2217
Analyses of Seismic Slope Stability and Subsequent Debris Flow Modeling
Analysis de stabilité de pente sous sollicitation sismique et modélisation des écoulements de boues
induits
Long X., Tjok K.-M.
Fugro GeoConsutling Inc., Houston, Tx
ABSTRACT: Earthquake-triggered slope failures and subsequent submarine debris mass flow can cause severe consequences and
jeopardize the integrity of offshore structures in the proximity. In this paper, a two-dimensional seismic slope stability analysis for an
offshore liquefied slope site located in American Petroleum Institute (API) seismic zone 4 with layered stratigraphy is addressed. A
nonlinear dynamic analysis using the stress-
strain law of “hysteretic modeling” allowing for soil weakening under large strains was
adopted to study flow failure instability. Laboratory tests were performed for the derivation of soil dynamic parameters and
calibration of hysteretic soil model. Using “Constant
-
volume” constraints, the run
-out distance of subsequent debris flow was also
estimated.
RÉSUMÉ : Lors de séismes, les ruptures de pentes et les écoulements (de boue) sous-
marin qui s’en suivent peuvent avoir des
conséquences graves et compromettre l’intégrité des structures «
offshore » voisines. Dans cet article, on présente une étude (2D) de
stabilité d’une pente dans un site «
offshore
» liquéfiable, dans un zone sismique de niveau 4 (référentiel de l’American Petroleum
Institute). Une analyse dynamique non-linéaire utilisant la loi « hysteretic modeling » (loi permettant la modélisation du
radoucissement en grande déformation) a été adoptée pour étudier la rupture des pentes et les écoulements induits. Des tests de
laboratoire ont été effectués afin d’identifier et de déterminer les paramètres numériques de la loi de comportem
ent. En se plaçant
dans le cadre des déformations à volume constant, la modélisation a permis d’estimer la distance parcourue par les écoulements de
boues.
KEYWORDS: earthquake slope debris flow non-linear dynamic.
1 INTRODUCTION
Earthquakes are one of the major casuses of submarine slides.
Large shear stresses and deformations may be generated due to
seismic vibrations and irrecoverable volume changes are
accumulated accompanied by the rise of pore water pressure
(i.e., decrease of effective stress) and cyclic degradation of
shear strength. Liquefaction occurs when the effective stress
equals zero and soil behaves as a liquid. The submarine debris
mass flow following the onset of liquefaction and instablity of
liquefied slope can pose significant impact force and thus
jeopardize the integrity of offshore structures in the flow path.
In this paper, a two-dimensional seismic submarine slope
stability analysis is performed using commercial computer
program FLAC (
F
ast
L
agrangian
A
nalysis of
C
ontinua) (Itasca,
FLAC 7.0, an explicit finite difference program operating in the
time domain. It simulates the behavior of soils which may
subject to plastic flow when their yield limits are reached. The
Finn Model- Byrne Formulation is implemented in the analysis
to assess dynamic pore pressure generation and liquefaction
potential. Based on
“
constant-
volume” constraints, the run
-out
distance of subsequent debris flow was predicted using the
one-dimensional subagqueous debris flow model BING (Imran
and Parker, 2001).
2 TWO-DIMENSIONAL NON-LINEAR DYNAMIC
ANALYSIS
The slope stability and liquefaction potential for a project site
under seismic loading is assessed using FLAC for a
two-dimensional non-linear dynamic analysis
.
The dynamic
stress-strain behavior for the soil layers are simulated using a
hysteretic model by means of a non-linear stress-strain back
bone curve and the associated stress-strain loops that represent
the energy dissipated in the soil during seismic loading.
Strain-controlled cyclic simple shear tests have been performed
to determine dynamic soil parameters and calibration of
hysteretic soil model.
2.1 Site Condition
The site is located in International Organization for
Standardization (ISO) seismic zone 4 (ISO 2004).
Soil
stratigraphy consists of soft to firm clay layers overlying dense
fine sand. Based on the in situ suspension P (compression wave
velocity)-S (shear wave velocity) logging and cone penetration
test (CPT) data, the upper 33m (100ft) of the effective seabed is
around 160 m/s (525 ft/s) and the site is classified as either
American Petroleum Institute (API) criteria Type C or ISO
criteria Type E.
Figure 1 below illustrates the soil strategraphy at the study site.
The subsurface conditions comprise of 50 m (160 ft) soft to firm
clay underlain by a layer of dense to very dense fine sand down
to 105 m (350 ft). The mean sea level is 61.0 m (200 ft) above
seafloor. Due to soil erosion, a 20
˚
slope is formed as detected
from the geophysical multibeam bathymetric data. In light of
variations of material parameters, three soil Strata, i.e., Stratum
I , II and III, respectively, are defined within the clay sediment.
For the study of effective stress analysis, four Strata with
Stratum I, II and III for cohesive materials and Stratum IV for
cohesionless sediment, are used. The soil parameters including
submerged unit weight
, average shear wave velocity V
s
,
cohesion c, and friction
for each stratum are also provided on
the figure. .
The laboratory test data from cyclic direct simple shear
(CDSS) along with the generic curves of similar soil types
(Seed & Idriss for sand) were used as guidance to develop
hysteretic model for the analysis. Figure 2 presents the dynamic
