1935
3D Finite Element Analyses for a Laterally Loaded Pile Wall in Marine Environment
–
Case History
Analyses 3D par éléments finis pour un mur de quai chargés latéralement dans un port
–
Etude de cas
Bahr M.A., Tarek M.F., El-Ghamrawy M.K., Abouzaid K.S.
Al-Azhar University, Egypt
Shaarawi E.M.
Fayum University, Egypt.
ABSTRACT: This paper presents a 3D finite element study on a model to simulate an horizontal load test on a retaining pile wall. The piles
wall was constructed at the site of Port Ghalib marina on the Red-Sea coast of Egypt which is considered as an active seismic area. The
subsoil layers consist of 2 m to 3 m gravelly sand followed by a deep clayey silt layer. The ground water was observed at a depth of about
1.10 m below ground surface. The purpose of the test is to evaluate the pile displacement characteristics under exposed loads. Numerically a
10 m
–
length pile was modelled to simulate the actual case. Effect of surcharge, earth pressure and earthquake loads were taken into
consideration. The numerical analysis was performed and the results have been found to be in good agreement with the measured field test
results. In addition the finite element method make an ability to predict the deflection along the pile length.
RÉSUMÉ : Cet article présente un modèle 3D éléments de dimension par élément finie pour' simuler un chargement horizontal d'un mur de
quai du port de plaisance de Port Ghalib sur les côtes égyptienne de la mer Rouge qui est une zone sismique. La stratigraphie est constitué
d'une couche de 2 à 3 m de sable graveleux, en surface, suivie d'une couche de limon argileux, en profondeur. La nappe phréatique est située
à 1,1 m de la surface. L'objectif de l'analyse est de caractériser le déplacement latéral du mur en fonction du chargement. Le modèle
numérique a été construit pour simuler le cas réel. Les effets de la surcharge, pression des terres et effets des seismes, ont été pris en compte
dans le modèle. L'analyse numérique et les résultats sont en accord avec les résultats des expérimentaux de terrain. En outre, la méthode des
éléments finis donne une prédiction de la déviation le long du mur.
KEYWORDS: 3-D FEE model, analysis, earthquake load, lateral loading, pile wall, Red sea.
MOTS-CLÉS : modèle FEE 3-D, analyse, la charge tremblement de terre, chargement latéral, palplanches, Mer Rouge.
1.
0B
SUBSOIL PROFILE CONDITIONS AND PILE
GEOMETRY
Soil investigation showed that the soil profile at the site is as
follows:
Top layer (yellowish brown, gravel and sand) from ground
surface with depth ranging from 2 to 3 m, followed by a layer of
grey clayey silt, some fine sand, traces of broken shells,
extended to the end of executed boreholes (at depth of about 20
m).
The Standard penetration test (SPT) shows the N values as
follows:
-
From depth of 2 m to 7 m level N has values between
2 and 13
-
From depth of 7 m to 11 m level N = 3
-
From depth 11m to the end of boring N has values
between 7 and 11
Ground water was observed e at 1.10 m below ground surface.
The pile wall consists of contiguous bored piles of 1.0m
diameter and 10m length.
2.
1B
FINITE ELEMENT MODELLING
The finite element mesh considered in three dimensional
nonlinear finite element analysis as discussed in
Abouzaid et. Al.
(2010)
is shown in Figure (1-a). Based on symmetry, only one
pile of the model is meshed. 20 nodes brick element Solid 95
were used to simulate both soil, and pile with cap. It should be
noted that these quadratic elements exhibit high accuracy even
for high aspect ratios and can model accurately bending of solid
piles. During mesh design stage, a study was performed to
decide on appropriate (balanced) mesh size. The study showed
that a much larger mesh, with more elements (of lower aspect
ratios) would account for a fairly small change in results, so the
current mesh is sufficient for analysis.
The concrete pile section, with a diameter of 1.0 m and pile cap
beam have been meshed by Twenty node brick elements as
shown in figure (1a-1b) with the elastic property of concrete.
The soil domain has been simulated by strip of 1 m width with
symmetry conditions on both sides.. The depth of soil
considered below the pile tip 10 times the pile diameter. The
domain of soil considered has been found very much suitable for
the analysis of the laterally loaded pile as when loaded till
failure the soil elements at and near boundary do not experience
any deflection. Also the soil elements at and near to the bottom
boundary do not experience any deflection. The soil has been
modeled as elastoplastic medium following Drucker-Prager
(1952) .
Soil domain has vertically been divided to 20 layers, each of
thickness 1/20 of the pile embedded length to allow the soil
variation with depth. The elastic modulus is taken proportional
to strength parameter (c).
The boundary conditions considered are shown in Figure (6-8a),
the translations UX have been constrained in outer YZ plans,
and only UZ and UY have been permitted whereas UY have
been constrained in outer XZ plans, and only UZ, and UX have
been permitted, also the translation UZ have been constrained
over the area the soil block bottom and the translations UX, and
UY have been permitted, this have been done to overcome the
singularity of matrices and to help to get convergence. All nodes
