Actes du colloque - Volume 2 - page 537

1415
The effect of fines type on correlation between shear wave velocity and liquefaction
resistance of sand containing fines
L'effet du type amendes sur la corrélation entre la vitesse des ondes de cisaillement
et de résistance à la liquéfaction du sable contenant des amendes
Akbari-Paydar N.
Sharif University of Technology, Tehran, Iran
Ahmadi M.M.
Civil Engineering Department, Sharif University of Technology, Tehran, Iran
ABSTRACT: The use of shear wave velocity (V
s
) measurements as an in-situ test for evaluation of liquefaction potential has
increased substantially due to its advantages. Relatively large numbers of studies have been performed to establish the correlation
between V
s
and liquefaction resistance (CRR) of clean sands. Usually natural sands contain silt and/or clay and previous studies have
shown that both the amount of fines and their nature influence the CRR and V
s
as well. Therefore, the V
s
-CRR correlations may also
be affected by fines content and type of sandy soils, but these effects are less studied so far. Cyclic triaxial and bender element tests
were conducted on samples of sand containing various amounts of different fines. Using this experimental data, and a proposed semi
empirical method, CRR-V
s
correlation was developed and the effect of fines type on the correlation is investigated. Based on the
results, the CRR-V
s
correlation is affected by both the amount and plasticity of the fines present in the sand and this issue should be
considered in evaluating the liquefaction resistance of sands containing fines.
RÉSUMÉ : L'utilisation de la vitesse des ondes de cisaillement (V
s
) de mesure comme un test in-situ pour l'évaluation du potentiel de
liquéfaction a considérablement augmenté en raison de ses avantages. Un nombre relativement important d'études ont été réalisées afin
d'établir la corrélation entre V
s
et résistance à la liquéfaction (CRR) de sable propre. Habituellement les sables naturels contiennent
limon et/ou de l'argile et des études antérieures ont montré que le montant des amendes et leur nature influencent le CRR et aussi V
s
.
Par conséquent, les corrélations V
s
-CRR peut également être affectée par la teneur en fines et le type de sols sableux, mais ces effets
sont peu étudiés jusqu'à présent. Cyclique essais triaxiaux et bender éléments ont été effectuées sur des échantillons de sable contenant
diverses quantités d'amendes différentes. A partir de ces données expérimentales, et un projet de méthode semi-empirique, le CRR-V
s
corrélation a été développé et l'effet du type amendes sur la corrélation est étudiée. D'après les résultats, la corrélation CRR-V
s
est
affectée à la fois par la quantité et la plasticité des fines présentes dans le sable et cette question devrait être pris en compte dans
l'évaluation de la résistance à la liquéfaction des sables contenant des amendes.
KEYWORDS: Liquefaction resistance, Shear wave velocity, Fines type, Fines content
1 INTRODUCTION
Simplified procedure developed initially by Seed and Idriss in
1971 is widely used for evaluating the liquefaction resistance of
soils from in-situ tests such as standard penetration test (SPT),
cone penetration test (CPT) and shear wave velocity (V
s
)
measurements. Compared with other indexes, V
s
offers
geotechnical engineers a promising alternative and a
supplementary tool toward the penetration-based methods (SPT
or CPT) to evaluate liquefaction resistance of sandy soils
(Andrus and Stokoe 2000). In recent years, the use of V
s
for
evaluation of liquefaction potential has increased substantially
due to its advantages especially for liquefaction potential
microzonation. Relatively large numbers of studies have been
performed to establish the correlation between V
s
and cyclic
resistance ratio (CRR: which is the cyclic liquefaction resistance
normalized by initial overburden effective stress). So, different
CRR-V
s
correlation curves for separating liquefaction and
nonliquefaction occurrences have been proposed by different
researchers (Zhou and Chen, 2007; Andrus and Stokoe, 2000;
Tokimatsu and Uchida, 1990).
Previous studies have shown that both V
s
and liquefaction
resistance are affected by fines content (FC) and fines nature.
Therefore, the V
s
-CRR correlations may also be affected by
these parameters. Given that usually natural soils contains silt
and/or clay, the effect of FC and fines type of sandy soils on the
correlation between V
s
and CRR should be investigated, but
these effects are less studied so far. It is interesting to note that,
in the separated V
s
-CRR correlation curves for FC≤5% and
other FCs proposed by Andrus and Stokoe (2000), the fines type
is not considered. These well known curves are recommended
by NCEER and are used widely.
In this study, in order to clarify the effects of fines type on
V
s
-CRR correlations, laboratory measurements of V
s
using
bender elements and cyclic triaxial tests have been conducted
on clean silica sand and sand containing up to 15% of different
fines including non-plastic, low-plastic and highly plastic fines.
A simple method based on theoretical considerations and
laboratory data is presented and used to develop the CRR-V
s
correlation. Using the developed CRR-V
s
correlations, the
effect of fines type on the correlation is investigated. The results
are then compared with the procedure of Andrus and Stokoe
(2000).
2 LABORATORY TESTS AND RESULTS
2.1 Tested materials
Standard Firoozkooh No.161 sand was used as the host sand.
This sand is of crushed silica type with angular grains and is
commonly used in Iran as the standard sand in geotechnical
testing. The fine part of the soil consisted of three types:
Firoozkooh micronized powder from the same mine of the host
sand as the non-plastic fine (silt), Kaolin clay as the low-plastic
fine and Bentonite clay as the highly-plastic fine. The physical
properties of these materials are summarized in Table 1 together
with the corresponding grain size distribution curves in Figure
1. Soil specimens of clean sand and also sand-fines mixture
with 5 and 15% fines contents were considered in this study
(Table 2). These samples were tested with different void ratios.
1...,527,528,529,530,531,532,533,534,535,536 538,539,540,541,542,543,544,545,546,547,...913