their water contents at failure; the salt samples
actually shrink more,
dry void ratios of the leached samples (initially, a "salt"
fabric) tends toward the value of dry void ratio of the water
samples.
These data suggest that the amount of fabric collapse and particle reorientation is greater in the leached samples; the leached samples shrink
to a noticeably lower void ratio upon-drying than do the salt samples at
the same water content.
AToM% yis M S-r?,FLSS
C Ils-24
CIUL-
3.0'
1,0
(5
a
.......
..
o
EFECN
~CIU-S-9
STRESS
ATH
( c qr. > Ays)
I- ciU-S-A
CoM~Ra1soce AcRG
Ruet
(V
1
to
C
0
A
a
a
0
r3
CIU-\,V- t7
clu-14- 2
clu-w- 8
(Rc-jEU) 46 t)RYc-;)
P-P-VzFC-rl\)F- -SMRU-,S FKTH5
WAI
:'
R
SAO PL.L S
COM-PRRV CN. Oiz- RGEA>
AWb k4CN-" G
SRMPU-S
CHAPTER V:
CONCLUSIONS
The CIU tests on the salt samples and the water samples lead to the
following conclusions regarding the effect of salt on the strength characteristics of the normally consolidated samples of MIT 1139 used in this
experiment:
1) Shear strength: at the same consolidation pressures and for the
void ratios encountered in these experiments, the water samples
and the salt samples are equally strong. If the salt samples and
the water samples had the same void ratio as well as the same
consolidation pressure, the salt sanples would surely be the
stronger.
2) Salt samples have greater maximum obliquity by about 2-1/20.
3) Salt samples exhibit higher effective stress parameters ('i) at
maximum principal stress difference by about 20. Since the salt
and water samples have the same Su/4c, Af in the salt -samples is
slightly greater. This behavior may be more rightly associated
with the differences in void ratio between the salt and water
samples and not to the 'difference in salt content.
The four tests on salt samples that were leached of their salt after
consolidation permit the following conclusions:
1) Shear strength: leaching out salt reduces the shear strength if
the salt content is lowered far enough - say to 5 gm/liter or less.
2) Leaching at constant consolidation pressure does not change the
void ratio.
3) The effective stress parameter ( ) at maximum principal stress
difference is definitely reduced by leaching. Increases in Af
are due to reduced shear strength, not to greater excess pore
pressures.
4)
) is the same whether the salt
The maximum obliquity (-/
is removed before or after coasolidation.
The effects of leaching (and, of salt, for that matter) are most
aptly summarized by comparison of the stress-strain data for two tests:
(L-3) and (S-10). These two tests had the same age at failure (46 days),
the same consolidation pressures (3.0 kg/cm 2 ), and the same void ratio at
failure (ef = 0.68). The salt concentration in the pore fluid of (S-10) was
35 gm/l; in (L-3), 4.0 gm/liter. The stress-strain data for these two
tests are superimposed in.Fig. 5.1. The following behavior can be observed:
80-
1) Both samples have virtually the same initial moduli of elastidity.
2) Pore pressures are equivalent in the two samples until about 0.81%
strain where the leached sample fails.
3)
Leaching has reduced the strain at failure from about 2 to less
than l1o.
4)
The leached sample develops only about 80% of the shear strength
of the salt sample.
5)
Subsequent to failure of the leached sample at 0.8% strain, the
leached sample develops greater pore pressures than the salt
sample.
6)
Remolding in the failure plane causes a considerable reduction of
shear strength in the leached sample; very little in the salt
sample.
7)
The leached sample never develops the obliquity attained by the
salt sample.
3.01
(A)
L
(~)
C6
-
-
,1~'
I
1.6
':2
1. 0
CoM
0.6
~.3
-I(-o
0.6
10.68
0.68
3.o
3.0
c; 46
46
rE
0.4
*.Z
0'
Leacvwes> 5AMPLE
AND A SALT SAMPLE (s-10)
PAR SON OF A
C
'
1
4.o
T
2
-L
**s
3
4
-
--- Ri .b
FGURE.
SLT
FiGuRE. 5.1
5
STRA\N w
6
8
9
ao
11
BIBLIOGRAPHY
1. ANDRESEN, et.al. (1957): "Triaxial Equipment Developed at the
Norwegian Geotechnical Institute", NGI Publ. #21.
2. BISHOP and HENKEL (1957): The Measurement of Soil Properties in
the Triaxial Test, Ed. Arnold, Ltd., London, 190 pp.
3. BJERRUM (1954): "Geotechnical Properties of Norwegian Marine Clays",
Norwegian Geotechnical Institute Publ. #4.
"Some Experiments With Artificially
Sedimented Clays", Geotechnique, Vol. 6.
4. BJERRUM and ROSENQVIST (1956):
5. GREEN (1956): "Effects of Pore Water Salt Concentration of Homoionic
Clays", M.I.T., S.M. Thesis, Dept. of Civil Engineering.
6. LADD (1961): "Physico-Chemical Analysis of the Shear Strength of
Saturated Clay", M.I.T., Sc. D. Thesis, Dept. of Civil Engineering.
"The Structure of Inorganic Soil", A.S.C.E.
Volume 79, Separate #315.
7. LAMBE (1953):
Proceedings
8.
LAMBE (1958): "Structure of Compacted Clay?, Journal of 8MFb, ASCE,
Proceedings, May 1958.
9.
LAMBE (1960): "A Mechanistic Picture of the Shear Strength of Clay",
Research Conference on the Shear Strength of Cohesive Soils,
ASCE, June, 1960.
10.
NASIM (1961): "Laboratory Investigation of the Effects of Rate of.
Strain on the Strength of a Saturated Soil", M.I.T., S.M. Thesis,
Dept. of Civil Engineering.
11.
SKEMPTON (1953): "The Colloidal 'Activity' of Clays", ICSMFE, #3,
Proceedings 1:57, Zurich, 1953.
12.
SKEMPTON and NORTHEY (1952):
Volume #3.
13.
WISSA (1961): "A Study of the Effects of Environmental Changes on
the Stress-Strain Properties of Kaolinite:, M.I.T., Dept. of
Civil Engineering, S.M. Thesis.
"The Sensitivity of Clays", Geotechnique,