1, 101-109
Development
C. M. R. TING*,
of KO in soft soils
G. C. SILLSt,
and D. C. WIJEYESEKERAf
laboratory tests.
INTRODUCTION
102
oedometer
wall. Davis & Poulos (1963) and
Lewin (1970) developed the controlled volume triaxial apparatus, in which the volume of the cell
fluid surrounding
the sample was maintained
constant
and hence it was presumed
that the
diameter of the soil sample remained unchanged.
Bishop (1958), Moore (1971) and Menzies, Sutton
& Davies (1977) used a conventional
triaxial
apparatus
with various lateral strain devices to
measure and regulate the cell pressure for zero
lateral strain condition.
Abdelhamid
& Krizek
(1976) used flush diaphragm transducers in a rigid
consolidometer
for lateral pressure measurement.
Since the publication of Jakys (1944) paper on
the theory of K,, many researchers
have suggested empirical or semi-empirical
correlations of
K, with the angle of shearing resistance 4 for
normally consolidated
soils. These relations are
summarized
in Table 1, which also gives the
values of K, calculated for 4 = 22, appropriate
for kaolin, for example. This Paper reports K,
measurements
made on very soft kaolin samples
in a consolidometer
under controlled
gradient
consolidation,
with measurement
of horizontal
stress by a transducer mounted flush in the wall
of the cell, and examines the results in the light of
these correlations.
EXPERIMENTAL
PROGRAMME
Controlled gradient consolidation
test
Table 1.
Summary of &-I$
remains
in a reasonably
uniform
condition
throughout
the consolidation
process, and a reliable estimate of K, can be obtained from measurements of the pore pressure distribution
and
the horizontal and vertical total stresses.
Consolidation
Transducers
K, equation
Reference
K, = 1 - sin 4
Jaky (1948)
K, = O-9 (1 ~ sin 4)
K, = tan
1 + sin f#~
45 -
cell
1,15(& - 9)
2
K, for kaolin
(& = 22)
O-63
(1957)
0.56
Jaky (1944)
Rowe (1957),
Abdelhamid
K, = 0.95 - sin 4
Brooker
& Ireland
K, = 1 - sin (1.2@)(OCR) I w)
Schmidt
(1967)
(1965)
0.59
0.58
0.56
DEVELOPMENT
OF K, IN SOFT SOILS
103
Piston total
stress transducer
Bleed
Horizontal
valve
Upper
drainage
port
total
Mid-plane
pore-pressure
transducer
Base total
stress transducer
3
Base pore-pressure
total and differential
Fig. 1. Consolidation
cell
Control system
The apparatus consists of a microcomputer,
the
autonomous
data acquisition
unit (ADU-ELE,
1984), the loading control system and the consolidation cell. The computer acts as a communicator with the ADU which records data, controls
the loading system and maintains the room temperature at 20C f 1C.
The loading system consists of an air pressure
actuator connected to an air-oil interface and a
hydraulic jack. The resolution of the air actuator
is 0.025% of the full-scale range of 840 kPa (i.e.
kO.21 kPa); the area ratio of the hydraulic jack
to the sample is 0.66 to 1. This gives an overall
loading resolution
of kO.14 kPa on the soil
sample, which was adequate
to maintain
the
lowest controlled gradient used in the tests. The
loading system is closed-looped
with the ADU,
which is programmed
to maintain the required
pore-water pressure gradient during the test. The
time interval for each looping and regulation was
set to 5 s.
104
Sample preparation
Kaolin was chosen as the soil for these tests
because of the large amount of reported research
on it, its homogeneity
when purchased
commercially, its convenience of preparation
and its
generally
low creep. The kaolin used in the
experimental work is a white powder marketed as
Speswhite
China clay, which is excavated
in
Cornwall. Two methods of sample preparation
were
adopted : remoulded
and
sedimented.
Samples were prepared with distilled water or sea
water; the latter was prepared by mixing goodquality sea salt (Tropical Marine) with distilled
water. The liquid limit and plastic limit are 58%
and 30% respectively in distilled water and 62%
and 36% with sea-water chemistry.
The remoulded
samples
were prepared
by
mixing dry kaolin with water at a water content
of 175%, more than twice the liquid limit. This
was wet enough
to allow thorough
mixing
without entrainment
of air into the soil. Sedimented samples were prepared by mixing 1.5 kg
of dry kaolin to form a dilute slurry of initial
density 1.055 g/cm. The slurry was then pumped
into a 2 m high sedimentation
column at whose
base the consolidation
cell was mounted. Sedimentation was carried out for a period of 3 days
before the careful removal of the consolidation
cell with the sample for the subsequent controlled
gradient consolidation
stage. The void ratios at
the end of sedimentation
for distilled and seawater samples were about 4.3 and 3.5 respectively.
preparation
technique (remoulded or sedimented)
and water chemistry (sea or distilled water). Tests
were carried out at one of three hydraulic gradients, i = 5, 10 or 50; some tests were repeated.
Table 2 summarizes
the test conditions.
For
clarity the results now given are for eight tests
only: the repeated ones are omitted. However, the
repeated tests are subsequently compared directly
with each other.
Stress-strain curves
The consolidation
curves e-log crVfor the distilled water remoulded
(KRD) and sedimented
(KSD) kaolin samples (Fig. 3) are almost identical, suggesting that the two fabrics are essentially
similar and no particle segregation has occurred
through sedimentation.
Also, there is no apparent
difference between the results of sea-water (KRS)
RESULTS
Test summary
Sample
Test
preparation
technique
Fluid
chemistry
KRS-CG-6
KRS-CG-7
KRS-CG-8*
Remoulded
Remoulded
Remoulded
Sea water
Sea water
Sea water
KRD-CG-10
KRD-CG-9*
KRD-CG-11
Remoulded
Remoulded
Remoulded
Distilled
Distilled
Distilled
KSS-CG-2
KSS-CG-14*
KSS-CG-8
Sedimented
Sedimented
Sedimented
Sea water
Sea water
Sea water
KSD-CG-6
KSD-CG-5
KSD-CG-1 I*
Sedimented
Sedimented
Sedimented
Distilled
Distilled
Distilled
* Repeated
test.
water
water
water
water
water
water
Hydraulic
gradient
1.34
5
50
50
1.29
1.27
50
10
10
1.32
1.32
1.31
5
5
50
1.14
1.16
1.16
10
50
50
1.33
1.32
1.33
DEVELOPMENT
OF
K,
IN
SOFT
105
SOILS
Top vertical
tots. stress
Base veriical
total stress
Horizontal
total stress
Percentage
settlement
60
40
Dlfferentlal
pore pressure
Time: h
K,
distilled
water remoulded
(KRD) samples.
This is in accordance
with the observations
of
Sides & Barden (1971), who reported difficulty in
flocculating large colloidally inert kaolinite.
However,
salt did affect flocculation
during
sedimentation.
For the entire range of effective
stress, Fig. 3 shows a considerable
difference
between the e-log eV results of sea water (KSS)
and distilled water (KSD) sedimented
samples.
The sedimented
sea-water samples have consistently lower compression
index C,, and lower
void ratio. The consistency of the test results for
the same water chemistry indicates good repeatability of the tests.
and
results
2.6-
2.4 -
Test
XI
0
Q.+
zx
0.6
5%
x
0
__
KRS-CG-6
KRS-CG-7
KRD-CG-10
KRD-CG-11
__
KSS-CG-2
fi
KSS-CG-6
__
KSD-CG-6
__
KSD-CG-5
fi
.
.
I1l111,
HO =
IC111,
100
10
a,: kPa
106
TING,
Range
of K,
+-
*:+-
SILLS
AND
WIJEYESEKERA
correlations
x KRS-CG-6
I
50
KRS-CG-7
KRD-CG-10
KRD-CG-11
KSS-CG-2
KSS-CG-6
KSD-CG-6
KSD-CG-5
I
I,
150
100
uv: kPa
+ 0.90
correlation
K, = -0.25e
coefficient
- 0.90
(1)
coefficient
-0.95
(2)
+ 0.87
correlation
for remoulded
and sedimented
kaolin respectively. If a realistic estimate of the lowest likely
values of void ratio e is taken, and it is assumed
that these equations will hold for the full range of
void ratios, they indicate that upper bounds on
K, of about 0.8 and 0.75 may exist for remoulded
and sedimented samples respectively. It must be
0.6
Ko = -0.22e
+ 0.90>.
6
Test
u 0.4
x KRS-CG-6
-F I
HTa
Ko = -0.25e
0 KRS-CG-7
+ KRD-CG-10
+ 0.67
0 KRD-CG-11
KSS-CG-2
A KSS-CG-6
.
KSD-CG-6
KSD-CG-5
I
2-4
I
2.2
I
2.0
I
l-6
I
1.6
e
I
1.4
I
1.2
I
1-O
DEVELOPMENT
OF
K,
IN
SOFT
107
SOILS
0.6 -
i-
*++
+f*
X.X
7
u
x
0.4 -
o-2
x+
lX
+*
KRD-CG-10
KRD-CG-9
lx
lX
l
x
l KSD-CG-5
x
KSD-CG-11
1.6
2.0
2.2
:X=xX&
I
1.2
1.4
1.6
160 < 0
I0.6 oy
0
u
0.4 -
Q @@@oxoo(
xx
m
2.0
1.6
xx$*Xxx
XX...
0.2
2.2
1.6
Xm
m
KRS-CG-7
KRS-CG-6
1.4
1.2
1
2.0
I
1.8
1.6
KSS-CG-2
KSS-CG-14
I
1.4
I
1.2
I
1.0
emphasized
that equations
(1) and (2) are the
results of the tests carried out in this research for
the specified stress range 5-150 kPa. Their applicability to lower (say ~5 kPa) or higher stresses
(> 150 kPa) must be investigated further.
DISCUSSION
400
m 300
4
b200 I
.
.
-iy;b,
~:, f ~
.
200
P = (~1 + 2os)/3:
400
kPa
600
108
TING.
SILLS
AND
thixotropic
effects to develop, it is clear that 4
values can be sufhciently high at low stresses to
explain the observed K, values within the previously proposed empirical correlation.
400-
WIJEYESEKERA
100
Fig. 8. Mohr
200
300
400
0,'.03'.kPa
500
600
I
700
tests on
kaolin
consolidation
can be identified with the condition
of yielding, and expansion of the yield envelope.
It is therefore not surprising to find some correof onelation between K,, a stress parameter
dimensional
normal
consolidation,
and
the
friction angle $, representing
the failure condition. Such a correlation is implicit in the empirical determinations
of K, given in Table 1.
Assuming the value 4 = 22 found from the triaxial tests for effective stress ranges of 200-600
kPa, the calculated values for remoulded kaolin
are shown in Table 1. The range is 0.5660.63. By
comparison,
the K, values measured in the consolidometer for remoulded kaolin mixed with distilled water lie within this range for effective
stresses above 70 kPa, and are therefore consistent with the original correlations.
Below about
60 kPa, however, the K, values are lower than
the empirical correlation
would suggest if the
value of 4 were constant independently
of stress
range. However, if 4 is assumed to be higher at
low effective stresses, the correlation may still be
valid. It is difficult to measure 4 at these stress
levels in the triaxial test, but Kamhawi
(1992)
reported experiments in direct shear under conditions of controlled
shear load increment
on
kaolin consolidated
from slurries of initial water
content about 400%. They found a failure angle
of 22.5 at vertical stresses > 10 kPa, with 38 for
~5
kPa
vertical
stress.
In displacementcontrolled
tests, on the other hand, they found
values of 22.5 at all stress levels. The high values
of 4 were attributed to thixotropy, developed in
the time
elapsed
between
the shear
load
increments,
and therefore not occurring in the
displacement-controlled
tests. If a value of
4 = 38 is taken, the corresponding
value of K,
calculated from (1 - sin 4) is 0.38. The consolidometer results for kaolin (both sedimented and
remoulded) shown in Fig. 4 lie between 0.33 and
0.44 for a vertical effective stress of 10 kPa. Thus,
although the conditions of the controlled gradient
test would not traditionally
be expected to cause
CONCLUSIONS
Experimental
observations
indicate that K, is
not a constant
parameter,
but increases nonlinearly with increasing effective stress, the lowest
values and the fastest changes of K, occurring at
low stress levels. The fact that the K, correlation
has been observed to vary linearly with void ratio
offers an intriguing possibility
of predicting
qS
values at lower effective stresses from the void
ratio+.ffective
stress relation for the soil. Thus, by
measurement
of K, at higher stress levels (say,
within an effective stress range of 100-500 kPa),
the appropriate
K,-e relation could be established. This could then be extrapolated
to higher
void ratios, corresponding
to softer soils. The
value of 4 could be calculated from K, = 1 - sin
effective
4: and related to the corresponding
stress level by way of the effective stress-void
ratio relation. The present data would support
this approach down to effective stress levels of
about 10 kPa.
Extrapolation
of the linear K,-e correlation in
the other direction, assuming a likely minimum
value of e, allows an upper bound to be placed on
K, and hence on 4. However, there is less justification for this from the experiments
reported
here, since none have been carried out to confirm
that the linear relation holds at effective stress
levels above 500 kPa. Comparison
of sea-water
remoulded and sedimented kaolin sample results
indicates a large effect of fabric on compressibility, with a much smaller effect on K,
The results of this experimental
programme
confirm that soft soils cannot, in general, be
treated simply by applying parameters
obtained
at higher stress levels. However, they also suggest
that the general relations still apply, provided
that appropriate parameters are used.
NOTATION
e
void ratio
cc compression index
HO initial height
4M
OCR
P
4
01) u3
0
hydraulic gradient
coefficient of lateral earth pressure at rest
critical stress ratio
overconsolidation
ratio
mean effective stress
deviatoric stress
principal effective stresses
vertical effective stress
shear stress
effective angle of shearing resistance
K, IN SOFT SOILS
REFERENCES
Abdelhamid,
M. S. & Krizek, R. J. (1976). At-rest lateral
earth pressure of a consolidating
clay. .I. Geotech.
Engng Div. Am. Sot. Civ. Engrs 102, GT7,721-738.
ADU-ELE
(1984). Autonomous data acquisition unit-a
new approach to data acquisition. Equipment
handbook. Engineering
Laboratory
Equipment
International.
Bishop, A. W. (1958). Test requirements
for measuring
the coefficient of earth pressure at rest. Proceedings
of conference on earth pressure problems, Brussels 1,
2-14.
Brooker, E. W. & Ireland, H. 0. (1965). Earth pressure
at rest related to stress history. Can. Geotech. J. 2,
1-15.
Davis, E. H. & Poulos, H. G. (1963). Triaxial testing
and three-dimensional
settlement analysis. Proc. 4th
Aus.-NZ ConjI Soil Mech., Adelaide, 233-243.
Edit, T. B. & Dhowian,
A. W. (1981). At rest lateral
pressure of peat soils. J. Geotech. Engng Div. Am.
Sot. Civ. Engrs 107, GT2,201-217.
Fraser, A. M. (1957). The influence of stress ratio on
compressibility and pore pressure coefficient in compacted soils. PhD thesis, University of London.
Jaky, J. (1944). The coefficient of earth pressure at rest.
J. Sot. Hung. Archit. Engrs 78, 355-358.
Jaky, J. (1948). Pressures in silos. Proc. 2nd Int. Conf.
Soil Mech., Rotterdam 1, 103-107.
Kamhawi,
K. Z. R. (1992). The shear strength of two
109