3475
Proceedings of the 18
th
International Conference on Soil Mechanics and Geotechnical Engineering, Paris 2013
1
Probabilistic Assessment of the Bearing Capacity of Shallow Strip Footings on Stiff-
Over-Soft Clay
Evaluation probabiliste de la capacité portante de semelles filantes peu profondes sur couche
d’argile rigide
recouvrant une couche
d’argile mo
lle
Y. Tian & M.J. Cassidy
Centre for Offshore Foundation Systems, UWA Oceans Institute & ARC CoE for Geotechnical Science and Engineering,
University of Western Australia, Perth, Australia
M. Uzielli
Georisk Engineering S.r.l., Florence, Italy
ABSTRACT: This paper focuses on the probabilistic assessment of the resistance factor for bearing capacity for a strip footing on a
stiff-over-soft clay profile. The analysis is performed by applying the Random Finite Element Method, which combines finite element
simulation, spatial variability analysis and Monte Carlo simulation. Finite-element analyses are performed in the program ABAQUS
on meshes in which undrained strength values are assigned on the basis of quantitative estimates of the vertical and horizontal spatial
variability and the probabilistically modelled scatter of undrained strength itself. The stochastic implementation of the numerical
analyses results in samples of bearing capacity factors which, when normalized by a deterministic bearing capacity factor, provide a
set of tabulated factors calibrated to user-defined target reliability levels. The results have application for the prediction of foundation
punch-through, where the footing pushes the upper strong layer of soil into the softer clay beneath.
RÉSUMÉ :
Cet article porte sur l’évaluation probabiliste du facteur de résistance de la capacité portante d’une semelle filante posée
sur une couche d’argile rigide recouvrant une couche d’argile molle. L’analyse applique la méthode des éléments finis
aléatoires, qui
associe simulation par éléments finis, analyse de variabilité spatiale et simulation par la méthode de Monte Carlo. Les analyses par
éléments finis sont réalisées avec le programme ABAQUS, en utilisant des maillages pour lesquels la valeur de résistance au
cisaillement non drainée est déterminée sur la base de l’estimation quantitative des variabilités verticale et horizontale,
ainsi que sur
la dispersion de la résistance au cisaillement non drainée modélisée de façon probabiliste. L’implém
entation stochastique des analyses
numériques conduit à des facteurs de capacité portante qui, lorsqu’on les normalise par le facteur de capacité portante déter
ministe,
fournit un ensemble de facteurs étalonnés pour des niveaux de fiabilité prédéfinis. Les résultats obtenus trouvent une application pour
la prédiction du poinçonnement des fondations, dans le cas où la fondation pousse la couche supérieure de sol dur dans la couche de
sol mou située en dessous.
KEYWORDS: Bearing capacity, shallow foundation, spatial variability, probability, random finite element analysis, strip footing
1 INTRODUCTION
The bearing capacity of a shallow foundation on two layered
clay soil is a classical problem in soil mechanics and one of
importance to many applications, including the punch-through
of offshore foundations. The problem being analyzed in this
paper is defined in Figure 1: what is the vertical load carrying
capacity of a strip footing of width (
B
) on a top layer of soil of
undrained shear strength (
s
ut
) overlying a weaker bottom layer
(
s
ub
). Using finite element analysis in conjunction with limit
theorems, Merifield et al. (1999) published extensive bearing
capacity factors (
N
c
*) defined to predict the vertical capacity as
a function of the strip footing width and the undrained shear
strength of the top clay layer. Conditions of varying top layer
thickness and shear strength ratio were analyzed. However,
these solutions were only provided for deterministic properties
of soil with no spatial variability accounted for.
This paper makes use of the Random Finite Element Method
(RFEM) (see Fenton and Griffith, 2008) to investigate the effect
of the spatial variability in undrained shear strength on the
bearing capacity of a shallow strip footing on two-layered stiff-
over-soft clay. In the RFEM the characterization of the spatial
variability enables the generation of random fields with spatially
varying values, all of which are mapped onto a finite element
mesh. The generation of multiple random fields associated with
the soil domain allows the repeated implementation of finite
element analysis, yielding multiple samples of outputs. These
can subsequently be analysed statistically.
Although RFEM has been used to estimate the statistical
distribution of the vertical undrained bearing capacity of a strip
foundation on a single layer (Paice et al. 1996, Nobahar and
Popescu 2000, Griffiths and Fenton 2001, Griffiths et al. 2002,
Fenton and Griffiths 2003, Popescu et al. 2005, Kasama and
Whittle 2011, Cassidy et al. 2012) it has yet to be applied to the
two-layered condition. The aim of this paper is to (i) provide a
methodology for doing so, (ii) discuss preliminary trends due to
changing variation in undrained shear strength distributions, and
(iii) estimate quantitatively the degree of unconservatism in
using deterministic bearing capacity factors.
Figure 1. Definition of problem being investigated.
2 METHODOLOGY
The FE analysis model used in this paper is illustrated in
Figure 2. Two-dimensional plane strain conditions were
assumed and the commercial ABAQUS finite element package
B
B
Top layer
Bottom layer
s
ut
s
ub
hb
vb
ht
vt
Tian Y., Cassidy M.J.
Uzelli M.
