1416
Proceedings of the 18
th
International Conference on Soil Mechanics and Geotechnical Engineering, Paris 2013
Proceedings of the 18
th
International Conference on Soil Mechanics and Geotechnical Engineering, Paris 2013
Table 1. Physical properties of the mixtures constituents
Material
Unified
Classification
D
50
C
u
LL PI
G
s
Firoozkooh
sand
SP
0.23
1.32
-
-
2.65
Firoozkooh
silt
ML
0.02
-
26
2
2.66
Kaolin clay
CL
0.003
-
43
18
2.69
Bentonite
clay
CH
-
-
160 116 2.75
#40
#100
#200
0
10
20
30
40
50
60
70
80
90
100
0.001
0.01
0.1
1
10
Passing (%)
Grain Size (mm)
Firoozkooh Sand
Firoozkooh Silt
Kaolin Clay
Bentonite Clay
Figure 1. Grain size distribution curves of the mixtures constituents
2.2 Cyclic triaxial tests
The cyclic resistance of the sands tested was determined using
undrained stress-controlled cyclic triaxial tests performed on
reconstituted specimens according to ASTM D5311 standard
testing procedure. The tested specimens were 70 mm in
diameter and 140 mm in height.
In view of the diversity of the specimen reconstitution
techniques (dry and water pluviation, moist tamping, slurry
deposition, etc.), obtaining homogeneous samples in terms of
void ratio distribution and consistency as well as covering a
wide range of void ratios are the basic requirements. Huang et
al. (2004) showed that specimen preparation method does not
affect the liquefaction resistance-V
s
correlations; moreover, it
gives the widest range in void ratio among others (Ishihara,
1993). Therefore, moist tamping method of sample
reconstitution was utilized to prepare the samples in the present
study. In order to obtain a uniform density, the specimens were
made in seven layers and the under-compaction method was
used.
To facilitate the saturation process, carbon dioxide (CO
2
)
was first passed through the samples. Subsequently deaired
water was allowed to flow in the specimens. Samples were then
saturated by applying proper back pressure in successive steps.
Samples were considered to be saturated if Skempton pore
pressure parameter (B) value was greater than 0.95.
Saturated samples were then consistently consolidated
uniformly in steps of 10 to 30 kPa. The consolidation process
continued until the effective confining stress reached a value of
200 kPa. The void ratio of the samples after consolidation was
determined by accurately measuring the moisture content at the
end of the experiment.
At the end of the consolidation process, a sinusoidal loading
with frequency of 1 Hz was applied to the sample having a
specified cyclic stress ratio (CSR: which is the ratio of cyclic
deviator stress to twice the initial consolidation stress). At least
3 cyclic tests were performed to obtain the cyclic resistance of a
soil sample having a specified void ratio. All parameters except
CSR were kept constant in these tests. The cyclic resistance
(CRR
tx
) is defined as the applied CSR required reaching 5%
double amplitude strain in 15 loading cycles (representing an
earthquake magnitude of 7.5). Generally, 110 cyclic triaxial
tests have been conducted in this study on 7 different
combinations of sand and fines with different void ratios.
Tests results, in the form of CRR
tx
versus void ratio (e) for
tested material are shown in Figure 2. As expected, the
liquefaction resistance decreases as the void ratio increases. A
power curve with the following expression can be fitted to these
points for each soil.
e.
CRR
tx
(1)
where
and
are constants for a given material and can be
obtained by fitting the obtained results;
these values are
presented in Table 2.
0.1
0.2
0.3
0.4
0.5
0.45
0.55
0.65
0.75
0.85
0.95
1.05
CRR
tx
e (Void Ratio)
F0-0
FS-5
FS-15
FK-5
FK-15
FB-5
FB-15
Power (F0-0)
Power (FS-5)
Power (FS-15)
Power (FK-5)
Power (FK-15)
Power (FB-5)
Power (FB-15)
Figure 2. The CRR
tx
versus void ratio for tested materials
2.3 Bender elements tests
In order to measure the V
s
and CRR
tx
on a single sample, the
bender elements were installed in a cyclic triaxial apparatus at
the top and bottom pedestal of the triaxial cell.
Immediately after the end of each consolidation stage
(ranging from 30 to 200 kPa), V
s
was measured using bender
elements. V
s
can be obtained from measuring the travel time
from the source to the receiver (t) and dividing the sample
length (L) to it. The value of “L” is assumed the tip-to-tip
distance of the bender elements (Lee and Santamarina, 2005). In
order to obtain the “t” value, the method of first arrival time was
used. First arrival time refers to the time interval between the
start of the source signal and the start of the major cycle of the
received signal by ignoring the initial portion of the weak signal
(Lee and Santamarina, 2005). In all the conducted bender
element tests, a single sinusoidal pulse having a frequency of 5
kHz and amplitude of ±10 V was used as the transmitted signal.
Sample result of a bender element test is represented in Figure 3
in which the first arrival time is shown.
The void ratio as well as the height of the samples changes in
each consolidation stage as the confinement stress increases. To
calculate the changes in the void ratio, the amount of water
expelled from the specimen during consolidation stage was
measured. Also, the water content of the samples was measured
carefully at the end of the experiment. As the sample is already
saturated prior to the consolidation phase, the void ratios at the
earlier stages of consolidation can be back-calculated from these
measured values. The settlement of the sample was also
measured during the saturation and consolidation phase and the
change in the height of the samples was accordingly used in
calculating the V
s
. Thus, from the bender element tests
performed on a certain sample, for different void ratios and
confinement effective stresses at successive stages of
consolidation, the shear wave velocity is conveniently achieved.
A total number of 1220 bender element tests were carried out on
110 different samples.
