2063
Mechanism of Settlement Influence Zone due to Deep Excavation in Soft Clay
Mécanisme de la zone d'influence de tassement dû à une excavation profonde dans l’argile molle
Ou C.-Y., Teng F.-C.
National Taiwan University of Science and Technology, Taipei, Taiwan
Hsieh P.-G.
Hwa Hsia Institute of Technology, New Taipei, Taiwan
Chien S.-C.
Aletheia University, New Taipei, Taiwan
ABSTRACT: The objective of this study is to examine the mechanism of settlement induced by deep excavation through finite
element analysis. The USC model was selected for this purpose through the calibration of different soil constitutive models. A series
of parametric studies were then performed. It was found that in addition to the excavation depth, excavation width, the soft clay
thickness and depth to the hard soil are also related to the settlement influence zone. A simple method derived from the basal heave
failure mechanism is proposed to predict the settlement influence zone. One case history and one hypothetical excavation with the 80
m thick soft clay were used to verify the proposed method. For comparison, the existing empirical formulae were also used for
prediction.
RÉSUMÉ : L'objectif de cette étude est d'examiner le mécanisme de tassement induit par une excavation profonde à travers l'analyse
d’éléments finis. Le modèle USC a été choisi à cet effet par le calibrage de différents modèles de sol . Une série d'études
paramétriques a ensuite été réalisée. Il a été constaté qu’en plus de la profondeur et de la largeur de l'excavation, l'épaisseur et la
profondeur de l'argile molle sur le sol dur sont également liées à la zone d'influence du tassement. Une méthode simple dérivée du
mécanisme de rupture parsoulèvement basal est proposée pour prédire la zone d'influence du tassement. Une étude de cas et un travail
d'excavation hypothétique de l’argile molle sur 80 m d’épaisseur ont été utilisés pour vérifier la méthode proposée. A titre de
comparaison, les formules empiriques existantes ont également été utilisées pour la prédiction.
KEYWORDS: Deep Excavation, Soft Clay, Settlement, Constitutive Model, Settlement Mechanism
1 INTRODUCTION
The finite element method and empirical methods are often used
to predict the ground settlement induced by deep excavation.
The finite element method usually gives better predictions for
wall deflection than for ground settlements unless small strain
characteristics of soil are taken into account. Ideally, empirical
methods should be able to predict ground settlements well
because they are mainly derived from field observations of case
histories. However, most of them yield poor prediction in
ground settlement because settlement mechanism is unclear,
case histories adopted is limited, and the excavation depth is the
only parameter used in formulas.
The objective of this paper is to investigate the mechanism
of ground settlement induced by deep excavation under the
plane strain condition through finite element analysis. The study
focuses on the settlement under the normal excavation condition,
that is, no dewatering induced settlement, no excessively long
construction duration causing the occurrence of creep, and no
serious construction defects. A suitable soil constitutive model
was selected through calibration process. Then a series of
parametric studies were performed and the settlement
mechanism is proposed.
2 CALIBRATION OF SOIL CONSTITITIVE MODELS
Since “settlement influence zone” is not rigorously defined, the
authors proposed the conception of the primary influence zone
(
PIZ
) and the secondary influence zone (
SIZ
) on the basis of the
principles of mechanics and regression analysis of excavation
case histories (Hsieh and Ou 1998). The settlement curve is
steep in the
PIZ
where buildings receive more influence and in
the
SIZ
the slope of the curve is gentle and its influence on
buildings is insignificant. Finite element analyses are used to
capture the characteristics of
PIZ
.
Four soil constitutive models including the Hardening Soil
(HS) model, Hardening Soil with Small Strain (HSS) model,
=0 Mohr-Coulomb (MC) model, and Undrained Soft Clay
(USC) model, were adopted. Of these, the HS and HSS model
are the effective stress model and the
=0 MC model and USC
model are the total stress model. Both the HSS model and USC
model take into account that the soil exhibits high stiffness at
small strain.
Though the USC model is a total stress model, it considers
the variation of undrained shear strength with principal stress
rotation, variation of Young’s modulus with the increase of
stress level, high stiffness of soil at small strain, and rational
way to determine the undrained shear strength (Hsieh and Ou
2011). Similar to Duncan and Chang’s model, the tangent
Young’s modulus (
E
t
) in the primary loading is derived as
2
)
1(
SLR E E
f
ur
t
(1)
where
R
f
is the failure ratio,
SL
is the stress level,
E
ur
is the
unloading/reloading Young’s modulus.
The
E
ur
should degrade with the increase of strain or stress
level. The degraded Young’s modulus is assumed to follow a
hyperbolic function as
)
(
1
i
i
i
ur
SL SLnm
SL SL
E
E
(2)
where
m
and
n
are the degradation parameters relative to the
stress level,
E
i
is the Young’s modulus at small strain,
SL
i
is the
stress level corresponding to the threshold value of the small
strain or the initial yield strain.
An elastic surface, ES, is defined to represent the small strain
characteristics for the state of stress inside the elastic surface.
Figure 1 shows the relationships of stress and strain and of
