1459
Challenges to the laboratory evaluation of field liquefaction resistance
Les défis de l’évaluation en laboratoire de la résistance à la liquéfaction de terrain
Coelho P.A.L.F., Azeiteiro R.J.N., Marques V.D.
University of Coimbra, Coimbra, Portugal
Santos L.M.A.
Coimbra Institute of Engineering, Coimbra, Portugal
Taborda D.M.G.
Imperial College London, London, UK
ABSTRACT: Soil liquefaction is one of the most feared phenomena in Geotechnical Earthquake Engineering, due to the serious
damage it can cause to modern societies in seismically active regions. Even if significant progress has been achieved in the laboratory
evaluation of liquefaction resistance of sandy soils, considerable challenges still remain. These mostly result from the complex and
variable nature of earthquakes, which apply non-uniform and multidirectional cyclic loading to soils and is therefore difficult to
reproduce in laboratory testing. This paper aims at discussing the impact of the loading conditions imposed during soil testing. In
particular, the effects of using uniform, axial and unidirectional cyclic loading are considered. The results show that the existence of a
singular peak load of larger amplitude considerably affects the number of cycles required to liquefy the soil, with the location of that
peak being of upmost importance for the evaluation of liquefaction resistance. Also, it was found that the loading mode affects the
cyclic response of the soil, with fewer cycles being necessary to liquefy the soil under radial cyclic loading. Moreover, the
liquefaction resistance of sand is substantially reduced when multidirectional loading is used.
RÉSUMÉ: La liquéfaction de sols est un des phénomènes les plus redoutés en Génie Parasismique Géotechnique à cause des graves
dommages qu’il peut causer aux sociétés modernes dans des régions sismiquement actives. Même si d’importants progrès ont été
réalisés dans l’évaluation en laboratoire de la résistance à la liquéfaction de sols sablonneux, des défis considérables demeurent.
Ceux-ci adviennent principalement de la nature complexe et variable des tremblements de terre, lesquels appliquent des chargements
cycliques non uniformes et multidirectionnels sur les sols qui sont très difficiles à estimer à partir des données de laboratoire. Cet
article vise à présenter l’impact des conditions de chargements appliqués durant des essais sur sols, en considérant en particulier les
effets de l’usage de chargements cycliques uniformes, axiaux et unidirectionnels. Les résultats démontrent que l’existence d’un pic de
charge singulier de plus grande amplitude affecte considérablement le numéro de cycles nécessaires à la liquéfaction d’un sol, ainsi
comme son emplacement est d’extrême importance pour l’évaluation da résistance à la liquéfaction. En outre, il a été constaté que le
mode de chargement affecte aussi la réponse cyclique des sols, en étant nécessaire moins de cycles pour liquéfier un sol sous un
chargement cyclique radial. De plus, la résistance à la liquéfaction de sables est considérablement réduite lors de chargements
multidirectionnels.
KEYWORDS: liquefaction; non-uniform cyclic loading; loading mode; multidirectional cyclic loading.
1 INTRODUCTION
The shear stresses induced on an element of soil in level ground
during earthquakes are mainly irregular and multidirectional.
Due to the difficulty in the application of such complex patterns
of stress changes in laboratory testing, it has been current
practice to use uniform and unidirectional loading conditions in
laboratory testing. The cyclic strength of soil obtained under
such conditions is then corrected in order to account for the
effect of load irregularity and multidirectionality (e.g. Ishihara
& Nagase 1988).
In this paper, laboratory test data obtained with a stress path
cell is used to point out the consequences of using different
loading conditions in the assessment of liquefaction resistance
of a sandy soil. In particular, it is shown that the occurrence of
singular peaks in uniform cyclic loading, as well as using
different loading modes, clearly affect the cyclic response of the
soil. This suggests that some challenges concerning laboratory
evaluation of liquefaction resistance still remain.
2 IRREGULAR LOADING
2.1
Motivation
In general, the evaluation of liquefaction potential of an element
of soil is based on the comparison of the cyclic shear stresses
induced in the field by the earthquake motion with the shear
stresses required to liquefy representative samples in the
laboratory. This comparison is hampered by the erratic nature of
seismic loading, varying, in general, in amplitude in each cycle.
The current methodology adopted to overcome this problem
consists of converting the random stress history of earthquake
motion into an equivalent uniform cyclic stress history, which is
expected to cause a similar effect on the soil (e.g., Seed et al.
1975). According to this methodology, the damage induced in
the material is proportional to the stress level and independent
of the location of the stress cycle in the loading history (Shen et
al. 1978).
In order to investigate the effect of the largest stress cycle on
the resistance of soil, as well as the importance of its location in
the overall load pattern, undrained cyclic triaxial tests having a
singular loading cycle with much larger amplitude were carried
out. The amplitude of the singular peak load and its location in
the time history of the uniform loading were defined in
accordance with the analyses of real seismic records (Azeiteiro
et al. 2012). Namely, the results of those analyses suggest that
largest peaks usually occur between 0.1 and 0.7 of the total
number of representative cycles, close to the middle cycle of the
earthquake. Furthermore, the authors concluded that maximum
and average accelerations of a given earthquake are
approximately directly proportional. Thus, a singular peak with
double amplitude of the remaining uniform loading was
