Effect of Deep Excavation on an Adjacent Pile

Foundation

Dimitrios Iliadelis

Diploma in Civil Engineering

Aristotle University of Thessaloniki, Greece

Submitted to the Department of Civil and Environmental Engineering in Partial

Fulfilment of the Requirements for the Degree of

Master of Engineering in Civil and Environmental Engineering

at the
Massachusetts Institute of Technology
June 2006

O 2006 Dimitrios Iliadelis

All rights reserved

The author hereby grants to MIT permission to reproduce and to

distribute publicly paper and electronic copies of this thesis document in whole or in
part in any medium now known or hereafter created.

Signature of Author ...................Y.Y.;.L-.z.,....w.....:. +. .: ...=...=......:. ........,............................................

Department of Civil and Environmental Engineering
May 23,2006

Certified by .-. .*.v.-.

.............................................................................r.Tvw. ........................
Andrew J. Whittle
Professor of Civil and Environmental Engineering
Thesis Supervisor

Accepted by ...........................................................................................................................
Andrew -J
Chairman, Departmental Committee for Graduate Students I
 Effect of Deep Excavation on an Adjacent Pile
Foundation
by
Dimitrios Iliadelis

Submitted to the Department of Civil and Environmental Engineering

on May 23,2006 in Partial Fulfilment of the
Requirements for the Degree of Master of Engineering in
Civil and Environmental Engineering

Abstract
The thesis studies the behavior of single axially loaded pile located close to a 30m
deep braced excavation in Marine Clay corresponding to site conditions of the
Kallang formation in Singapore. Parametric analyses were carried out, using non
linear three-dimensional finite element methods (with Plaxis Foundation 3-D),
comparing different pile lengths (17m, 30m and 42m) cross sections (solid concrete
sections of 0.4m and 1.Om diameter) and proximity to the excavation (2m-lorn). The
results focus on the development of horizontal deformations and bending moments
due to the excavation process. The computed results for end-bearing piles (42m long)
are compared with semi-empirical design methods proposed by Poulos and Chen
(1997). In general, this design method substantially underestimates the computed wall
deflections and bending moments. This result confirms the importance of site-specific
analyses for these types of complex soil-structure interactions.

Thesis Supervisor: Andrew J. Whittle

Title: Professor of Civil and Environmental Engineering
Table of Contents
Table of Contents................................................................................3
List of Tables ........................................................................................5
List of Figures....................................................................................-6
1. Introduction....................................................................................9
. .
2. Descript~onof the problem .................................................................13
2.1. Characteristics of Deep Excavation ..............................................13
2.2. Characteristics of the Piles........................................................17
2.2.1. Bearing Capacity of the Piles.................................................17
2.2.2. Bending Moment Capacity of the Piles .....................................19
2.2.3. Pile Flexibility..................................................................20
3. 3-D Finite Element Model .................................................................21
3.1. Introduction to the Finite Element Analysis....................................21
3.2. Geometry of the 3-D model .......................................................23
3.3. Material Input ......................................................................-26
3.3.1 . Soil Properties.................................................................26
3.3.2. Embedded Pile Properties...................................................-32
3.3.3. Diaphragm Wall Properties..................................................35
3.3.4. Strut Properties.................................................................36
3.4. Mesh Generation .. 38
3.5. Calculation Steps...................................................................41
4. Results of Finite Element Analyses........................................................43
4.1. Shallow Piles (1 7m deep) ........................................................-43
4.2. Results for Long Floating Piles (30m deep)....................................47
4.3. Long, end-bearing Piles (42m long)............................................-50
5. Method proposed by Poulos and Chen (1997)............................................58
5.1. Description of the Design Charts................................................58
5.2. Application of the method for the 42m long piles .............................63
6. Conclusions................................................................................68
7. References. ...............................................................................-70
Appendix A. Calculation of End Bearing Resistance and Skin Friction.. ................71
List of Tables
Table 2.1 Skin Friction for the piles ..........................................................17
Table 2.2 Load carried in end bearing. Qbf...................................................18
Table 2.3 Pile Bearing Capacities (kN)......................................................18
Table 2.4 Flexibility factor Kn.....................................................................................20
Table 2.5 Pile Characterization................................................................21
Table 3.1 Pre-load forces of Struts and applied horizontal point loads..................25
Table 3.2 Material Properties of soil layers.................................................26
Table 3.3 Embedded Pile Characteristics..................................................-32
Table 3.4 Skin Friction and Tip Resistance introduced into Plaxis......................33
Table 3.5 Axial Load of piles (kN) ...........................................................34
Table 3.6 Struts used for the supporting system of the Excavation......................37
Table 3.7 Characteristics of Struts inserted into the 3-D model .........................37
Table 3.8 Calculation Steps..................................................................-41
Table 4.1 Tensile stresses for 17m long piles ...............................................46
Table 4.2 Tensile stresses for 30m long piles ...............................................50
Table 4.3 Tensile stresses for 42m long piles ...............................................56
Table 5.1 Comparison of Poulos and Chen method with finite element
results for 0.4m diameter piles ....................................................64
Table 5.2 Comparison of Poulos and Chen method with finite element
results for 1m diameter piles ......................................................... 65
Table A.1 Skin friction at different soil layers..............................................77
List of Figures
Figure 1.1 Cross Section through Excavation (Finno. 1991)..............................10
Figure 1.2 Computed Sheet-Pile and Pile movements (Finno. 1991)....................11
Figure 2.1 Cross Section of Excavation with a 17m long adjacent pile
at a distance of 4m .................................................................13
Figure 2.2 Undrained shear stress of Marine Clay (Whittle & Davies. 2006)..........14
Figure 2.3 Construction Sequence of the C&C Excavation...............................16
Figure 2.4 Tensile and compressive stresses in the cross section of the pile ............19
Figure 3.1 Plan view of working plane a) at 0 m (el. 103 mRL) and
b) at -2m (el. IOlmRL).............................................................24
Figure 3.2 Details of the borehole introduced in the model. with all elevations........27
Figure 3.3 Details of the borehole introduced in the model. with the KOvalues ........28
Figure 3.4 General Properties of the Lower Marine Clay .................................29
Figure 3.5 Strength and Stiffness parameters for the Lower Marine Clay ...............30
Figure 3.6 Advanced Parameters for the Lower Marine Clay .............................31
Figure 3.7 End-bearing and Skin Resistance of a 30m long. 1.Om diameter
pile. as they are introduced in Plaxis 3-D Foundation program ..............33
Figure 3.8 Properties of solid section concrete pile. 42m long. withml .Om diam ......34
Figure 3.9 Properties of Diaphragm Wall ...................................................36
Figure 3.1 0 Characteristics of HEM type struts.............................................36
Figure 3.1 1 Properties of the strut type HEM300 introduced into the model ..............38
Figure 3.12 Generation of the 2-D model .................................................. -39
Figure 3.13 Generation of the 3-D model ...................................................40
Figure 3.14 View of final calculation step for a pile located 4m away
from the excavation face ........................................................42
Figure 4.1 Maximum deformations for 17m long piles located 2m from
the excavation face. for a 18m deep excavation ................................43
Figure 4.2 Maximum deformations for 17m long piles located 4m from
the excavation face. for a 18m deep excavation...............................44
Figure 4.3 Bending moment envelopes for 17m long piles located 2m distance
from the diaphragm wall, for 18m deep excavation.. ........................ .45
Figure 4.4 Bending moment envelopes for 17m long piles located 4m distance
from the diaphragm wall, for 18m deep excavation. ..........................46
Figure 4.5 Maximum deformations for 0.4m and lm diameter, 30m long piles
at a 2m distance from the diaphragm wall, for a 30m deep excavation.. ...47
Figure 4.6 Maximum deformations for 0.4m and 1m diameter, 30m long piles
at a 4m distance from the diaphragm wall, for a 30m deep excavation.. ...48
Figure 4.7 Bending moment envelopes for 30m long piles located 2m from
the diaphragm wall, for 30m deep excavation.. ................................49
Figure 4.8 Bending moment envelopes for 30m long piles located 4m from
the diaphragm wall, for 30m deep excavation.. ................................49
Figure 4.9 Maximum deformations for 0.4m and 1m diameter, 42m long piles
at a 2m distance from the diaphragm wall, for 30m deep excavation.. .....5 1
Figure 4.10 Maximum deformations for 0.4m and 1m diameter, 42m long piles
at a 4m distance from the diaphragm wall, for 30m deep excavation.. .....5 1
Figure 4.1 1 Maximum deformations for 0.4m and 1m diameter, 42m long piles
at a 1Om distance from the diaphragm wall, for 30m deep excavation.. ....52
Figure 4.12 Bending moment envelopes for 42m long piles located 2m from
the diaphragm wall, for 30m deep excavation.. ...............................53
Figure 4.13 Bending moment envelopes for 42m long piles located 4m from
the diaphragm wall, for 30m deep excavation.. ..............................53
Figure 4.14 Bending moment envelopes for 42m long piles located 30m from
the diaphragm wall, for 30m deep excavation.. ..............................54
Figure 4.15 Bending moment envelopes for a 42m long pile with 1 .Om
diameter, located at different distances from the diaphragm wall,
for 30m deep excavation.. .......................................................55
Figure 4.16 Horizontal Deformations of a 42m long pile with 1.Om diameter,
located at different distances from the diaphragm wall,
for 30m deep excavation.......................................................56
Figure 4.17 Distribution of axial forces for a 42m long pile with 1 .Om diameter.
located 2m from the diaphragm wall ...........................................57
Figure 5.1 Basic Problem analyzed by Poulos and Chen ..................................58
Figure 5.2 Basic Bending Moment and Basic Deflection versus
Distance from Diaphragm Wall ..................................................60
Figure 5.3 Correction Factors fpr Bending Moment (Poulos and Chen. 1997).........61
Figure 5.4 Correction Factors for Deflection (Poulos and Chen. 1997)..................62
Figure 5.5 Estimation of active and passive pressures below the excavation base .....63
Figure A .1 Values for the N factor............................................................71
Figure A.2 Values for the parameter a.......................................................72
Figure A.3 Values for the parameter J3..........................................................75
1. Introduction

Piles may be designed to control lateral soil movements in applications such as

stabilization of unstable slopes or potential landslides. However, in other cases piles
undergo lateral soil movements, for which they were not designed. Examples include
cases of existing piles adjacent to pile driving operations, piles supporting bridge
abutments (i.e. adjacent to approach embankments) or piles close to deep excavations.

In dense urban environments where land is scarce and buildings are closely
spaced, cut-and-cover excavations are widely used for basement construction and
development of underground transit facilities. One of the main design constraints in
these projects is to prevent or minimize damage to adjacent buildings. To date, much
of the research has focused on the lateral movements of the retaining wall system and
predictions of ground movements. Since many buildings are supported on deep
foundations, there is a concern that lateral ground movements resulting from the soil
excavation can damage the piles. Although an excavation will cause both vertical and
lateral soil movements, the second component is considered to be more critical, as
piles are usually designed to sustain significant vertical loads. In contrast, lateral loads
imposed by soil movements induce bending moments and deflections on the pile,
which may lead to structural distress and even failure. For this reason, this thesis pays
special attention to the development of lateral pile deformations caused by cut and
cover excavations.

Finno (1991) gives an interesting example of this class of problem He reports on

the performance of groups of step-tapered piles located adjacent to a 15m deep
tieback excavation, Figure 1.1.
 O I P
s a l e (It)

Figure 1.1 Cross Section through Excavation (Finno, 1991)

The excavation was performed through primarily granular soils (hydraulic fill,
alluvial sand; Figure 1.I) within the footprint of an existing framed structure. The
main columns were supported by groups of 21m long unreinforced or lightly
reinforced concrete piles. The temporary tieback sheet-pile wall was located as close
as 0.60m to the pile caps. Field observations, including lateral deformations of the
sheet pile wall and lateral and vertical deformations of the main columns, found that
several of the pile caps displaced up to 7.6cm towards the excavation (Figure 1.2).
Movements were two times higher than expected for excavations made under similar
conditions. Thus, the project engineers were concerned about the potential for
cracking and damage to the piles. Fortunately, field observations and finite element
analyses of the construction process showed that the recorded movements were not
large enough to cause serious damage.
 - - - - .. ..Flrsl level tCbadc8 In.1all.d (pwlod 2)
-- -- - f xcnvallon complete (pwrlod 7 )
.-. . .. ... Bsckflfl complete and tkbaeks releatud (psrlod 11)
---..-- Shmet plk ontraded (prlod 12)

Figure 1.2 Computed Sheet-Pile and Pile movements (Finno, 1991)

The objective of this thesis is to analyze the behaviour of a single pile, under
axial load, due to construction of an adjacent deep excavation. Parametric finite
element analyses have been carried out to investigate factors including the pile length,
flexural stiffness and proximity to the excavation. The results of the analyses are
compared with the semi-empirical design methods proposed by Poulos and Chen
(1 997).

The current analyses focus on soft ground conditions encountered in deep

excavations for the Circle Line project in Singapore (Contract Section C824), based
on a group design project (MFish, 2006). Therefore, the characteristics of the
diaphragm walls, the cross-lot bracing with basal improvement (using a jet grout pile
raft) are derived from this previous study.
 The analyses are performed using computer software, Plaxis 3-D Foundation (v.
1.5 beta), which is capable of simulating deep excavations and embedded pile
elements (friction or end-bearing). Three-dimensional analyses are necessary to
capture the lateral displacements around a single pile due to a long excavation (i.e.,
free field ground movements are two-dimensional).

Chapter 2 of the thesis introduces the characteristics of the reference problem,

including the geometry of the deep excavation, soil and adjacent pile properties.
Chapter 3 describes in detail the process of simulating all the components of the
problem within a 3-D analysis. Chapter 4 presents the results of the analyses, for a
series of different piles. Chapter 5 the semi-empirical design methods proposed by
Poulos and Chen (1997) is applied to the reference problem. In Chapter 6 the results
are being discussed, and useful conclusions are made. Finally, Appendix A contains
the calculations of the skin friction and tip resistance for all piles analyzed.
2. Description of the Problem

2.1 Characteristics of the Deep Excavation

Figure 2.1 shows the cross section of the reference geometry to be considered in
the thesis. This geometry considers a 20m wide, 30m deep cut-and-cover excavation.
Based on prior design studies (MFish, 2006) the excavation support system 1.2m
thick, 45m deep diaphragm walls, 7 levels of pre-loaded cross-lot bracing and a 1Om
thick jet-grout pile (JGP)raft below the final excavation grade.

C
1.2m
* ?Om -- I
, !> -. F
L oOu nCd a~t i o n

I O A CZMFETENT

Figure 2.1 Cross Section of Excavation with a 17m long adjacent pile
at a distance of 4m

The soil profile emulates ground conditions of the Kallang formation ground in
coastal Singapore comprising 42m of Marine Clays overlying units of Old Alluvium
(OA). The Marine Clay is further subdivided into Upper and Lower units (UMC and
LMC) each of which is underlain by a thin layer of stiffer desiccated clay (F2 Clay).
The groundwater table is located at the ground surface and water pressures are
assumed to increase hydrostatically through the Marine Clay and Old Alluvium.
Figure 2.2 shows the undrained strength profile of the Marine Clay.

Undrained Shear Strength, su (kPa) Stresses (kPa) Eu(MPa)

Figure 2.2 Undrained shear strength profile of Marine Clay (Whittle & Davies, 2006)
(analyses assume EJs,=400)

Most of the Old Alluvium is classified as very dense silty sands transitioning
with depth to very stiff to hard, silty clay. The shear modulus, G=40MPa at the top of
the OA layer (61.6 mRL) and increases significantly with depth. The SPT blow count
increases remarkably with depth in the Old Alluvium, ranging from N=10-20 bpf near
the upper surface to N> 100 bpf typically over a depth of 6- 1Om. The undrained shear
strength of the Old Alluvium is routinely assigned, su(kPa) = 5*N(bpf) in local
practice. Finally, the JGP layer is assumed to have high undrained shear strength
equal to s, = 300kPa and shear modulus increasing with depth (G=lOOMPa at El.
mRL 6 1.6).

The deep excavation is completed in eight stages. The main purpose of the JGP
layer is to brace the toe of the wall, to reduce bending moments and control lateral
wall deflections. Cross-lot struts are installed after each excavation stage at intervals
of 4m along the excavation. In order to be more effective and to reduce the wall
movement, the struts are pre-stressed (to 50% of the maximum computed loads). In
the excavation, two different types of struts are used (HEM260 and HEM300). Table
3.1 summarizes the pre-load schedule and maximum strut loads found from prior 2D
analyses (MFish, 2006). Figure 2.1 also shows two rows of king piles, inside the
excavation area. These are steel piles (type HEA400) and their main purpose is to
carry the weight of the struts and to prevent them from buckling.

This study assumes a pre-existing pile foundation at a working load

corresponding to 50% of the ultimate pile capacity (prior to excavation). The
construction sequence can be summarized as follows (Figure 2.3): 1) Installation
diaphragm walls (each panel is 4m long and 1.2m wide) 2) Installation of the JGP raft
(10m thick) within the Lower Marine Clay 3) Driving of king piles (49m long,
HEA400) 4) The first stage of excavation comprises a 2m deep unsupported cut 5)
The first level of struts are installed (and pre-loaded) at grade steps. 4 and 5 are then
repeated for the remaining excavation assuming 4m excavation intervals. The struts
are installed at 4m intervals.

Installationof DiaphragmWalls Installationof JOP

lnrhllatlon of Klng Poles

P ExcavationStage

The srrn procedure contlnws untll we reach

the final excrvatlon bvel

I
Figure 2.3 Construction Sequence of the C&C Excavation
2.2 Characteristics of the Piles

2.2.1 Bearing Capacity of the piles

The analyses consider circular solid section concrete piles, with diameter 0.4m
and 1.Om and lengths 17m, 30m and 42m. In the first case, the pile tip is bearing on
the F2 Clay layer, the second is a floating pile lying in the middle of the Lower
Marine Clay layer and the third case is end-bearing on the deep F2 clay. The
calculations also consider a range of pile- excavation offset distances ranging from
2m to 1Om.

Plaxis 3-D Foundation has the feature to model the skin friction resistance
depending on depth, e.g. constant or linearly. However, version 1.5 beta does not
include the feature of entering different skin resistance for each soil layer, when soil
profile is not uniform. Therefore, the skin friction of the pile is assumed to increase
linearly with depth along the pile (Table 2.1). Since the piles are embedded in clay, it
is recommended that the skin friction resulting from p-Methods is used, (Appendix
A)*
Table 2.1 Skin Friction for the piles

Skin Friction (kPa)

- L(m) top bottom
17 0 25.5
30 0 49.25
42 0 7 1.25
The ultimate load carried in the end bearing Qbfis equal:

Qbf = qbf Ab

where,
qbf is the end bearing capacity (Appendix A)
Ab is the base area of the pile
Table 2.2 presents the end bearing capacity Qbffor each pile.

Table 2.2 Load carried in end bearing, QbC

Qbf (kN)

679
1272

The bearing capacity of a single vertical pile under axial load is the sum of the skin
friction and the tip resistance.

where,
Qsf is the ultimate load carried in side friction: Qsf= fs As
fs is the limiting skin friction (Table 2.1)
As is the embedded surface area

Table 2.3 summarizes the bending capacities of the piles.

Table 2.3 Computed Pile Bearing Capacities (kN)

2.2.2 Bending Moment Capacity of the piles

The deep excavation induces bending moments on the pile which should be
checked if they can cause cracking of the pile. Due to the bending moments, tensile
stresses are induced in the cross section of the pile, while compressive stresses exist
due to the axial load from the foundation. In the simplest case of elastic stresses on an
unreinforced circular section the stresses can be represented as shown in figure 2.4.

Figure 2.4 Tensile and compressive stresses in the cross section of the pile

In this case the maximum net axial stress is

Where M is the bending moment induced in the pile, I is the second moment of area
cross section,

Clearly the potential for cracking will depend on the tensile or bending capacity
of the pile section. For the worst case scenario of unreinforced concrete piles,
cracking will occur if tension develops, hence onepOis required at all depths.
 According to the Eurocode, the minimum reinforcement for a concrete pile is
Astee]= 0.0 1*Aconcrek
which, (for reinforcement cover = 1Ocm) corresponds to Astee]=
7cm2and 8 1cm2for the 0.4m and 1.Om diameter piles, respectively.

2.2.3 Pile Flexibility

The flexibility factor KR proposed by Poulos is used, in order to evaluate the

flexibility of each pile. The flexibility factor KR is given from the following equation:

in which Epis the Young's modulus of the pile, I, is the moment of inertia of the pile,
Es is the averaged elastic modulus of the soil and L is the embedded pile length.
Table 2.4 presents the flexibility factor used in the current analysed for concrete piles.

Table 2.4 Flexibility factor KR

t

KR d(m)
0.4 1.O
17 4.6010~~ 1.8*lo5
L (m) 30 4.7010" 1.8*10-~
42 1.3 1 0-
"-7. lo-'

For the calculation of KR,we assumed E, = 40MPa (for concrete) and Es =

12.7MPa, which is the average elastic modulus of the reference soil profile. For each
pile, the parameter I, = 1.256010-3 m4 and 4.9010-2m4 for the 0.4m and 1 .Om diameter
piles, respectively.
The piles used in the finite element analyses range in flexibility as indicated in Table
2.5.
 Table 2.5 Pile Characterization
,

KR d(m)
0.4 1 .O
Relatively
17 flexible
stiff
very Medium
(m) 30
flexible flexibility
42 Very flexible
flexible

3.3-D Finite Element Model

3.1 Introduction to the Finite Element Analysis

For the analysis of the deep excavation, the software Plaxis 3D Foundation,
version 1.5 beta was used. Plaxis 3D Foundation is a commercial finite element
package intended for the three-dimensional deformation analysis of foundation
structures. The project geometry is modelled using a top view approach (i.e. model is
extended in a vertical direction). The input of soil data, structures, construction stages,
loads and boundary conditions is based on convenient graphical user interface, which
allows for a detailed and accurate modelling of the major features. From this
geometry a 3D finite element mesh is generated. Soil layers are defined by means of
boreholes. The current analysis assumes horizontal layers under KOstress conditions
which are defined using a single borehole. The excavation support structures are
defined in horizontal "work planes". A series of work planes are needed to represent
each excavation stage and to define the top and base elevations of the pile.

The Plaxis 3D Foundation program allows for automatic generation of

unstructured 2D finite element meshes based on the top view. In order to achieve
numerically accurate solutions, the mesh is refined in the vicinity of the pile. The soil
is discreted using 15-node wedge elements with quadratic variation of displacements.
The undrained shear behaviour of the clay layers (Marine Clay and Old Alluvium) is
modelled using the linear-elastic Mohr-Coulomb model. Undrained shear strength
properties are simulated using the conventional cp=O assumption (i.e. q=O, c'=su),
together with effective stress stiffness properties (E', v'). The vertical undrained shear
strength profile (Figure 2.1) is replicated by specifying linear variation, s,,(z) , within
each soil unit.

The struts are modeled using 2-D linear elastic plate elements with axial
stiffness. The current beta-version does not enable pre-loading of these elements, so
additional point loads are necessary to represent the strut installation. Embedded piles
are a new and innovative feature of Plaxis 3D Foundation (v. 1.5 beta). These
elements are not connected directly to the finite element mesh and can be placed at
arbitrary locations within the model. The current analyses consider circular piles with
specified end bearing and skin friction will be simulated by introducing appropriate
values for the factors Ttop,Tbotand F,, respectively (Table 2.1). The embedded piles
are axially loaded by a vertical point load corresponding to the design working load,
Q=Qu1d2.
3.2 Geometry of the 3-D model

The 3-D finite element model assumes symmetry such that, only half of the
excavation is simulated., The model considers the effect of a 16m length of a
excavation on the response of adjacent pile. These assumptions are essential to limit
computation demands associated with large 3-D models.

The project geometry is modelled using a top view approach. We should

introduce working planes at the top and bottom of the pile and the diaphragm wall
(corresponding to elevations 103 mRL, El. 58mRL, respectively). Moreover, working
planes corresponding to the excavation stages are needed. In reality, when
constructing a deep excavation, the struts are installed approximately I m above each
excavation stage. In our model, we assumed that the strut elevation coincides with the
excavation level for every excavation grade, in order to simplify the model. This
assumption is justified for current purposes to investigating the response of the pile
but is not realistic for designing the lateral earth support system.

Figure 3.la presents the top working plane of the model (El. 103 mRL)
corresponding to the ground surface. The dimensions of the model are 16m x 32m and
the vertical dimension is defined by the program to be 3m below the lowest working
plane. The figure shows the diaphragm wall, the borehole, the embedded pile and strut
locations. All the geometry lines perpendicular to the diaphragm wall is where the
struts are going to be placed, in deeper working planes.
 tiiilE Pile

a b

Figure 3.1 Plan view of working plane a) at 0 m (el. 103 mRL) and b) at -2m (el. 10lmRL)

As mentioned before, Plaxis 3D Foundation, version 1.5 beta does not include
the feature of pre-stressing the struts, which are simulated as "beams" in the model. In
order to account for the pre-stressing of the struts, horizontal point load are applied, at
the points where the struts are connected with the diaphragm wall (Figure 3.lb). In
order to activate the pre-loading, the following procedure is adopted: We first create
point loads at the strut ends and then install the loads while the strut (beam) is
inactive, to set the preload. Then, in the next phase, we remove the loads and activate
the struts (beam). The struts are generally sufficiently stiff to pick up at least 90% of
the load. In Figure 3.lb the working plane at elevation -2m (el. 101 MRL) is
presented. Apart from the diaphragm wall and the embedded pile, we can also see the
struts and the horizontal point loads acting at the strut ends.
 Based on prior analyses by MFish (2006) the pre-load schedule of the struts is
shown in Table 3.1.

Table 3.1 Pre-load forces of Struts and applied horizontal point loads

Strut Elevation Pre-Load Point Load

No. (m) Type (Wm) (kN)
I -2 HEM260 375 1500
2 -6 HEM300 450 1800
3 -10 HEM300 600 2400
4 -14 HEM300 700 2800
5 -18 HEM300 550 2200
6 -22 HEM300 700 2800
7 -26 HEM260 275 1100

Since the struts are installed every 4m in plan view, the applied horizontal point
loads are equal to 4 times the pre-stress of the strut at every elevation, and their values
are listed in the last column of Table 3.1.
3.3 Material Input

3.3.1 Soil Properties

Table 3.2 presents all the soil properties used for the different layers in the 3-D
model, based on recommendations from Whittle and Davies (2006).

Table 3.2 Material Properties of soil layers

Bulk Shear Poissonts Permeability

RL Density s, ' Modulus Ratio Coefficient
Stratum K O
(m) Yt (") G V'
k
(k~lm~) (MPa) (m/day)

18
Upper MC 102.9 l6 0 3 0.25 8.6* 10-5 0.7
25
F2 Clay 85.6 19 88 0 11.7 0.25 8.6* 10-5 0.7
31
.Lower MC 83.4 16.8 0 5.2 0.25 8.6* 10-5 0.7
47
F2 Clay 63.2 20 88 0 11.7 0.25 8.6* 10-5 0.7
OA 100
61.6 2o 40 0.25 8.6* 10-4 1
weathered 500
OA
53.9 500 0 67 0.25 8.6* 10-5 1
Competent
JGP 16 300 0 108 0.15 8.6* 10-5

Since the soil stratigraphy remains the same in the whole area of the excavation
which will be simulated, we need only one borehole in the model in order to introduce
the different soil layers with the appropriate elevations. The details of the borehole,
the location of which in plan view was shown in figure 3.3, are given in the following
figure. On the left of figure 3.3, we can see the soil profile of the borehole and the
level of the groundwater table.
 UMC

F2

LMC

F2

OAW

OAC

- - 1

Figure 3.2 Details of the borehole introduced in the model, with all the elevations

If we choose the "Soil" option, in the window which opens we can input the
value of the & for each layer. This process is presented in figure 3.4 in which it is
also shown that for the analyses we assumed that the value of the KOis the same in
both horizontal directions. The input of this soil parameter in the 3-D model is
important because it will later be used to calculate the initial conditions of the
problem.
 Figure 3.3 Details of the borehole introduced in the model, with the KOvalues for each layer

In the next figures, the process of defining the material properties of the Lower
Marine Clay layer is briefly described. On the right part of figure 3.5 the unit weight
and the permeability of the material for all three directions is defined. In the "Material
Set" part of the window, it is important to define the correct material model and
material type for the mode. In our case, the Mohr-Coulomb model will be used, and
"Undrained" properties are assigned, to enable simulation of the undrained response
of the low permeability clays (computing effective stresses and pore pressures.
 Figure 3.4 General Properties of the Lower Marine Clay

Figure 3.6 presents how the Stiffness and Strength Parameters are introduced
into the finite element model for the Lower Marine Clay. For this layer, the Poisson's
ratio v = 0.25 and the shear modulus GEf = 5200kPa. The value of the Elastic
Modulus befis automatically calculated by the program.
 -
-Coulomb Lower MC
_I

Figure 3.5 Strength and Stiffness parameters for the Lower Marine Clay

According to Table 3.2, the undrained strength on the Lower Marine Clay,
increases with depth. In order to simulate this behavior in the model, we have to select
the "Advanced" properties for the Mohr-Coulomb criterion and assign the correct
value of Cinnemen~ It is also important to specify the depth from which this parameter
will be used by the model, which is equal to elevation -20m for the case of the Lower
Marine Clay layer.
 ~cedpuameters Mohr-Coulomb
rn

Figure 3.6 Advanced Parameters for the Lower Marine Clay

The procedure for introducing the properties for all the soil layers, including the
JGP layer, is the same as this described in the current chapter.
3.3.2 Embedded Pile Properties
Two different piles are studied in the current thesis. They are solid section
concrete piles with diameter 0.4m and 1.Om respectively. All the necessary properties
which were introduced in the program are listed in the following table.

Table 3.3 Embedded Pile Characteristics

Pile 1 Pile 2
d (m) 0.4 1.O
A (m2) 0.1256 0.785
E (MP~) 40000 40000
I2 (m4) 0.00 1256 0.049
v 0.2 0.2
Y (mlm3) 24 24

For a circular solid section the moment of inertia is equal to:

Parameters Ttop,Tbt are used to specify the skin friction at the top and at the tip
of the pile respectively. The skin friction at the ground surface (Ttop)is taken equal to
zero, and for every pile, the skin friction at its tip (Tbot)is taken equal to the skin
friction corresponding to that point (Table 2.1) multiplied by the perimeter of the pile.
Parameter F, corresponds to the tip resistance (Table 2.2).

Table 3.4 presents the values for Ttop,Tbotand F, for all piles analyzed and
Figure 3.7 illustrates an example for a 30m long, 1.Om diameter pile.
 Table 3.4 Skin Friction and Tip Resistance introduced into Plaxis

Figure 3.7 End-bearing and Skin Resistance of a 30m long, 1.0m diameter pile, as they
are introduced in Plaxis 3-D Foundation program
 Figure 3.8 shows the input data of all the material properties mentioned above
for the 0.4m diameter solid section concrete pile with length 30m.

I Pile Properties
Matorialset --

, [I .Om diameter

Figure 3.8 Properties of solid section concrete Pile, 42m long, with 1.0m diameter

In the analyses performed, each pile is axially loaded with a point load. A Factor
of Safety equal to 2 was used and the axial load of each pile is presented in Table 3.5.

Table 3.5 Axial Load of piles (kN)

3.3.3 Diaphragm Wall Properties
In the particular model, a 45m deep diaphragm wall is installed. This wall is
1.2m wide and is made of reinforced concrete. Therefore, in the program we insert the
value d = 1.2m for the width of the wall and y = 24kPa for its unit weight.
In order to define the Young's Modulus, we will use the following equation:

where E ,is the Young's Modulus for the reinforced concrete and f,is the strength of
the concrete. Usually it is f,=34-55 MPa and by assuming that high quality concrete
is used for the construction of the diaphragm wall, we have:

As a result, the values we insert into the model are:

The Poisson's ratio for the concrete is vl2 = v13 = v23 = 0.2, SO by using the equation
G = E*2.(l+v),we get G12= GI3= G23= 9.601o7 kN/m2.

All these properties of the diaphragm wall are shown in the next figure.
 r brotoPic
I
tii Lineor
C' Non l i a r
81 : - 1 kN/d 012: [9.600~*07
I wd

Figure 3.9 Properties of Diaphragm Wall

3.3.4 Strut Properties

As mentioned before, two different types of struts are used in the deep
excavation model studied, which are HEM260 and HEM300. The characteristics of
these struts were taken from design charts, and are presented next.

Figure 3.10 Characteristics of HEM type struts

 Table 3.6 Struts used for the supporting system of the Excavation

Strut Type h (mm) b (mm) sa (mm) e (mm)

HEM260 290 268 18 32.5
HEM300 340 310 21 39

All the necessary properties introduced into the finite element analysis, regarding the
strut material properties, are shown in the following table.

Table 3.7 Characteristics of Struts inserted into the 3-Dmodel

The following figure schematically shows the input of these parameters in Plaxis 3-D
Foundation for the HEM300 type strut.
 " . -

II
..I
1. ' -
IS Linear I
"

r IE -1 w/~Z- ;

Figure 3.11 Properties of the strut type HEM300 introduced into the model

3.4 Mesh Generation

After the geometry and the properties for all the materials have been introduced
into the model correctly, the mesh is ready to be generated. This is done first by
generating the 2-D mesh, and then extending for the 3-D geometry. Figure 3.1 1 shows
the plan view of the 2-D mesh for the problem we are studying. We can notice that the
mesh is more refined in the area included by the diaphragm wall and the line 14-15
described in section 3.2. This is done in order to obtain numerically more accurate
results.
 Refined
Mesh

Figure 3.12 Generation of the 2-Dmodel

After the 2-D mesh has been created, we are ready to generate the 3-D one,
which is shown in figure 3.12. In this figure we can determine the different soil layers.
We also notice that the mesh around the pile is more refined than the rest of the area,
in which the finite element mesh is coarser.
Figure 3.13 Generation of the 3-D model
3.5 Calculation Steps

After the procedure described in the previous sections has been completed, we
are ready to set up the analysis procedure. The analysis of our 3-D model is done in
28 calculated phases, which are listed below. The reason we have so many phases,
while there are only eight excavation stages is the fact that for each excavation stage
we first have to apply the horizontal point loads at the strut ends, which represent the
pre-stress force, then deactivate these loads and activate the struts, and finally perform
the actual excavation step.

Table 3.8 Calculation Steps

In Table 3.9 we can see that the first phase includes the determination of the
initial conditions of our problem, which is done with the KOprocedure. After this is
done, we sequentially activate the embedded pile with the axial load on its top, the
diaphragm wall and finally the grouting, inside the excavation. Before beginning with
the excavation stages, it is important to reset all displacements from the first 4 phases
to zero. This is done because our main concern is to analyze the behavior of the pile
when the deep excavation is performed and as a result, the impact of the construction
of the diaphragm wall to the pile is of minor importance.

Figure 3.12 presents a 3-D view of the final calculation phase which
corresponds to excavation to -30m, and identifies the diaphragm wall the struts and
the JGP raft. Finally, the vertical point load at the top of the embedded pile can be
seen. In this figure, the embedded pile is located 4m from the deep excavation.

Diaphragm
'all

Struts

JGF

Figure 3.14 View of final calculation step for a pile located 4m away
from the excavation face
4. Results of Finite Element Analyses

This chapter presents results of finite element analyses for single pile
foundations next to the 30m deep excavation in Singapore Marine Clays. The
excavation support system and modelling details are given in the preceding chapter.
According to previous 2-D finite element analyses (MFish, 2006) the maximum
deflection of the diaphragm wall is expected to be approximately 50mm and it occurs
at depth of 22.5m. This was confirmed by the 3-D analyses.

4.1 Shallow Piles (17m deep)

Figure 4.1 shows the maximum horizontal displacements of 0.4m and 1.0m
diameter piles for a 18m deep excavation. The figure compares the lateral deflections
of piles located 2m from the wall with those of the diaphragm wall itself. The results
show that the top of the wall is displaced 24mm from the excavation while maximum
wall deflection 49mm occurs at a depth of 22.5m. The piles exhibit very similar
deflections. The maximum deformation occurs at the toe of the piles 6h=
-
-
r
-
lorn 2m

*
* Upper marine clay

- __+

F2 clay Dirphrclgmwall
0.401die pik
1.hdii pile

Lower marine clay

- JGP

F2 clay

Old Alluvium, Weathered

50
-60 -40 -20 0 20 40

Deformation (mm)
Figure 4.1 Maximum deformations for 17m long piles located 2m from the excavation face, for a
18m deep excavation
 The results in Figure 4.1 show that for a small distance between the piles and the
excavation face, the deformations of a floating pile are very similar to those of the
wall (and consequently of the soil). Moreover, increased bending stiffness of the pile
has no effect on the lateral pile deformations.
When the distance between the pile and the excavation face is increased from
2m to 4m, Figure 4.2, there is still little change in the computed pile deflections. In
this case, the pile head deformations are 22mm, while the toe movement is 34mm.
The toe movement is 3mm less than the wall at the same elevation.

Upper marine clay

, Diaphragm wall

t - - - 0.4mdia pile

Lower marine clay

Old Alluvium, Weathered

50
-60 -40 -20 0 20 40

Deformation (mm)
Figure 4.2 Maximum deformations for 17m long piles located 4m from the excavation face, for a
18m deep excavation

In general, these observations are consistent with assumptions made by Poulos

and Chen (1997). The maximum deflection for each pile is less than the allowable
(68mm) which was estimated in chapter 2.2.4.
Figure 4.3 presents the bending moment envelopes for 17m long piles located
2m from the excavation face. As expected, the predicted bending moments are much
larger for the rigid pile (1.0m diameter) than for the flexible one (0.4m diameter) and
equal to 420kN.m and 29kN.m respectively.

Bending Moments (kNm)

100 200 300

-..-. Enwlope of 0.4m diam.

pile
--.-...
Moments of 0.4m diam.
Pile for 18rn deep exc.
- Enwlope of 1.Om diam.
pile
---- Moments of 1.Om diam.
Pile for 18m deep exc.

Figure 4.3 Bending moment envelopes for 17m long piles located 2m distance from the
diaphragm wall, for 18m deep excavation

When the pile-excavation face distance is increased to 4m (figure 4.4) the

bending moment distribution along the piles does not change significantly, but the
maximum values observed are decreased significantly, especially for the rigid pile. As
shown, in this case, the maximum bending moments are 26kN.m for the 0.4m
diameter pile and 260kN-m for the lm diameter pile, respectively.
 Bending Moments(kNm)
50 100 150 200 250 300

I
.---...
Enmlope of 0.4m diam.

1
pile
Moments of 0.4m diam.
-.-em

Pile for 18m deep exc.

~~~nw
pile
lopeofl.Omdiam.

Moments of 1.Om diam pile

for 18m deep exc.

Figure 4.4 Bending moment envelopes for 17m long piles located 4m from the diaphragm wall,
for 18m deep excavation

Table 4.1 presents the tensile stresses which develop on the pile sections for a
18m deep excavation. The results show that if the piles are unreinforced, then
cracking is likely to occur. However, the minimum reinforcement suggested by
Eurocode 2 (section 2.2.2) is enough to prevent cracking.

Table 4.1 Tensile stresses for 17m long piles

max M
M at As*
L(m) d(m) I (m4) H(m) (m) N(kN) (MPa) (cm2)
(kNm)
2m 17 0.4 1.25 18 12 29 85 4 5
dist. 17 1.0 4.910-~ 18 12 420 523 3.5 5
4rn 17 0.4 I.25 18 12 26 111 3 4
dist. 17 1.0 4.910-~ 18 12 260 530 2 3

* Area of steel required to resist computed tensile stress

4.2 Results for Long Floating Piles (30m deep)

For the 30m long piles, the maximum horizontal displacements for both the
0.4m and 1.0m diameter piles occur at a depth of approximately 24m for excavation
to the final grade (30m deep excavation). The lateral deformations of the piles and of
the diaphragm wall are plotted in Figure 4.5. All the pile deformations in this figure
are for a pile to excavation face distance equal to 2m. According to this figure, the
horizontal deformations at the top of the piles are 25mm and 24mm for the 0.4m and
1.0m diameter piles, respectively, while the maximum deformations are 47mm and
46mm, respectively.

- -
>
i
I
- Upper marine clay

>

-
> F2 clay
-
-
-- Diaphragm wall
O.4m dia pile
.-.------1.m&*pile

- JGP

F2 clay
Lower marine clay

Old Alluvium, Weathered

-60
\;
-40 -20

Deformation (mm)
0
,,,,,,,,
20 40

Figure 4.5 Maximum deformations for 0.4m and lm diameter, 30m long piles at a 2m distance
from the diaphragm wall, for a 30m deep excavation

When the distance between the pile and the excavation face is to 4m, Figure 4.6,
there is negligible change in pile head deflections (23mm and 2 1mm for the 0.4m and
1.0m diameter piles, respectively) while the maximum movements are reduced by
about 10% (44mm and 4 1mm,respectively).

- 1

- Upper marine clay

Diaphragm wall
- - F2 day
- --.-------
-- 0 . h dia pile
I dia

Lower marine clay

- JGP

F2 clay

Old Alluvium, Weathered

I s ~ ~ l a ~
* l I ~a l * ~I i ~ ~ r

-60 -40 -20 0 20 40

Deformation (mm)
Figure 4.6 Maximum deformations for 0.4m and l m diameter, 30m long piles at a 4m distance
from the diaphragm wall, for a 30m deep excavation

In figure 4.7 the bending moment envelopes for both piles are presented for pile-
excavation face distance equal to 2m. The maximum predicted bending moments are
much larger for the lm diameter pile than for the 0.4m diameter one and equal to
1122kN-m and 45kN.m respectively. The reason for the negative bending moments
for the 0.4m diameter pile at elevation approximately -18m is because at this depth a
stiffer layer of F2 Clay is present.
 Bending Moments (kNm)
600 0 500 1000

0
11 Enwlope of 0.4m diam.

1 -.-.-Moments of 0.4m diam.

-~nwlope
pile
Pile for D m deep exc.
of 1." diam.

Moments of 1.Om diam.

1
Pile for 30m deep exc.

Figure 4.7 Bending moment envelopes for 30m long piles located 2m from the diaphragm wall,
for 30m deep excavation

When the pile-excavation face distance is increased to 4m (figure 4.4) the

bending moment distribution along the piles does not change, but the maximum
values observed are decreased, especially for the rigid pile. Similar behaviour was
previously observed for the 17m long piles as well. As shown, in this case, the
maximum bending moments are 39kN.m for the 0.4m diameter pile and 871kN-m for
the lm diameter pile.
.-
I

Bending Moments (kNm)

200 400 600

..-....Enwlope of 0.4m diam.

pile
-.---Moments of 0.4m diam.
Pile for 30m deep exc.
-Enwlope of 1.Om diam.
pile
- - - - Moments of 1.Om diam.
Pile for 30m deep exc.

Figure 4.8 Bending moment envelopes for 30m long piles located 4m from the diaphragm wall,
for 30m deep excavation
 Table 4.2 presents the tensile stresses which develop on the pile sections for a
30m deep excavation. The results show that no tensile stresses will develop for the
0.4m diameter piles, but cracking is likely to occur for the 1 .Om diameter piles, if they
are unreinforced. However, the minimum reinforcement suggested by Eurocode 2
(section 2.2.2) is enough to prevent cracking.

Table 4.2 Tensile stresses for 30m long piles

max M
M ut As*
L(m) d(m) I (m4) H(m) z(m) N(kN)
( k ~ ~ ) (MPa) (cm2)
2m 30 0.4 1.2510" 30 23 45 522
dist. 30 1.0 4 . 9 1 0 ~ ~ 30 23 1122 814 10 16
4m 30 0.4 1.2510" 30 23 39 556
dist. 30 1.0 4.910-~ 30 23 871 769 8 13

* Area of steel required to resist computed tensile stress

4.3 Long, end-bearing piles (42m long)

For the 42m long end-bearing piles, additional analyses have been carried out
for pile-wall spacing up to 1Om. The maximum horizontal displacements of the wall
and for both the 0.4m and 1.0m diameter piles occur at a depth of approximately
22.5m during excavation to the final grade at a depth of 30m. The deformations of the
piles and of the diaphragm wall are plotted in Figures 4.9 and 4.10 for pile-wall
spacing of 2m and 4m, respectively. The figures show that there is negligible change
in pile head deflections (24mm and 23mm for the 0.4m and 1.0m diameter piles,
respectively, while maximum movements are reduced by about 10% (47mm and
46mm, respectively).When the distance from the diaphragm wall is increased from
2m to 4m, the maximum movements are reduced by about 10% (44mm and 42mm,
respectively).
 - C

- Upper marine clay

*
- F2 clay -
-- - -
Diaphragm wall
0.4mdim pile
> ----.---- 1,omdja pile

>

Lower marine clay

- JGP

F2 clay
k
k Old Alluvium, Weathered
-60 -40 -20 0 20 40

Deformation (mm)

Figure 4.9 Maximum deformations for 0.4m and l m diameter, 42m long piles at a 2m distance
from the diaphragm wall, for 30m deep excavation

Upper marine clay

-
--- Diaphragm wall
0.4mdia pile
---------I .Om dia pile

Lower marine clay

Old Alluvium, Weathered

1
50
-60 -40 -20 0 20 40

Deformation (mm)

Figure 4.10 Maximum deformations for 0.4m and l m diameter, 42m long piles at a 4m distance
from the diaphragm wall, for 30m deep excavation
 When the pile is located 10m from the diaphragm wall, there is little difference
in the computed mode shape, Figure 4.1 1. However, there is a significant reduction in
the maximum deformations. The results show 6, = 29-30mm for both 0.4m and 1.Om
diameter piles.

0 -

Upper marine clay

10 -

F2 clay
20 -

30 -
Lower marine clay
JGP

40.
F2 clay

Old Alluvium, Weathered

50
-60 -40 -20 0 20 40

Deformation (mm)
Figure 4.11 Maximum deformations for 0.4m and l m diameter, 42m long piles at a 10m distance
from the diaphragm wall, for 30m deep excavation

The bending moment envelopes for both piles are presented in figures 4.12 and
4.13 for pile-diaphragm wall distances equal to 2m and 4m, respectively. The
predicted bending moments are much larger for the 1.0m diameter than for the 0.4m
diameter pile (MMax= 1190kN-mand 972kN-m,respectively for piles located 2m from
the diaphragm wall and MMax= 43kN-m and 39kNem at 4m).
 Bending Moments (kNm)
-1000 -500 0 500 1000 1500

-.---Enelope of 0.4m diam.

pile
A .......Moments of 0.4m diam.
E
V
Pile foe 30m deep exc.
Enelope of 1.Om diam.
pile
----Moments of 1.0m diam.
Pile for 30m deep exc.

Figure 4.12 Bending moment envelopes for 42m long piles located 2m from the diaphragm wall,
for 30m deep excavation

Bending Moments (kNm)

-1000 -500 0 500 1000 1500

- - - -Envelope
- of 0.4m diam. pile

Moments of 0.4m diam. Pile

.+..--a

A for 30m deep exc.

Y
E -Envelope of 1.Om diam. pile
Eaa ----Moments of 1.Om diam. Pile
0
for 30m deep exc.

Figure 4.13 Bending moment envelopes for 42m long piles located 4m from the diaphragm wall,
for 30m deep excavation
 Finally, when the pile-wall separation is increased to 10m (Figure 4.14), then
there is a large reduction in the maximum moments (MM, = 18kN.m and 505kN*m,
for the 0.4m and 1.Om diameter pile, respectively).

Bending Moments (kNm)

-200 0 200 400 600

-.-.-Enwlope of 0.4m diam.

pile
A .-.....Moments of 0.4m diam.
E
V
Pile for 30m deep exc.
-Enelope of 1.Om diam.
pile
----Moments of 1.Om diam.
Pile for 30m deep exc.

AC
.w
1

Figure 4.14 Bending moment envelopes for 42m long piles located 10m from the diaphragm wall,
for 30m deep excavation

The figures show that the maximum moment for the 42m long piles with 1.0m
diameter is 1190kN-mand occurs at a depth 22.5m. According to the analyses (Figure
4.1 5), at this point the axial load is 289 1kN.

The values of both the maximum bending moment and horizontal deflection
depend a lot on the pile to excavation face distance and decrease significantly when
this distance increases. In Figure 4.1 6, the maximum bending moment is reduced by
approximately 60% when the pile to excavation distance is increased from 2m to 1Om.
 Bending Moment (kNm)
-1000 -500 0 500 1000 1500

E
Y

5
#
,I

45 -
Figure 4.15 Bending moment envelopes for a 42m long pile with 1.0m diameter, located at
different distances from the diaphragm wall, for 30m deep excavation

Similarly, as shown in figure 4.17, the pile to excavation distance affects the
maximum observed horizontal deformation on the pile as well. For example, the
maximum deformation of the pile decreases by approximately 40% when the distance
increases from 2m to 10m. On the other hand, the diameter of the pile, and
consequently its rigidity does not play an important role for the maximum horizontal
deflections, as it was clearly shown in the figures of chapter 4. As it was observed, the
increase of the pile diameter from 0.4m to 1.0m slightly reduces the horizontal
deformations. It does, however, influence the maximum moment of the piles, as it was
expected. In all cases these were lower than the pile's bending moment capacity.
 r
-
I
I
I
D
I
I
I
D
Upper marine clay
- I
I
I
I
I
I
- 11 1 F2 clay -.-.-..---
-- Displnagm wall
2m &st
+
I
I
I

I
-- - 4m dist
lhdist

> I D

I
I I
I

I I Lower marine clay

JGP
II I
D

I f I F2 clay

Old Alluvium, Weathered

-60 -40 -20 0 20 40

Deformation (mm)

Figure 4.16 Horizontal Deformations of a 42m long pile with 1.0m diameter, located at different
distances from the diaphragm wall, for 30m deep excavation

Table 4.3 presents the tensile stresses which develop on the pile sections for a
30m deep excavation. The results show that no tensile stresses will develop for the
0.4m diameter piles, but cracking is likely to occur for the 1 .Om diameter piles located
close to the diaphragm wall (2m-4m), if they are unreinforced. The minimum
reinforcement suggested by Eurocode 2 (section 2.2.2) is enough to prevent cracking.

Table 4.3 Tensile stresses for 42m long piles

max M
M ut As*
L(m) d(m) I (m4) H(m) z(m) N(kN) (MPa) (cm*)
(kNm)
2m 42 0.4 1.25 30 22.5 43 1363 - -
dist. 42 1.0 4.9 10 -' 30 22.5 1190 2890 25 38
4m 42 0.4 1.25 10" 30 22.5 39 1457 -
dist. 42 1.0 4.9 10 -' 30 22.5 972 2314 7 11
1Om 42 0.4 1.25 10" 30 22.5 18 1464
dist. 42 1.O 4.9 I 0-' 30 22.5 505 2582 -

* Area of steel required to resist computed tensile stress

Figure 4.18 presents the distribution of the axial forces for a 1.0m diameter piles,
located 2m from the excavation. As shown, the axial force slightly increases until a
depth of 18m. Same behaviour was observed for most piles analyzed in this thesis.
This happens because settlements occur at the vicinity of the excavation and,
consequently, the ground movements of the surrounding soil generate loads on the
pile (negative skin friction). However, this behaviour does not impact the axial loads
at the depth where the maximum bending moments occur.

Axial Force (kN)

0 500 1000 1500 2000 2500 3000 3500
0 I I I I 1

5 -

10 -
15 -
A

g 20
25 -
30
-

35 -

40 -

45 --
. -- .- .- . .- . . .... ---. . -. .. . -- - -. -- .- ..---.... --.-.----.. . ---.....-.
. - .-- .
,-. -------. -

Figure 4.17 Distribution of axial forces for a 42m long pile with 1.0m diameter, located 2m from
the diaphragm wall
5. Method proposed by Poulos and Chen (1997)

5.1 Description of Design Charts

Poulos and Chen (1997) developed a series of design charts for predicting the
maximum deflection and bending moment of a single pile due to excavation-induced
lateral soil movements, with specific attention being focused on braced excavations in
clay layers. The basic problem analyzed and the parameters selected are shown in
Figure 5.1 where B, the half-width of the excavation, H is the total thickness of the
soil layer, X is the distance of the pile from the excavation face, c, is the undrained
shear strength of the soil, Esis the Young's modulus of the soil, y is the unit weight of
the soil, Lp is the pile length, d is the pile diameter, Ep is the Young's modulus of the
pile, Elwis the stiffness of the wall, s is the spacing of the struts, L, is the wall length
and h, is the maximum depth of the excavation.
For the simulation of the plain-strain excavation a two-dimensional finite element
program was used and the analyses of the pile response were carried out using a
boundary element program.

Figure 5.1 Basic Problem analyzed by Poulos and Chen

 It was found that pile deflections follow soil movements closely at all distances
from the excavation face and therefore the maximum pile deflections can be
conservatively equated with the soil movements. The pile bending moment profiles at
various distances X are quite similar in shape, but the maximum value decreases with
increasing distance X.
Several cases with different key parameters were studied by Poulos and Chen (1997)
which revealed the following:
1. Pile response (bending moment and deflection) increases with increasing c.
and Esdue to an increased ultimate lateral soil pressure.
2. Pile response increases with increasing stability number N, due to larger lateral
soil movements.
3. Pile response decreases with stiffer excavation support conditions (i.e. larger
wall and/or strut stiffness, smaller strut spacing) because such support
conditions result in smaller soil movements.
4. Pile bending moment increases with increasing pile diameter, due to its larger
stiffness (for a solid pile) and pile deflection tends to decrease slightly with
pile diameter but generally follows the soil movement unless the pile is very
stiff.
Based on the parametric studies which they studied, Poulos and Chen (1997)
found that the maximum pile bending moment and deflection can be approximated by
using the following equations:

Where ,
M is the maximum bending moment (kNom), pmaxis the maximum
deflection (mm), Mb, pb are the basic bending moment and deflection respectively, k ,
k',, are the correction factors for undreamed shear strength, kd, k'd are the correction
factors for the pile diameter, kNc,k'Ncare the correction factors for the excavation
depth, k E l , k'EIware the correction factors for wall stiffness, kk,k'k are the correction
factors for strut stiffness and k , k', are the correction factors for the strut spacing.
Figures 5.2, 5.3 and 5.4 present the variations of Mb,pband of all the correction factors
respectively. It should be noted that all these values correspond to stability factor

Nc=3, where N, = -(h is the excavation depth).

cu
In figure 5.2 we can see that according to Poulos and Chen (1997), the influence
of the distance between the pile and the excavation face X is significant until X=lOm
and then it keeps decreasing until X=l8m. Moreover the increase of this distance from
2m to 4m has no influence on the value of the maximum deflection of the pile. This
recommendation is verified by the results presented in the previous chapter.

0
0 4 8 12 16 21
Distance from excavation face, X (m)

0
0 4 8 12 16 20
Distance fmm oxcavarion h e , X (m)

Figure 5.2 Basic Bending Moment and Basic Deflection versus Distance from Diaphragm Wall
 1
El,r 40. MI%

Figure 5.3 Correction Factors for Bending Moment (Poulos and Chen, 1997)
 Oh%
0.1 I 10 100
1.6 2
El, x IO' (kNm2)

Figure 5.4 Correction Factors for Deflection (Poulos and Chen, 1997)
5.2 Application of the Method for the 42m long Piles

In principle, the method previously described in section 5.1 can be applied to the
end-bearing 42m long embedded piles presented in section 4.3.
In figure 5.3 we notice that the charts provide predictions of the correction
factors kNcand k'Ncfor a range 0.3<Nc<4.5. For the 30m deep excavation analyzed in
Chapter 4 using an average value of the undrained shear strength c, of the soil profile
(Figure 2.2), we get a very high stability factor, Nc = 17.

PILE
'A

Figure 5.5 Estimation of active and passive pressures below the excavation base

If the JGP layer inside the excavation is considered (Figure 5.5), then the
difference between the active and passive pressures is:
Where A, P are the active and passive pressures, respectively, y is the unit weight of
the soil for depth H, ~,,~l, and S~,JGPare the undrained shear strength of the clay and
the JGP layer, respectively and h is the depth below the excavation where the
diaphragm wall has negligible deformation (Figure 5.4). For the base of the
excavation (h=O) and for equilibrium conditions (A=P), we get:

The factor of safety is equal to:

22
F.S.= -= 1.3
17

Therefore, for the prediction of the maximum bending moment and deflection of
the pile we will assume a value Nc= 4.3/1.3 = 3.4 which is also very close to the value
of the stability factor for which the design charts of Figures 5.2 and 5.3 were
developed.
Tables 5.1 and 5.2 present the predictions of the maximum bending moment and
deflection according to the design charts proposed by Poulos and Chen, for 42m long
piles with 0.4m and lm diameter. The tables show the values of the correction factors
which were used for each case, and the predictions made by this method. The last
column of the tables includes the predictions of the finite element program Plaxis 3-D,
which were discussed in Chapter 4.

Table 5.1 Comparison of Poulos and Chen method with finite element results
for 0.4m diameter piles

2m distance

Pmax F.E.
fb
' k'cu k'd ~ ' N C ~'EIW k' k k's (mm)
(mm)
14 0.99 1 1.3 1 1.22 1.02 22.4 46.5
 4m distance

Mmax FOE-
Mb kcu kd ~ N C ~EIW, kk ks
(mem) ( ~ r ~ * m )
27.5 0.95 0.8 1.4 1 1.2 1.15 40 50

Qmax F.E.
Pb k'cu k'd ~'NC k'Elw k'k k's
(mm) (mm)
13.5 0.99 1 1.3 1 1.22 1.02 21.6 44

10m distance

Mma, F*E*
Mb kcu kd ~ N C ~EIW, kk ks
( ( ~ r ~ - m () m - m )
20 0.99 0.8 1.1 1 0.92 0.97 16 25 ,
Qmax F.E.
Pb k'cu k'd ~ ' N C ~'EIW k'k k's
(mm)
(mm)
10.5 0.97 1 1.2 1 1.22 1.02 15.2 28

Table 5.2 Comparison of Poulos and Chen method with finite element results
for lm diameter piles

2m distance

Qmax F.E.
Pb k'cu k'd ~ ' N C k' EIW k'k k's
(mm)
(mm)
14 0.99 0.96 1.3 I 1.22 1.02 21.5 43.3

4m distance

Mmax FOE*
Mb kcu kd ~ N C ~EIW, kk ks
(m*m) (m-m)
27.5 0.95 11 1.4 1 1.2 1.15 555 967

Qmax F.E.
Pb k'cu k'd ~ ' N C ~'EIW k'k k's
(mm)
(mm)
13.5 0.99 0.96 1.3 1 1.22 1.02 20.7 40
 10m distance

Pmrx F.E.
Pb k'cu k'd ~ ' N C ~ ' E I W k'k k's
(mm) (mm) .
10.5 0.97 0.96 1.2 1 1.22 1.02 14.6 27.2

The results presented in Tables 5.1 and 5.2 indicate that the design charts
proposed by Poulos and Chen (1997) underestimate the deflections and bending
moments indeed by the excavation in Singapore marine clay. This result is due, in
part, to the use of JGP to stabilize the base of the excavation.
According to the design charts presented in Figures 5.2 and 5.3, for the
prediction of the maximum bending moment, the pile to excavation face distance not
only determines the value of the basic bending moment Mb but also influences the
value of all the correction factors. Similarly, for the predictions of the maximum
deflection, the pile to excavation face distance determines the value of the basic
deflection pb and influences the value of three correction factors. This means that
Poulos and Chen regard the pile to excavation face distance to be the most important
parameter which influences the final predictions. This can be confirmed by the finite
element results according to which both the maximum bending moment and the
maximum deflection decrease significantly when the distance is increased to 1Om.
On the other hand, the rigidity of the pile influences only one correction factor,
which is kd for the bending moment and k'd for the maximum deflection. The value of
the correction factor kd has a wide range from 0 to 100, which is much bigger than the
range of the other correction factors. However, the range of the correction factor k'd =
0.8-1.0. Consequently, according to Poulos and Chen (1997) the pile rigidity
influences significantly the maximum bending moment observed and very little the
maximum deflection. This is also verified by the results from the finite element
analysis where we notice that the maximum deflection is slightly reduced when the
pile diameter is increased form 0.4m to 1 m.
6. Conclusions

The thesis has presented results of numerical analyses to evaluate the lateral
deflections and bending moments induced in a single foundation pile due to adjacent
excavation. The analyses are based on a case study derived from related group design
project (MFish, 2006) and focus on ground conditions and excavation support systems
selected for the Circle Line in Singapore. The project comprises 30m deep
excavations in Marine Clay which are supported by 1.2m thick diaphragm walls, 7
levels of cross-lot bracing and a IOm thick jet grout pile raft. The current analyses
used Plaxis 3-D Foundation (v. 1.5, beta) to compute the response of an adjacent
single pile. A short parametric study has compared the effects of the pile embedment
depth, proximity to the diaphragm wall and flexural stiffness.
The main results of the numerical analyses can be summarized as follows:
The maximum bending moment and lateral deflection depend on the pile-
diaphragm wall distance and decrease significantly when this distance
increases. The maximum bending moment is reduced by approximately 50%
when the pile to excavation distance is increased from 2m to 1 Om.
The pile-diaphragm wall distance affects the maximum observed horizontal
deformation on the pile. The maximum deformation of the pile decreases by
approximately 40% when the distance increases from 2m to 1Om.
The diameter of the pile does not play an important role for the lateral
deflections. As it was observed, the increase of the pile diameter from 0.4m to
1.Om slightly reduces the horizontal deformations. It does, however, influence
the maximum bending moment of the piles, but in all cases these were much
lower than the pile's bending moment capacity.
If the piles are unreinforced, cracking is likely to occur, especially at the 1.Om
diameter piles. If the minimum reinforcement suggested by the regulations is
installed, no damaging of the piles is expected to happen.
 Throughout the thesis, it was clear that analyzing the effect of a deep braced
excavation on an adjacent pile by using non-linear finite element methods has many
benefits: First of all, the problem is modelled in detail and all the important
parameters relating soil properties and excavation support system are induced.
Moreover, the model can be established for the site specific conditions. Results of the
numerical solutions for 42m long piles have been compared with design charts
proposed by Poulos and Chen (1997) for estimating the maximum lateral deflections
and bending moments in the pile. For the example geometry considered in this thesis,
the Poulos and Chen (1997) analyses are highly sensitive to the N, factor, estimated
basal stability. In using the Poulos and Chen (1997) method, basal stability is
estimated taking into account the JGP raft. The results suggest that the Poulos and
Chen (1997) method can underestimate substantially the lateral deflections and
bending moments induced in end-bearing piles.
7. References

Finno R.J., Lawrence S.A., Allawh N.F. & Harahap I.S. (1990). "Analysis of
performance of pile groups adjacent to deep excavation", Journal of Geotechnical.
Engineering, ASCE, 117(6).
Goh A.T.C., Wong K.S., Teh C.I. & Wen D. (2003). "Pile response adjacent to braced
excavation", Journal of Geotechnical and Geoenvironmental Engineering, ASCE,
29(4).
Leung C.F., Chow Y.K. & Shen R.F.(2000). "Behavior of pile subject to excavation-
induced soil movement", Journal of Geotechnical and Geoenvironmental
Engineering, ASCE, 126( 1 1).
Plaxis 3-D Foundation Manual (2004). http://www.plaxis.nl
Poulos H.G. & Chen L.T. (1997) "Pile response due to excavation-induced lateral soil
movement", Journal of Geotechnical and Geoenvironmental Engineering, ASCE,
123(2).
Whittle A.J. (2005). "1.364:Advanced Geotechnical Engineering" class notes,
Massachusetts Institute of Technology.
Whittle A.J. & Davies R.V. (2006). "Nicoll Highway Collapse: Evaluation of
Geotechnical Factors Affecting Design of Excavation Support System", To appear
International Conferenceof Deep Foundations, Singapore.
Whittle A.J. & Hashash Y.M.A. (1994). "Soil modelling and prediction of deep
excavation behav iour", International Conference on Deformation Properties of
Geomaterials, Sapporo Japan, Balkema, 589-594.
Appendix A. Calculation of End Bearing Resistance and Skin
Friction
Appendix A presents analytical calculations of the tip resistance, skin friction and
bearing capacity of each pile used in the finite element analyses.

Tip Resistance in Clays

For the estimation of the tip resistance in clays, we use the following equation:

in which the value of Nc can be calculated from the following figure.

Z
~ 1 0 ~ " " ' ' " ' ' " " ' " " " " ' ~
-4d -

Embedment Depth Ratio, d/B

Figure A.1 Values for the N factor

According to the figure, the range of Nc is 7-13. In the following calculations, we will
use Nc=9, which is considered to be a good average for this parameter.

For Pile Length 17m

In this case, the pile lies at the top of the F2 Clay layer, which, at this depth has su=88
kPa (Figure 2.2). At this depth, it is oo= 287kPa.
For Pile L e n ~ t h30m

In this case, the pile lies at the top of the middle Lower MC layer, which, at this depth
has su=39kPa (Figure 2.2). At this depth, it is oo= 5 15kPa.

For Pile L e n ~ t h42m

In this case, the pile lies at the top of the OA Weathered layer, which, at this depth has
s.=100 kPa (Figure 2.2). At this depth, it is oo= 72 1 kPa.

Skin Resistance in the Clays

For the estimation of the skin resistance in the different clay layers, we will use the a-
Methods and P-Methods.

According to this method, the skin friction of the pile is equal to f, = a. su, where suis
the undrained strength of the clay and the parameter a can be derived from the
following figure:

Figure A.2 Values for the parameter a

For each pile, we then calculate the skin friction, according to this method.

For Pile Len~th17m

Upper MC Layer:

At the top of the layer, it is s,=18 kPa, so according to API(1981) the value for
parameter a is equal to 1. Thus, fs = a*s, => fs= 18 kPa.

At the bottom of the Upper MC layer, it is su=25kPa and in this case we have f, = a*
s, => f, = 25 kPa.

For Pile Lenpth 30m

Upper MC Layer:

The value of the skin friction is the same as the previous pile, so we have:

Top of Upper MC: fs= 18 kPa

Bottom of Upper MC: fs= 25 kPa.

F2 Clay Layer:

The value of the undrained shear strength s, is 88kPa and for this value, according to
API(198 1), it is a=O.S.

Thus, the skin friction is equal to fs = a*s, => fs= 44 kPa

Lower MC Layer:

At the top of the layer, it is su=31 kPa, so according to API(1981) the value for
parameter a is equal to 1. Thus, fs = a*s, => f, = 3 1 kPa.

At the tip of the pile (depth 30m) the undrained shear strength of this layer is su=39
kPa and in this case we have fs = a- s, => fs= 39 kPa.
For Pile Lenpth 42m

Upper MC Layer:

The value of the skin friction is the same as the previous piles, so we have:

Top of Upper MC: f, = 18 kPa

Bottom of Upper MC: f, = 25 kPa.

F2 Clay Layer:

The value of the skin friction is the same as the previous pile, so we have

f, = 44 kPa.

Lower MC Layer:

At the top of the layer, it is su=31 kPa, so according to API(1981) the value for
parameter a is equal to 1. Thus, f, = a- su=> fs= 3 1 kPa.

At the bottom of this layer the undrained shear strength of this layer is s U 4 7kPa and
in this case we have f, = a- s, => f, = 47 kPa.

F2 Clay Layer (lower):

The value of the skin friction is the same as the previous pile, so we have

According to this method, the skin friction of the pile is equal to f, = P*o',, where a',
is the effective stress of the layer at the particular depth and the parameter P can be
derived from the following figure. According to Burland(1973), we choose f3=0.25.
 0.0 0.1 0.2 0.3 0.4 0.5 0.t
Parameter

Figure A.3 Values for the parameter P

For Pile Length 17m

Upper MC Layer:

At the top of the layer, it is rr',,=O, so at this point we f,=O.

At the bottom of the Upper MC layer, it is o',,=102 kPa (Figure 2.2) and in this case
we have f, = p*rr9,,=> f, = 25.5 kPa.

For Pile Len~th30m

Upper MC Layer:

The value of the skin friction is the same as the previous pile, so we have:

Top of Upper MC: f, = 0 kPa

Bottom of Upper MC: f, = 25.5 kPa.

F2 Clay Layer:

At the top of the layer, it is a',,=102 kPa, so at this point we f,= p*o',, => f, = 25.5
kPa.
At the bottom of the F2 Clay layer, it is a',,=129 kPa (Figure 2.2) and in this case we
=> fs= 32.25 kPa.
have fs = $-o'~,

Lower MC Layer:

At the top of the layer, it is 0',,=129 => fs =

kPa, so at this point we have fs= $*oYvo
32.25 kPa.

At the tip of the pile (depth 30m) the effective stress is a', = 197 kPa (Figure 2.2)
and in this case we have fs= => fs= 49.25 kPa.

For Pile Lenpth 42m

Upper MC Layer:

The value of the skin friction is the same as the previous pile, so we have:

Top of Upper MC: fs= 0 kPa

Bottom of Upper MC: fs= 25.5 kPa.

F2 Clay Layer:

The value of the skin friction is the same as the previous pile, so we have:

Top of Upper MC: fs= 25.5 kPa

Bottom of Upper MC: fs= 32.25 kPa.

Lower MC Layer:

At the top of the layer, it is a',,=129 kPa, (Figure 2.2) so at this point we have fs=
P-a', => fs= 32.25 kPa.

At the bottom of this layer, the effective stress is a', = 265 kPa (Figure 2.2) and in
this case we have fs= => fs= 66.25 kPa.
F2 Clay Layer (lower):
At the top of the layer, it is 0',,=265 kPa (Figure 2.2), so at this point we have fs=
p*a9,=> fs= 66.25 kPa.

At the bottom of this layer, the effective stress is o', = 285 kPa (Figure 2.2) and in
this case we have fs= p-o'vO=> fS= 71.25 kPa.

Table A. 1 summarizes the skin friction for the three different pile lengths.

Table A.l Skin friction at different soil layers (kPa)

a-Method fl-Method
Layer Top Bottom Top Bottom
18 25 0 25.5
Pile 1 MC
F2 Clay -
Lower
(17m deep) . M c
F2 Clay
-
-
-
Upper ,g 25 0 25.5
Pile 2 MC
F2 Clay 44 44 25.5 32.25
Lower
31 39 32.25 49.25
(30m deep) MC
F2 Clay -
18 25 0 25.5
Pile 3 MC
F2 Clay 44 44 25.5 32.25
Lower
31 47 32.25 66.25
(42m deep) Mc
F2 Clay 44 44 66.25 71.25