333
Proceedings of the 18
th
International Conference on Soil Mechanics and Geotechnical Engineering, Paris 2013
1
Tensile Strength of Lightly Cemented Sand through Indentation Tests
Résistance à la traction de sable légèrement cimenté par des tests d'indentation
Ge L.
Department of Civil Engineering, National Taiwan University
Yang K.-H.
Department of Construction Engineering, National Taiwan University of Science and Technology
ABSTRACT: Compared to the compressive or shear strength of soil, its tensile strength is generally assumed to be zero, or
insignificant, in geotechnical engineering practice because of its relatively small value and lack of a satisfying laboratory technique.
The tensile strength of soil is, however, an important parameter in the design of geosystems, where tensile cracks contribute to
progressive erosion or landslides in excavation, slopes, dams, highway embankments, riverbanks, hydraulic barriers, and other earth
structures. In this paper, a recent development of the indentation test is presented. The theoretical framework is re-evaluated followed
by a series of test results on lightly cemented sand.
RÉSUMÉ : Par rapport à la résistance à la compression ou de cisaillement du sol, sa résistance à la traction est généralement
considérée comme nulle ou insignifiante dans la pratique de la géotechnique en raison de sa valeur relativement faible et le manque
d’une technique de laboratoire satisfaisante. La résistance à la traction de sol est, cependant, un paramètre important dans la
conception des géosystèmes, où des fissures de traction contribuent à l’érosion progressive des glissements de terrain dans
l’excavation, les pistes, les barrages routiers, des remblais, des rives, des barrières hydrauliques, et d’autres structures terrestres. Dans
cet article, un développement récent de l’essai d’indentation est présenté. Le cadre théorique est réévalué, suivi d’une série de
résultats d’essais sur le sable légèrement cimenté.
KEYWORDS: tensile strength, indentation, limit analysis.
1 INTRODUCTION
Determination of tensile strength of soil can be categorized to
direct and indirect methods. In the direct method, the soil
specimen is directly pulled apart and assumed to split in the
middle (e.g. Das and Dass 1995, Tang and Graham 2000,
Nahlawi et al. 2004, Tamrakar et al. 2007, Zeh and Witt 2007).
Its tensile strength is computed as the measured maximum pull-
up force divided by the cross sectional area. However, it is
normally quite challenging to have the soil specimen clamped
or glued in the split mold. In the indirect method, tensile failure
is induced by compressive load. A well-known Brazilian test is
of such kind. The indentation test, also called unconfined
penetration tests (Fang and Chen 1972, Fang and Fernandez
1981, and Kim et al. 2012) is an indirect method for tensile
strength determination. It uses a pair of cylindrical metal
indenters, or punches, to compress a cylindrical soil specimen
as shown in Figure 1. The tensile strength is computed through
the equation developed from the limit analysis (e.g. Chen 1975).
The test gives an applied compressive axial load and indenter
displacement curve. The maximum load is identified for further
data reduction to computer the tensile strength. The limitation
of this test method is that a certain level of brittleness of the
specimen is required so that a split tension failure would occur.
2 BACKGROUND
As described in Chen (1975), an ideal failure mechanism
developed in the soil specimen can be presented in Figure 1. By
equating the external work with the internal work, an upper
bound solution can be obtained as follows.
(1)
p
u
is the upper bound axial compressive load that causes the
split tension failure.
a
is the radius of the indenter.
b
is the
radius of the soil specimen.
is the internal frictional angle of
the soil.
is the developed angle underneath the indenter when
failure occurs.
q
u
is the unconfined compressive strength while
q
t
is the tensile strength to be determined.
The upper bound solution has a minimum value when
p
u
/
= 0, where
(2)
By equating (1) and (2), the equation below can be obtained,
(3)
Substituting (3) into (1) and let
p
is the maximum
compressive load applied from the test, which is less than the
upper bound load,
p
u
. Therefore,
(4)
(5)
where
K
= tan(2
+
)
The value of
K
is challenging to determine in laboratory
testing as it is a function of the internal frictional angle of the
soil, the unconfined compressive-tensile strength ratio, as well
as the size of the indenter. The recommended values of
K
based
t
u
u
q
bH
q
p
) co
n(
n
t
u
q
p
2
a
bH
a
1)
tan(
2
2
2/1
2
sin
2
sin 1
cos
1 sec
tan
cot
t
u
q
q
a
bH
t
u
qa
2
2 ta
bH p p
n
2
a
bHK
p
q
t
a
a
t
)(
ta
2)
cos(
sin2
si 1
2
2
