1535
Normalized Shear Modulus of Compacted Gravel
Module de cisaillement normalisé des graviers compactés
Liao T., Massoudi N., McHood M.
Bechtel Power Corporation, Frederick, MD, USA
Stokoe K.H., Jung M.J., Menq F.-Y.
University of Texas, Austin, TX, USA
ABSTRACT: Compacted gravel is often used as engineered fill to provide the needed bearing capacity for structures. The dynamic
properties of the gravel fill, such as nonlinear shear modulus, are required in seismic analyses to evaluate the response to dynamic
loading. From a series of Resonant Column and Torsional Shear (RCTS) tests on two types of crushed gravel fill, normalized shear
modulus reduction curves were obtained as a function of cyclic shear strain. These curves are presented and compared to empirical
relationships in the literature that have been proposed for gravelly soils.
RÉSUMÉ: Le gravier compacté est souvent utilisé comme remplissage pour fournir la capacité portante nécessaire aux structures. Les
propriétés dynamiques du remblai de gravier, tels que le module non linéaire de cisaillement, sont requises dans les analyses
sismiques pour évaluer la réponse à un chargement dynamique. A partir d'une série d’essais à la colonne de résonance et d’essais de
cisaillement en torsion sur deux types de gravier concassé de remplissage, les courbes d’évolution du module de cisaillement
normalisé ont été obtenues en fonction de la contrainte de cisaillement cyclique. Ces courbes sont présentées et comparées à des
relations empiriques provenant de la littérature qui ont été proposées pour les sols graveleux.
KEYWORDS: Shear Modulus, Resonant Column Test, Torsional Shear Test, Fill, Gravel.
1 INTRODUCTION
Compacted gravel is frequently used as engineered fill beneath
the foundations of important structures, such as nuclear power
facilities and high-rise buildings. To evaluate the seismic
response of the ground supporting these structures, the dynamic
properties of the gravel fill (i.e., shear modulus G and material
damping ratio D) need to be determined. Due to the limited
paper length, only normalized shear modulus is discussed
herein.
Although the small-strain shear modulus (G
max
) can be
determined under in-situ conditions from shear wave velocity
(V
s
) measured in the field, it is very difficult to obtain strain-
dependent curves of G and D directly from in-situ tests (Ishihara
1996). In current engineering practice, the effects of confining
pressure (
0
) and shear strain (
) on G and D are primarily
evaluated through laboratory tests, such as cyclic triaxial
(CTX), cyclic simple shear (CSS), cyclic torsional shear (TS),
and resonant-column (RC) tests. These tests not only give the
values of G and D at small strain, but also yield the variation of
G and D with
and
0
. However, such tests are rarely
performed on gravels, due to the large size of the testing
apparatus required to test representative specimens.
Additionally, because it is difficult to obtain undisturbed
samples of gravelly soils, such tests on natural gravelly soil
deposits have been mainly limited to high-quality undisturbed
gravel samples obtained by in-situ ground freezing (e.g., Goto et
al. 1992, 1994; Hatanaka and Uchida 1994; Kokusho and
Tanaka 1994). Comparison of test results between undisturbed
and reconstituted specimens show that the effect of sample
disturbance is significant on G, but most researchers believe that
it is quite small on G/G
max
~
curves (e.g., Hatanaka and
Uchida 1994; Rollins et al. 1998), although Kokusho and
Tanaka (1994) indicate that undisturbed specimens exhibit
greater degradation at relatively small strain levels.
2 LITERATURE REVIEW
By removing (or scalping) particles of size greater than 51 mm
in diameter, Seed et al. (1986) performed a series of cyclic
triaxial tests on 305-mm diameter specimens of four different
types of well-graded gravels, which were isotropically-
consolidated and tested under undrained cyclic loading
conditions. Based on the test results, the G/G
max
~ log(
) curves
of gravelly soils are suggested to be in the range shown in
Figure 1.
0
0.2
0.4
0.6
0.8
1
0.0001
0.001
0.01
0.1
1
Cyclic Shear Strain,
(%)
G/G
max
Average for Gravel
(Seed et al., 1986)
Variation Range for
Gravel (Seed et al., 1986)
Variation Range for Gravel
(Rollins et al., 1998)
Average for Gravel
(Rollins et al., 1998)
Figure 1. Typical G/G
max
~
relationships for gravels.
By analyzing the test results for gravelly soils mainly found
in literature, Rollins et al. (1998) also suggested a slightly
different range for the G/G
max
~ log(
curves for gravelly soils
(Figure 1). Most of the data used by Rollins et al. (1998) came
from cyclic triaxial tests (CTS) typically performed on
specimen of 300 mm in diameter and 600 mm in height, and a
small portion of the tests are cyclic torsional simple shear tests
performed on specimens of larger diameters. All these tests
