ing=0 width=100% bgcolor=ccccff> « back to results for ""
Below is a cache of http://scholar.lib.vt.edu/theses/available/etd-11012000-08130056/unrestricted/06Chapter_3.pdf. It's a snapshot of the page taken as our search engine crawled the Web.
The web site itself may have changed. You can check the current page or check for previous versions at the Internet Archive. Yahoo! is not affiliated with the authors of this page or responsible for its content.
Chapter 3 Experimental Testing Program
50
Chapter 3
Experimental Testing Program

3.1
Introduction
The focus of this chapter is to describe the engineering properties of the sand used in
the study and to outline the equipment and testing program used in the investigative
analysis. The chapter begins with a summary of the index and strength testing properties
of the sand used in the study. Descriptions of the different cone penetrometers, vibratory
units, other equipment used in the testing are then presented. The calibration procedures
and corresponding calibration factors generated for each of the pieces of equipment that
had testing sensors is also included. A discussion of the problems encountered during the
testing regarding the influence of temperature on pore pressure measurements is also
presented. The chapter is concluded with the presentation of typical results of the
vibratory motion recorded during penetration tests using both the rotary turbine and
counter rotating mass vibrators.
3.2
Index Properties of Light Castle Sand
Light Castle sand was used for all of the calibration chamber testing performed as
part of this study, mainly due to the large supply readily available at the testing facility.
Light Castle sand was used in other research studies at Virginia Tech (e.g. Filz and
Duncan 1992; Porter 1998, Gomez et al. 1999). Light Castle sand has a grain shape and
distribution similar to Monterey 0 sand, which has been widely used in liquefaction
evaluation analyses (e.g. Silver et al. 1976). Light Castle sand is a poorly graded quartz
sand (SP) consisting predominantly of subangular to subrounded particles. Scanning
electron microscope photographs of a grab sample from the sand stockpile are shown in
Figure 3.1, while a grain size distribution curve is presented as Figure 3.2. Also included
in Figure 3.2 is the grain size distribution curve defined by Silver et al. (1976) for
Monterey 0 sand and the zone of liquefiable soils defined by Ishihara (1985). As shown
through this comparison, the Light Castle sand used in the study has a grain size John A. Bonita Chapter 3: Experimental Testing Program
51
distribution almost identical to that of Monterey 0 sand and falls within the zone of
liquefiable soils suggested by Ishihara (1985).
Specific gravity and maximum and minimum density tests were also performed as
part of the study. Table 3.1 is a summary of the index testing performed for Light Castle
sand. Also included in this table are the corresponding properties for Monterey 0/30 sand
given by Porter (1998). All of the index testing performed as part of the investigation was
in accordance with ASTM procedures.
3.3
Strength Testing of Light Castle Sand
3.3.1
ICU Triaxial Tests
Estimations of the strength parameters of the soil were obtained from information
available in the literature and from monotonic and cyclic triaxial testing performed in the
laboratory. Porter (1998) performed a series of monotonically-loaded isotropically
consolidated triaxial tests (ICU) on Light Castle sand to evaluate the effects of the testing
procedure, testing equipment, and applied correction factors on the undrained steady-state
shear strength. Some of the tests performed as part of his study were on lose samples
within the loose density range considered in this investigation, the results of which are
considered herein. Monotonically loaded ICU triaxial tests were performed on samples
air pluviated to medium dense densities as part of this study to determine the strength
parameters of the soil at the higher density state and further the information related to the
undrained steady-state strength generated by Porter (1998).
ICU triaxial tests were performed on medium dense samples as part of this
investigation. The samples were 71mm in diameter by 152mm tall and were formed
through an air pluviation technique. Each of the samples was pluviated into a latex
membrane that was positioned within a forming mold. A relative density of 55% was
used to simulate the average relative density present in the calibration chamber testing.
An end platen was then placed over the specimen top and the membrane was secured
over the platen, resulting in complete separation of the sample from the outer air surface.
The sample was then subjected to a vacuum of 12 kPa to keep the sample intact once the
mold was removed. The sample was then placed into the triaxial cell, which was sealed
and filled with water. The vacuum was removed and the sample was purged with CO
2
to
displace any air entrapped within the soil voids. The sample was subjected to a John A. Bonita Chapter 3: Experimental Testing Program
52
confining stress and then inundated with de-aired water. The confining stress and
backpressure were then increased until a B-value greater than 0.95 was achieved. The
sample was consolidated to the desired effective stress and then sheared at a constant
strain rate of 0.1%/min.
Presented in Figure 3.3a are the effective stress paths presented by Porter (1998) for
a series of ICU triaxial tests on samples at a relative density of 22%. Presented as Figure
3.3b are the test results determined through this study for samples at a relative density of
55%. The peak friction angles determined through this testing are 30.8
o
and 40.5
o
for the
loose and medium dense conditions, respectively, while the steady-state friction angles
were 29.1
o
and 28.9
o
. As shown in Figure 3.3, a significant post peak strength loss was
only observed in the medium dense samples at the intermediate and high stress levels,
suggesting that the flow liquefaction condition defined by Robertson and Fear (1997) was
not encountered in the loose samples using the ICU testing procedure.
Included as Figure 3.4 is the steady-state friction angle suggested by Porter (1998)
for Light Castle sand (i.e.

= 29.1
o
). Superimposed on this plot are the undrained
steady-state strengths determined through the ICU testing performed as part of this study.
As shown through the comparison, a good agreement exists between the test data from
the two different sources.
Figure 3.5 is the steady-state line formulated by Porter (1998) for Light Castle sand.
Also included in the plot are test data generated by Porter (1998) for loose samples and
data obtained from this investigation for medium dense samples. The position of the
medium dense samples relative to the steady-state line suggests that each of the samples
would tend to dilate during shear, which is further indicated by the stress path presented
in Figure 3.3. However, the magnitude of any reduction in pore water pressure observed
through the test data is considerably less than that suggested by the horizontal distance
from the steady-state line, particularly for samples tested at the low and intermediate
stress levels. Similar observations are noted for the contractive behavior of the loose
samples. Thus, the concept of the steady-state line did not provide a reliable estimation of
the volume change or pore pressure behavior during loading for the Light Castle sand. John A. Bonita Chapter 3: Experimental Testing Program
53
3.3.2
Cyclic Triaxial Testing
Cyclic triaxial tests were performed to evaluate the behavior of Light Castle sand
during dynamic loading. The triaxial system used in the testing is a completely
automated closed-loop feedback system that is controlled by a five-channel signal
processor. A full discussion of this system is presented in Polito (1999). Software run
through a personal computer was developed specifically for the testing equipment to
control both the signal processor and the data acquisition system. The software can be
used to run drained and undrained tests on monotonic or cyclically loaded samples. The
sample in any of these tests can be consolidated to either isotropic or anisotropic stress
states. The data acquisition system also automatically checks the B-value of the sample
during the backpressure saturation stage of the test and converts the recorded data to an
ASC