Documenti di Didattica
Documenti di Professioni
Documenti di Cultura
Faculty of Engineering
BY
A Thesis
Submitted in Partial Fulfillment
For the Requirements of the Degree of
Doctor of Philosophy
In
Civil Engineering
Supervised by
2009
Supervision Committee
i
Approval Sheet
BY
ii
Acknowledgements
I wish to express my deep thanks and gratitude to Prof. Dr. Mohamed Mosaad
Engineering, Suez Canal University, for his kind supervision and useful
Professor, Civil Eng. Dept., Suez Canal University, for his constructive
criticisms, precious and generous support that have truly helped in bringing
great mother soul who she is in my heart all the time and also to my family
specially my dear wife for their continuos sacrifice and fruitful care.
iii
ABSTRACT
SYNOPSIS
iv
ABSTRACT
exposed to vertical loads and seismic loads are carried out by ASTNII and
ASTNIII respectively. Discussion for the numerical results is performed and
conclusions are drawn.
v
TABLE OF CONTENTS
TABLE OF CONTENTS
Page
CHAPTER 1: INTRODUCTION
1.1. Background....... 1
1.2. Objective of Thesis.... 3
1.3. Layout of Thesis........ 4
v
TABLE OF CONTENTS
Page
2.4.3 The Finite Element Method.............................................. 26
2.4.4 Combined Boundary Element and Finite Element Method 29
2.4.5 Combined Finite Layer and Finite Element Method 30
2.4.6 Variational Approach.. 30
2.5. Lateral Load-Resisting Systems 31
2.5.1. Types of lateral loads. 31
2.5.2 Seismic loads... 31
2.5.3 Distribution of shear forces among stories 34
2.5.4 Types of resisting systems... 35
vi
TABLE OF CONTENTS
Page
4.2.6. Modeling the layered Soil... 63
4.2.7. Influence Matrix for the Soil... 64
4.2.8. Linear analysis of the raft 65
4.3. Analysis single pile and pile groups.. 66
4.3.1. Modeling single pile in half-space soil (vertical loads).. 66
4.3.2. Modeling the layered soil 68
4.3.3. Soil flexibility for a single pile 68
4.3.4. Linear analysis. 68
4.3.5. Nonlinear analysis... 70
4.3.6. Modeling pile groups... 72
4.3.7. Soil stiffness for freestanding raft (pile groups under vertical
loads)... 72
4.4. Modeling piled raft 75
4.4.1. Linear analysis 79
4.4.1.I. Analysis of piled flexible raft.. 79
4.4.1.II. Analysis of piled rigid raft. 79
4.4.1.III. Analysis of piled elastic raft... 80
4 4.4.2. Nonlinear analysis. 82
4 4.4.3. Iterative Procedure 82
4.5. The pile model against lateral loads.. 84
4.5.1. Modeling of the layered soil... 85
4.5.2. Soil stiffness for pile groups against lateral
forces. 86
4.5.3. Linear analysis 87
4.5.4. Nonlinear analysis.. 89
vii
TABLE OF CONTENTS
Page
4.5.6. Limiting pile-soil stresses... 89
4.5.6.1. Cohesive soil. 90
4.5.6.2. Cohesionless soil 90
4.6. Soil-structure interaction... 90
4.6.1. Effect of superstructure on foundations.. 91
4.6.2. Piled raft-3D superstructure interaction.. 93
viii
TABLE OF CONTENTS
Page
Gendy,2007). 134
5.3.5.1. Soil properties.. 135
5.3.5.2. Raft and pile material.. 136
5.3.5.3. Analysis of the piled raft. 136
5.3.5.4. Comparison with three-dimensional finite element
analysis and field measurements.. 136
5.3.5.5. Comparing among different analysis types. 138
5.3.6. Test problem 6: Pile Group under General Loading
Conditions (Basile, 2003).. 142
5.3.7. Test problem 7: Lateral load distribution (Basile,2003). 150
5.4. conclusion.. 151
ix
TABLE OF CONTENTS
Page
6.4.1.2. Moment 162
6.4.1.3. Pile load... 162
6.4.1.4. Bearing factor.. 162
6.4.2. Effect of pile diameter 163
6.4.2.1. Settlement and differential settlement. 163
6.4.2.2. Moment 164
6.4.2.3. Pile load... 164
6.4.2.4. Bearing factor 164
6.4.3. Effect of varying pile length. 165
6.4.3.1. Settlement and differential settlement. 165
6.4.3.2. Moment 166
6.4.3.3. Pile load... 166
6.4.3.4. Bearing factor.. 167
6.4.4. Effect of superstructure.. 167
6.4.4.1. Settlement and differential settlement 167
6.4.4.2. Moment 168
6.4.4.3. Pile load... 168
6.4.4.4. Bearing factor.. 168
6.5. Conclusions 183
x
TABLE OF CONTENTS
Page
7.3.1. Description, geometry and properties of models... 186
7.3.2. Soil properties 186
7.3.3. Modulus of compressibility... 187
7.3.4. Limited shaft friction. 188
7.3.5. Concrete material properties.. 188
7.4. Analysis. 188
7.5. Results and discussion... 190
7.5.1. Effect of pile length 190
7.5.1.1. Vertical settlement and differential settlement. 190
7.5.1.2. Lateral settlement and drift ratio... 191
7.5.1.3. Moment. 193
7.5.1.4. Vertical pile load... 193
7.5.1.5. Lateral pile load 194
7.5.1.6. Bearing factor... 194
7.5.2. Effect of pile diameter... 194
7.5.2.1. Vertical settlement and differential settlement. 194
7.5.2.2. Lateral settlement and drift ratio... 195
7.5.2.3. Moment. 197
7.5.2.4. Vertical pile load... 197
7.5.2.5. Lateral pile load 197
7.5.2.6. Bearing factor 198
7.5.3. Effect of varying in pile length.. 198
7.5.3.1. Vertical settlement and differential settlement. 198
7.5.3.2. Lateral settlement and drift ratio... 200
7.5.3.3. Moment. 201
7.5.3.4. Vertical pile load 202
xi
TABLE OF CONTENTS
Page
7.5.3.5. Lateral pile load 202
7.5.3.6. Bearing factor... 203
7.5.4. Effect of raft thickness... 203
7.5.4.1. Vertical settlement and differential settlement 203
7.5.4.2. Lateral settlement and drift ratio... 204
7.5.4.3. Moment. 204
7.5.4.4. Vertical pile load... 205
7.5.4.5. Lateral pile load.... 205
7.5.4.6. Bearing factor... 205
7.6. Conclusions... 206
REFERENCES... 240
APPENDIX I Displacement at the center due to a cylindrical
surface stress..... 252
APPENDIX II Displacement at the center due to a circular uniform
load 257
APPENDIX III Lateral displacement at the center due to a
rectangular uniform load........................ 259
xii
LIST OF TABLES
LIST OF TABLES
xiii
LIST OF FIGURES
LIST OF FIGURES
Figure (2.1) Stresses on piles and the soil for a two pile group
under the axial load 36
Figure (2.2) Stresses acting on piles and soil for a two pile
group under lateral loads 36
Figure (3.1) Port-Said zones 39
Figure (3.2) Main soil properties 43
Figure (3.3) Modulus of compressibility with depth for all
46
zones
Figure (3.4) Main soil profile of Port-Said 48
Figure (3.5) Typical soil properties of Port-Said area 49
Figure (4.1) Framework representation of plate (Lattice of
96
plate)
Figure (4.2) Plate element and equivalent Framework model 96
xiv
LIST OF FIGURES
Figure (5.8-a) Load capacity of each pile = 875 [kN] Pile and
127
load configuration
Figure (5.8-b) Model condition and material properties 127
Figure (5.9-a) Comparison of various methods for load-settlement
128
analysis
Figure (5.9-b) Effect of raft thickness on maximum bending
128
moment
Figure (5.9-c) Effect of raft thickness on maximum settlement 129
Figure (5.9-d) Effect of raft thickness on diffrential settlement 129
Figure (5.10-a) Distribution of shear stress along 130
xv
LIST OF FIGURES
xvi
LIST OF FIGURES
xvii
LIST OF FIGURES
Figure (6.2) Typical floors for the two models with column
arrangement 154
xviii
LIST OF FIGURES
xix
LIST OF FIGURES
xx
LIST OF FIGURES
Figure (6.29) Bearing factor for pile diameter = 0.8 [m] 181
xxi
LIST OF FIGURES
Title Page
Figure No.
Figure (7.5) Lateral settlement x-direction at the center of the
building for pile length =20 [m] 209
Figure (7.6) Lateral settlement y-direction at the center of the
building for pile length =20 [m] 209
Figure (7.7) Vertical settlement at the center of the building
when varying pile length 210
Figure (7.8) Lateral settlement x-direction at the center of the
building when varying pile length
210
Figure (7.9) Lateral settlement y-direction at the center of the
building when varying pile length
211
Figure (7.10) Vertical settlement at the center of the building
when varying raft thickness 211
Figure (7.11) Lateral settlement x-direction at the center of the
building when varying raft thickness
212
Figure (7.12) Lateral settlement y-direction at the center of the
building when varying raft thickness
212
Figure (7.13) Vertical differential settlement of the building for
pile diameter =0.8 [m] 213
Figure (7.14) Drift ratio in x-direction of the building for pile
diameter =0.8 [m] 213
Figure (7.15) Drift ratio in y-direction of the building for pile
diameter =0.8 [m] 214
Figure (7.16) Vertical differential settlement of the building for
pile length =20 [m] 214
xxii
LIST OF FIGURES
Page
Figure No. Title
Figure (7.17) Drift ratio in x-direction of the building for pile
length =20 [m] 215
Figure (7.18) Drift ratio in y-direction of the building for pile
length =20 [m] 215
Figure (7.19) Vertical differential settlement of the building with
varying pile length 216
Figure (7.20) Drift ratio in x-direction of the building with
varying pile length 216
Figure (7.21) Drift ratio in y-direction of the building with
varying pile length 217
Figure (7.22) Vertical differential settlement of the building with
varying raft thickness 217
Figure (7.23) Drift ratio in x-direction of the building with
varying raft thickness 218
Figure (7.24) Drift ratio in y-direction of the building with
varying raft thickness 218
Figure (7.25) Moment in the raft at section I-I for square-shape
model for pile diameter = 0.8 [m] 219
Figure (7.26) Moment in the raft at section I-I for square-shape
model for pile length = 20 [m] 219
Figure (7.27) Moment in the raft at section I-I for square-shape
model with varying in pile length 220
Figure (7.28) Moment in the raft at section I-I for square-shape
model with varying in raft thickness 220
xxiii
LIST OF FIGURES
xxiv
LIST OF FIGURES
xxv
CHAPTER 1 INTRODUCTION
CHAPTER 1
INTRODUCTION
1.1. Background
For most piled raft foundations, the primary purpose of the piles is to act as
settlement reducers. The proportion of load carried by the piles is considered
as a secondary issue in the design. Over the past decades, extensive research
work has been presented, aimed at improving the accuracy in the prediction
of the behavior of piled rafts. In the design of piled rafts, design engineers
have to understand the mechanism of load transfer from the raft to the piles
and to the soil to predict (i) the behavior of the raft which includes the
settlements, bending moments and the proportion of load carried by the raft,
and (ii) the behavior of the piles which includes the displacements and load
distributions along the piles. Interactions between piles, raft and soil are of
major concern in the analysis. The concept of interaction between piles
introduced by Poulos (1968) was used in the analysis of pile groups and can
be extended to the analysis of piled rafts.
Analyzing piled raft is a complex task because it is a three-dimensional
problem including many capabilities. The main capabilities that must be
considered in the analysis are: the interaction between all piled raft and soil
elements; taking into account the actual loading and geometry of the piled
raft; representing the soil by a real model and treating the problem as
nonlinear analysis. Considering all these capabilities requires great
experience and effort. Besides such a problem requires long computational
time where huge size soil matrix is required for a large piled raft due to
discretized nodes along piles and under the raft. For these reasons many
1
CHAPTER 1 INTRODUCTION
2
CHAPTER 1 INTRODUCTION
3
CHAPTER 1 INTRODUCTION
building as a small tilt can mean a large horizontal movement at the top of
the building. Most of the recent research has been performed for piled rafts
with piles of identical size.
The objective of this thesis is to develop a numerical method for the analysis
of piled rafts with piles of different lengths and diameters by the use of the
finite layer method for the analysis of the layered soil and the finite element
method for the analysis of the piles and the raft. This new method has the
following features:
1) Both horizontal and vertical loadings can be applied to the piled raft.
2) The piled raft is analyzed as a whole structure by attaching the piles to
the nodes on the raft.
3) Piles can have different lengths and diameters and the raft can have
different thicknesses and any shape.
4) Applied moments are directly transferred from the raft to the pile
heads.
5) This model has the ability to save the time and save the memory of the
computer.
Effects of the dimensions of piles on different kinds of interaction and on the
overall behavior of the piled raft subjected to horizontal and vertical
loadings will be examined by the use of the new method developed in this
thesis. Results obtained from this new method are compared with those from
existing methods and also with field measurements.
4
CHAPTER 1 INTRODUCTION
Chapter three: states the used soil properties of Port-Said and the method
of deducing the modulus of soil elasticity and modulus of the soil
compressibility .
Chapter four: presents the numerical models used in the present study to
analyze the piled raft-superstructure. Finite Element models, Framework
Analogy method, Composed coefficient technique and Equivelant stiffness
technique which are used to simulate the different elements.
Chapter six: a parametric study using the proposed method of analysis for
different construction models on Port-Said soil meduim is carried out. It
demonstrates the behavior of piled raft-structure resting on nonlinear elastic
medium and subjected to vertical loads using Composed coefficient
technique.
5
CHAPTER 1 INTRODUCTION
6
CHAPTER 2 LITERATURE REVIEW
CHAPTER 2
LITERATURE REVIEW
2.1. Introduction
Piled raft has attracted many researchers during the last four decades. Early
researches focused on hand calculation techniques with the help of empirical
charts and formulas for single piles and pile groups. With the advent of
computer and numerical procedures, Finite Element techniques were
developed to solve piled foundation. The different analysis methods can be
classified into several categories: empirical, analytical and numerical
methods. In this chapter, a brief review of the techniques developed for the
analysis of raft, pile group and piled raft are presented. Also, a short brief
about seismic forces has been introduced.
2.2. Rafts
In the design of rafts, the soil may be treated as one of the following
methods.
a) A series of individual springs known as Winkler model.
b) A continuum model.
The Winkler model provides sufficient accuracy with reasonable time of
solution and storage requirements. The soil is represented by an infinite
number of isolated elastic springs. The deflection of the soil medium si at the
point i on the surface is directly proportional to the soil contact pressure qi at
that point regardless of the pressure at other locations. However, it neglects
the interaction between individual springs and therefore, the soil medium is
not modeled as a continuum soil. The other approach, which considers the
7
CHAPTER 2 LITERATURE REVIEW
supporting soil as elastic continuum, can represent the soil more accurate
than the first one. The soil parameters used in this approach depend on the
field stress state as Hain and Lee (1974).
E (1 2 ) t r
3
K= r ( 2.1)
Es a
where
Er denotes Young's modulus of the raft.
Es is the soil modulus.
vs is Poisson's ratio of the soil.
t, is the raft thickness.
a is the raft radius.
8
CHAPTER 2 LITERATURE REVIEW
9
CHAPTER 2 LITERATURE REVIEW
10
CHAPTER 2 LITERATURE REVIEW
11
CHAPTER 2 LITERATURE REVIEW
12
CHAPTER 2 LITERATURE REVIEW
13
CHAPTER 2 LITERATURE REVIEW
an atypical pile cap, which consisted of a large, deep circular embedded cap.
These studies indicate that the lateral resistance of pile caps can be quite
significant, especially when the pile cap is embedded beneath the ground
surface. There is clearly a need for a rational method to evaluate the
magnitude of the pile cap resistance, and for including this resistance in the
design of pile groups to resist lateral loads.
14
CHAPTER 2 LITERATURE REVIEW
15
CHAPTER 2 LITERATURE REVIEW
16
CHAPTER 2 LITERATURE REVIEW
In the analysis, each pile was divided into a number of cylindrical elements.
Each element was subjected to a uniform load around the periphery of the
element and a uniform circular load at the circular base of the pile as shown
in Fig. 2.1. The shaft of the pile was assumed to be perfectly rough while the
base was assumed to be perfectly smooth such that the shear stresses were
not developed on the base. The displacement factors were obtained by
integration of the Mindlin equation for vertical displacement due to a point
load within a semi-infinite soil mass. By considering the compatibility of the
vertical displacement The distribution of shear stress can be obtained along
the pile shaft and the vertical stress on the pile base and subsequently the
displacement of the pile can be determined. For a group of piles, the
displacement of an individual pile in the group was obtained by
superposition. Poulos further extended the method to laterally loaded pile
groups by introducing the interaction factor for rotation, which is defined
as following:-
17
CHAPTER 2 LITERATURE REVIEW
18
CHAPTER 2 LITERATURE REVIEW
19
CHAPTER 2 LITERATURE REVIEW
20
CHAPTER 2 LITERATURE REVIEW
21
CHAPTER 2 LITERATURE REVIEW
soil component was subjected to the interaction forces along the pile-soil
interface which were represented by a series of ring loads on the soil nodes
and a uniform load on the base. The displacements at all soil nodes, which
from the soil flexibility matrix, were determined by applying a unit load at
each soil node in turn. The soil stiffness matrix was obtained by inverting
the flexibility matrix. Combining the pile and soil stiffness formed the
stiffness matrix for the system and displacements of the pile group were then
obtained by solving a set of stiffness equations. Lee and Small (1991)
modified this approach for the analysis of laterally loaded piles by replacing
the vertical loads acting along the shaft and base with horizontal loads.
Southcott and Small (1996) extended the approach of Lee and Small (1991)
to the analysis of vertically loaded pile group but the loads applied to soil
nodes, which were assumed to be a series of uniform annular nodes rather
than ring loads.
Zhang and Small (1991) proposed two methods based on the finite layer
theory to analyze axially and laterally loaded pile groups embedded in
homogeneous and nonhomogeneous soils. The principle of the methods is
similar to that used by Lee and Small (1991). The finite layer theory was
employed for the layered soil and simple beam theory for the piles. The piles
were divided into a series of finite elements and the soil was divided into
corresponding layers. Interaction and stiffness methods were developed to
generate the influence of matrices for the soil and the pile group. In the
interaction method, each pair of piles in the pile group was considered in
turn to compute the soil influence matrices The soil influence matrix was
formed by applying unit ring loads ( to each node along the shaft in turn) or
a circular load ( to the base node ) to compute the displacement at each node
of the layered soil . The pile influence matrix was formed by pinning the top
of the pile to stop rigid body rotations and translations of the pile and a
22
CHAPTER 2 LITERATURE REVIEW
similar method was used to form the soil influence matrix .For the stiffness
method, the soil stiffness matrix was formed by first generating the soil
influence matrix. The influence matrix was formed by applying a unit ring
load or circular load (base node) to each node on all piles in turn to obtain
the displacement of the soil. The soil stiffness matrix was obtained by
inverting the soil influence matrix. The pile stiffness matrix was generated
by assembling the stiffness matrix of all the elements of the pile.
Comparison of results between the two methods has shown that the stiffness
method is the most accurate method. However, it is not practicable for large
pile groups, as it requires a large amount of computer memory. The
shortcoming of the interaction method is that the formation of the pile
influence matrix only considers the interaction between a pair of piles and
ignores the effects of other piles in the pile group. Therefore, the stiffness
method is more suitable for small pile groups with any pile spacing while the
interaction method is suitable for large pile groups with large pile spacing.
Zhang and Small (2000) further extended the method to include the pile cap
in the analysis. The analysis was separated into three parts: the cap, the piles
and the layered soil. The cap was assumed a thin elastic plate and analysed
by the finite element method. The element division of the cap was such that
the pile head fitted within one element of the cap. In the analysis of the cap,
the cap has to be restrained from rigid body rotations and translations by
pinning two corner nodes. The influence matrix of the pinned cap was
generated by applying a uniform horizontal or vertical load to each of the
cap elements in turn to obtain the central displacements of all cap elements.
The influence matrices of the piles and the layered soil were obtained from
the iterative procedure presented by Zhang and Small (2000). By considering
the equilibrium of interaction forces and compatibility of displacements
23
CHAPTER 2 LITERATURE REVIEW
between the pile heads and the cap, the contact stresses can then be obtained
and displacement of the raft can then be computed by solving the equations.
One of approaches that treated the raft as a thin plate, the piles as springs and
the soil as an elastic continuum, was used by Hongladaromp et al. (1973) in
which the interaction effects between the piles were ignored. Poulos (1980)
developed a program GARP (Geotechnical Analysis of Raft with Piles)
which employed a finite difference method for the raft with the
consideration of the interaction effects between the piles and raft. Russo
(1998) employed a similar method where the piles and soil were modeled by
linear or non-linear interacting springs. The soil displacements were
calculated using Boussinesqs solution thus yielding a closed form solution.
The non-linear behavior of the piles was modeled by the assumption of a
hyperbolic load-settlement curve for a single pile.
Randolph (1983) presented a method to compute the interaction between a
single pile and a circular raft. A flexibility matrix method was then used to
calculate the overall stiffness of the piled raft foundation by combining the
individual stiffness of a single pile-raft unit. Clancy and Randolph (1993)
employed a hybrid method which combined finite elements and analytical
solutions. The raft was modeled by two-dimensional thin plate finite
elements, the piles were modeled by one-dimensional rod finite elements
and the soil response was calculated by using an analytical solution. This
method took into account the non-linearity the behavior of the pile and the
slip was allowed to occur at the pile-soil interface. However, this method is
limited to homogeneous soil conditions. Kitiyodom and Matsumoto (2003)
presented a similar approach to Hain and Lee (1978), but the piles were
24
CHAPTER 2 LITERATURE REVIEW
25
CHAPTER 2 LITERATURE REVIEW
between the piled raft and a soil subjected to externally imposed vertical
movement. The analysis is implemented via a computer program PRAWN
(Piled Raft with Negative Friction).
Mendona and de Paiva (2000) presented a boundary element method for
the analysis of piled rafts in which full interaction between the raft, piles and
the soil is considered. Unlike the other approaches, discretization of the
foundation system was not required in this approach. The soil was
represented by a Mindlin elastic linear homogeneous half space. The raft was
assumed a thin plate and was represented by integral equations. The pile was
represented by a single element and the shear stresses along it were
approximated by a second-degree polynomial. The interaction between the
raft and soil was analyzed by dividing the interface into triangular elements
and the subgrade reaction was assumed to vary linearly across each element.
26
CHAPTER 2 LITERATURE REVIEW
27
CHAPTER 2 LITERATURE REVIEW
28
CHAPTER 2 LITERATURE REVIEW
29
CHAPTER 2 LITERATURE REVIEW
30
CHAPTER 2 LITERATURE REVIEW
used by Shen and Teh (2000) which has incorporated lateral loading into the
analysis. The piles were modeled as flexible beams and the non-linear
behavior of the pile head was approximated in the elastic analysis by a
reduction in the soil modulus near the ground surface. An extension of the
above method for vertically loaded pile groups was made by Chow et al.
(2001) for the analysis of piled raft foundations.
Liang and Chen (2004) presented a modified variational approach based on
the above method. In their modified approach, the flexible raft was isolated
from the piled raft foundation and the deformation of the raft was
represented by an analytical function. The base of the raft was assumed to be
smooth and the connection between the raft was assumed to be a sliding ball
joint. Interactions between the raft, piles and soil were described by an
approximate closed-form analytical method.
31
CHAPTER 2 LITERATURE REVIEW
following paragraphs.
Fb = Sd(T1). W / g (2.3)
Where, W is the weight of the structure used to calculate the seismic load.
Sd(T1) is the horizontal design spectrum for elastic analysis, is the
corrective factor and g is the earth acceleration and equals to 9.81 [m/s2].
The type of soil is an effective factor affects on Fb as well as other factors
like the type of the construction, the geographic zone according to the
earthquake map,
32
CHAPTER 2 LITERATURE REVIEW
the earthquake such as hospitals, power plant, nuclear facilities, fire and
police stations. Values given in the Egyptian Building code are listed below:
- Essential facilities I=1.2~1.4
- All others I=.8~1.2
33
CHAPTER 2 LITERATURE REVIEW
T1 = Ct H 3 / 4 (2.4)
34
CHAPTER 2 LITERATURE REVIEW
zW
Fi = n .Fb
i
(2.5)
z jW j
j =1
Where, zi, zj are the height of mi, mj calculated from the level of foundation.
35
CHAPTER 2 LITERATURE REVIEW
P P
S
Pj
Pb Pb
Pb Pb
Pile 1 Pile 2
(a) Stresses on piles (b) Stresses on soil
Fig.2.1 Stresses on piles and the soil for a two pile group under the
axial load
H, M d
d
S H, M
Fig.2.2 Stresses acting on piles and soil for a two pile group under
lateral loads
36
CHAPTER 3 PORT-SAID SOIL PROPERTIES AND PARAMETERS
Chapter 3
3.1. Introduction
Port-Said lies on the eastern side of the Nile Delta at the north end of the
Suez Canal on the Mediterranean Sea. Most dry land in Port-Said has been
reclaimed except for a narrow beach which separated Lake Manzala from
the sea. The material used in reclamation came from the Suez Canal and
harbor excogitations and raised the ground surface to its present height of
some 1 to 2 m above sea level. This thesis will focus on the study of piled
raft 3D-space buildings with different shapes exposed to both vertical loads
and seismic loads due to construction in Port-Said, taking into account the
soil data in the reclaimed zones as source data. This source data has been
based upon the extensive geotechnical study performed by Golder
Associates (1979).
37
CHAPTER 3 PORT-SAID SOIL PROPERTIES AND PARAMETERS
38
CHAPTER 3 PORT-SAID SOIL PROPERTIES AND PARAMETERS
39
CHAPTER 3 PORT-SAID SOIL PROPERTIES AND PARAMETERS
- Fill
- Surface clay
- Sand stratum
- Transition zone
- Lower clay
- Basal deposits
Table 3-1 shows the thickness of each main layer in the reclaimed zones.
The next paragraphs describe the soil layers in details according to EL Azab
(1988).
Surface clay
40
CHAPTER 3 PORT-SAID SOIL PROPERTIES AND PARAMETERS
The surface clay is typically grey, very soft and highly plastic. Sand and silt
are present in varying proportions such that in some areas it could be
classified as sandy silty clay and fine sand. It has high silt content that when
dry it appears to be silt.
The surface clay contains occasional layers of fine sand ranging up to 0.20
[m] in thickness. Shell fragments embedded in clay are frequently
concentrated in brands up to 0.15 [m] thick. The sand is micaceous. In some
areas there are thin horizontally stratified sand and silt layers.
Sand stratum
The sand stratum is under the surface clay. The thickness of the sand varies
between 4.2 [m] and 10 m with an average for whole site of 6.5 [m].
Transition zone
The transition zone is under the sand stratum, at an elevation of about 7 m.
thin layers of silt and clay occur in the sand stratum. This marks the start of
the gradational change from the sand stratum to the lower clay. The layers of
clay increases in frequency until the soil are predominantly clay with some
sand layers. They increase in thickness and frequency towards the base of
the stratum. These bands of sand and clay have been termed the transition
zone.
Lower clay
41
CHAPTER 3 PORT-SAID SOIL PROPERTIES AND PARAMETERS
The lower clay is firm becoming stiff, dark grey, organic and micaceous. Its
structure characteristics are laminations and fissuring. The laminar structure
is more dominant at the top of the clay deposit than at the base. To about
elevation 22 [m], there are frequent thin bands and parting of sand and silt.
At greater depth there are occasional pockets of very finely laminated clay.
The bottom of the lower clay was reached at elevation ranging between 40
[m] to 52 [m].
42
CHAPTER 3
Sand stratum
mS mS mS mS mS
10.00
7.40 8.30 9.10 7.50 7.60
The layers of clay increases in frequency until the soil is
U a clay with some sand layer U U U U 15.00
11.00 12.00 14.50 11.50 10.50
20.00
25.00
30.00
35.00
T
40.00
41.50
Firm becoming stiff, dark grey, organic and micaceous
T T
structure characteristics are laminations and fissuring 45.00
47.00 46.50
T
T 50.00
51.50
52.50
43
55.00
PORT-SAID SOIL PROPERTIES AND PARAMETERS
Modules of compressibility:
Analysis of foundation using continuum model required to obtain the
modules of compressibility of the clay Es or modulus of Elasticity E as a
main soil parameter. Considering the available water content for clay layers
from Golder Associates (1979), this modules can be presented in general
equations related to the depth.
44
CHAPTER 3 PORT-SAID SOIL PROPERTIES AND PARAMETERS
1 Cc +
H = log o H (3.3)
Es (1 + eo ) o
where:
Es Modulus of compressibility of Port-Said clay, [kN/m2].
Average vertical stress increase in clay, [kN/m2].
H Layer thickness, [m].
o Initial overburden pressure in a layer, [kN/m2].
Reda (2009) had determined the modulus of compressibility for each zone
and verified it. He had found that the difference in Es for all zones is small
and may be neglected. Values of Es for all zones in Port-Said are plotted in
the 0. Fitting values of Es is shown in the same figure and indicates that Es
increases with depth and can be approximated by the following linear
relation:
45
CHAPTER 3 PORT-SAID SOIL PROPERTIES AND PARAMETERS
2
Modulus of compressibility E s [kN/m ]
0 2000 4000 6000 8000 10000
0
10
Depth under the surfacez [m]
20
30
Zone 1
Zone 2
40
Zone 3
Zone 4
50 Zone 5
Fitting
60
Modules of Elasticity:
Modulus of elasticity E [kN/m2] of the soil is given by:
1 s 2 s2
E= Es = 0.83Es (3.5)
1 s
46
CHAPTER 3 PORT-SAID SOIL PROPERTIES AND PARAMETERS
Groundwater:
Groundwater in Port-Said lies in within 2 [m] from the ground surface. The
groundwater level is assumed to be lie directly below the raft.
47
CHAPTER 3 PORT-SAID SOIL PROPERTIES AND PARAMETERS
U, Silt 20.00
T, Clay
25.00
30.00
35.00
Lower clay
T firm becoming stiff, dark grey, organic and micaceous 50.00
50.0
48
CHAPTER 3 PORT-SAID SOIL PROPERTIES AND PARAMETERS
B1
A
A
GW 2.0 2.0
5.00
Es = 72289[kN/m2]
mS Ws=72289 [kN/m2]
Gam=8[kN/m2]
8.00
Es = 7831[kN/m2]
U Ws=7831 [kN/m2] 10.00
Gam=8[kN/m2]
12.00
Es = 2866[kN/m2]
T Ws=2866 [kN/m2]
14.50 Gam=8[kN/m2]
15.00
A A, Filling
Es = 3339[kN/m2]
T Ws=3339 [kN/m2]
17.50 Gam=8[kN/m2]
mS, Medium sand
Es = 3810[kN/m2]
T Ws=3810 [kN/m2] 20.00
U, Silt 20.50 Gam=8[kN/m2]
Es = 4281[kN/m2]
T Ws=4281 [kN/m2]
T, Clay 23.50 Gam=8[kN/m2]
Es = 4753[kN/m2]
T Ws=4753 [kN/m2] 25.00
26.50 Gam=8[kN/m2]
Es = 5224[kN/m2]
T Ws=5224 [kN/m2]
29.50 Gam=8[kN/m2]
Es = 5695[kN/m2] 30.00
T Ws=5695 [kN/m2]
32.50 Gam=8[kN/m2]
Es = 6167[kN/m2]
T Ws=6167 [kN/m2]
Gam=8[kN/m2]
35.00
35.50
Es = 6639[kN/m2]
T Ws=6639 [kN/m2]
38.50 Gam=8[kN/m2]
40.00
Es = 7110[kN/m2]
Ws=7110 [kN/m2]
Gam=8[kN/m2]
T
50.00 50.00
49
CHAPTER 4 ANALYSIS OF PILED RAFT IN LAYERED SOIL
CHAPTER 4
4.1. Introduction
The analysis of a raft foundation involves the separation of the foundation
system into an isolated raft and the supporting soil and takes account of the
interactions between the soil and the raft. The distribution of contact
pressures acting on the interface between the raft and the soil depends on the
rigidity of the raft. For pile groups, the analysis may be performed by
separating the system into a pile group and the supporting soil. In a similar
way to the raft analysis, the full soil-structure interactions have to be taken
into account.
Finite element techniques have often been used for the analysis of rafts by
different researchers such as Cheung and Zienkiewicz (1965), Cheung and
Nag (1968), and Svec and Gladwell (1973). The raft can be treated as a plate
and the soil can be treated as (I) a series of isolated springs known as a
Winkler model, in which the contact pressure at any point on the base of the
raft is proportional to the deformation of the soil at that point or (II) as an
elastic half-space (or elastic continuum) in which the behavior of the soil can
be obtained from a number of closed-form solutions. For a multilayered soil
system, Wardle and Fraser (1974), Fraser and Wardle (1976) used the
Finite element technique to analyze a rectangular raft subjected to uniformly
distributed loads. The finite layer method was used by Zhang and Small
(1991) for the analysis of rafts on an elastic half space subjected to different
types of loadings. The behavior of pile groups has been examined
extensively by the use of different techniques. The most widely used
technique is the interaction method which was introduced by Poulos (1968)
50
CHAPTER 4 ANALYSIS OF PILED RAFT IN LAYERED SOIL
to calculate the displacements between two piles and hence the pile group.
Other techniques including the boundary element method (Butterfield and
Banerjee, 1971a and b; Banerjee and Davies, 1977) and the Finite element
method (Ottaviani, 1975; Valliappan, et al. 1974; Pressley and Poulos,
1986; Chow, 1987a) have been used for settlement analysis. The finite layer
method (Lee and Small, 1991a and b; Ta and Small, 1995; Southcott and
Small, 1996; Zhang and Small, 1999 and 2000) was used for the analysis of
pile groups subjected to vertical or horizontal loadings. Moreover,
Steinbrenner employed an approximation technique to solve such problems
of finite layer by Finite element method based on Mindlins equations.
In this chapter, the analysis of raft and pile groups through the use of
Composed coefficients technique and Finite element methods will be
presented. Loadings and moments in different directions can be applied to
the foundation. The soil behavior is modeled by the use of Finite layer
techniques and the raft and piles are modeled by Finite element techniques.
51
CHAPTER 4 ANALYSIS OF PILED RAFT IN LAYERED SOIL
52
CHAPTER 4 ANALYSIS OF PILED RAFT IN LAYERED SOIL
k a M1
1 = 3
(4.1)
h
E( )
12
where, a and k a are the side lengths of the element as shown in Fig. 4.2.a., h
is its thickness, and E is the elastic modulus of the material. The rotation of
the element in the orthogonal direction is given as follows:
a M1
2 = 3
(4.2)
h
E( )
12
a M2
3 = 3
(4.3)
h
E( )
12
and,
k a M2
4 = (4.4)
h3
E( )
12
k a H (1 + )
5 = (4.5)
h3
E( )
12
53
CHAPTER 4 ANALYSIS OF PILED RAFT IN LAYERED SOIL
M1 a 2 K r 3Is + I d
6 = (4.6)
2E r 3Is Ic + I d Ic + K 3Id Is
and,
M1 a 2 K2 Id
7 = (4.7)
2E r Is I c + Id Ic + K 3Id Is
3
M 2a 2 K r 3Ic + K 3Id
8 = (4.8)
2E r 3Is I c + Id Ic + K 3Id Is
and
M 2a 2 K 3 Id
9 = (4.9)
2E r Is I c + I d I c + K 3I d Is
3
54
CHAPTER 4 ANALYSIS OF PILED RAFT IN LAYERED SOIL
H r3 a2 K
10 = (4.10)
GJ
2E r 3 c + 2KI d
E
w4 = -a 10 (4.11)
1 = 6 (4.12.a)
2 = 7 (4.12.b)
3 = 8 (4.12.c)
4 = 9 (4.12.d)
5 = 10 (4.12.e)
(K 2 ) a h 3
IS = . (4.13)
2K (1 2 ) 12
(1 K 2) a h 3
IC = . (4.14)
2 (1 2 ) 12
r3 a h3
Id = . (4.15)
2K (1 2 ) 12
55
CHAPTER 4 ANALYSIS OF PILED RAFT IN LAYERED SOIL
The torsional stiffness is obtained from Equation 4.10 and may be given by
the following Equations:
GJ c (1 3 ) a h 3
= . (4.16)
E 2 (1 2 ) 12
GJ s K(1 3 ) a h 3
= . (4.17)
E 2 (1 2 ) 12
{} T ={ui, vi, wi, xi, yi, zi, uj, vj, wj, xj, yj, zj} (4.19)
u = a1 + a2 X (4.20.a)
v = a3 + a4 X + a5 X2 + a6 X3 (4.20.b)
w = a7 + a8 X + a9 X2 + a10 X3 (4.20.c)
56
CHAPTER 4 ANALYSIS OF PILED RAFT IN LAYERED SOIL
w
y = = a8 + 2a9X + 3a10X2 (4.20.e)
x
v
z = = a4 + 2a5X + 3a6X2 (4.20.f)
x
{} = [A] {} (4.21)
The flexural strain (curvature) is obtained from Equations 4.20 (b) and (c) in
terms of the displacement, thus:
2 w/x 2
{} = 2
(4.22.b)
v/x
2
From Equations 4.21 and 4.22 (b) and (c), the strain {} can be formed in
terms of {} as follows.
{} = [C]{} (4.22.c)
{} = [B]{} (4.23)
where,
where, {f} is the stress resulted from the axial force. The moment M is
obtained from Equations 4.25(b) and (c)
{
M y = [D] 2 w/x 2 } (4.25.b)
{} = [D]{} (4.25.d)
By equating the external and the internal work done in the frame element ,
thus:
L
{} {F} = (1/2){}T[B]T[D][B]{}x
T
(4.26)
0
Where, {F} is the nodal force vector which is given by the following
equation.
58
CHAPTER 4 ANALYSIS OF PILED RAFT IN LAYERED SOIL
L
[K] = (1/2) [B]T [D][B]dx (4.28)
0
The element stiffness matrix of the 3D-space frame [K] is always square and
symmetric.
After deriving the local stiffness matrix of the frame element, the
transformation matrix of forces and displacements is needed to obtain the
global stiffness matrix of the mentioned element. The member stiffness in
the local system of coordinates x, y and z is transferred to the global system
of coordinates X/, Y/ and Z/ using the transformation matrix. The nodal
forces in local coordinates can be expressed in terms of those in the global
coordinates as follows.
where,
[F] and [F\] are column matrices representing the nodal forces in local and
global coordinates respectively and [T] is the transformation matrix and
given by following forms.
R 0 0 0
0 R 0 0
[T] = (4.30)
0 0 R 0
0 0 0 R
59
CHAPTER 4 ANALYSIS OF PILED RAFT IN LAYERED SOIL
and
l x mx nx
[R ] = ly my n y (4.31)
l z mz n z
where, lx, mx, nx are direction cosines of the angles between x-axis and each
X\, Y\ and Z\ axes respectively as shown in Fig. 4.4. Ly, my, ny are direction
cosines of the angles between y axis and each X\, Y\ and Z\ axes respectively.
Lz, mz, nz are direction cosines of the angles between z-axis and each X\, Y\
and Z\ axes respectively. The orientation of the member x-axis should be
defined to find the remaining elements of [T] as lx, mx and nx (i.e. both
members y and z-axes rotate about the member x-axis to take up any
orientation) as shown in Fig. 4.5. These axes should be defined to proceed
the orientation of one of them.
If, however the member x axis lies along the global Z\ axis then, the
condition that the member y axis must lies in the global X\ and Y\ plane is
automatically satisfied and a further condition. For members not laying
along the global Z\ axis, the evaluation of the direction cosines of members y
and z axes is less straightforward, the problem is best tackled using vector
cross products. Therefore, the cross product C of two vectors A and B is
defined as follows.
A = i Xa + j Ya + k Za (4.32.a)
B = i Xb + j Yb + k Zb (4.32.b)
and
i j k
C = A x B = Xa Xb Xc (4.32.c)
Ya Yb Yc
60
CHAPTER 4 ANALYSIS OF PILED RAFT IN LAYERED SOIL
T1 T4 T7
R = T2 T5 T8 (4.33)
T3 T6 T9
61
CHAPTER 4 ANALYSIS OF PILED RAFT IN LAYERED SOIL
wi = fij Pj (4.36)
The service area of each element is a rectangle or a square area. For the
nodes i in Fig. 4.6, the flexible factor fii is determined using Bousinesque
concept as follows
qB
S e (corner ) = (1 2 )[I 5 ] (4.37)
2Es
Se is the elastic settlement beneath the corner of the rectangular area (B*L), q
is the contact pressure, Es is the modulus of elasticity of the soil and is
Poisson ratio of the soil. The factor m is a ratio between the length and the
width of the service area.
I5 in equation (4.37) can be written in the following manner.
1 1 + m 2 + m 1 + m 2 + 1
I5 = ln + m ln (4.38)
1 + m 2 m 1 + m 2 1
62
CHAPTER 4 ANALYSIS OF PILED RAFT IN LAYERED SOIL
1 8(1 ) 2
f ij = (4.39)
16G (1 ) r
where, is the Radial distance between points i and j and G is the Shear
modulus of the soil.
The Mindlin's solutions which are employed in this work are derived for
homogeneous soil medium. In the fact, the practical problems are almost
nonhomogeneous soil. It leads to a need to promote this model to be able to
treat with the said kind of problems. Mindlins solution has been used to
obtain approximate solutions for a layer of finite thickness by employing the
Steinbrenners approximation (Steinbrenner, 1934) to allow for the effect of
the underlying rigid base in reducing the soil displacements (Poulos, 1989;
Poulos & Davis, 1980). For n layers under the i which are shown in Fig. 4.6,
the layer which contains i is denoted by "". The settlement in each layer k
due to a point load at j can be calculated from the following equations (4.40)
and (4.41).
fkj = f ( ztop
k
) f ( zBottom
k
) (4.40)
n
f + ( fij f ( zBottom ))
\
fij = kj (4.41)
k=1, where k
where, fij\ is the modified flexible factor for the case of layered soil.
63
CHAPTER 4 ANALYSIS OF PILED RAFT IN LAYERED SOIL
64
CHAPTER 4 ANALYSIS OF PILED RAFT IN LAYERED SOIL
Pj = Kij wi (4.44)
where, ki,j is the stiffness coefficient of the soil stiffness. Accordingly, the
relationship between contact forces under raft nodes and settlements is
expressed as follows:
where:
{P} 3*nr vector of applied loads and moments on the raft nodes.
[kr] 3nr*3nr piled plate stiffness matrix.
65
CHAPTER 4 ANALYSIS OF PILED RAFT IN LAYERED SOIL
where:
[X] 3*nr matrix xi, yi and zi are coordinates of node i.
{} 3 vector of translations wc and rotations tan y and tan x.
66
CHAPTER 4 ANALYSIS OF PILED RAFT IN LAYERED SOIL
Elements which are distributed along the shaft present the shear strength of
the soil. The bearing resistance is presented by an element at the base of the
pile as shown in Fig. 4.6. The flexible factors for shear resistance and
bearing are fijsh and fijb respectively which may be determined according to
equations (4.48) and (4.49) for bearing will be determined approximately
related to Eqs. (4.49) and (4.51). The vertical displacement w of the node i
caused by the vertical load P at j for any depth beneath the surface, may be
expressed as follows:
1 3 4 8 (1 ) 2 ( 3 4 ) ( Z i Z j )
2
G (i, j ) = + +
16 G (1 ) R 1 R2 R1
3
1 ( 3 4 )( Z i + Z j ) 2 2 Z i Z 6Z i Z j (Z i + Z j ) 2
+ j
+ (4.49)
16 G (1 ) R2
3
R2
5
For uniform distributed tractions t over the surface area of the shaft denoted
by S, the vertical displacement is obtained as follows:
w(i ) = t ( j )
S
G (i, j ) dS ( j ) (4.50)
The flexible factor fiish can be gotten subsequently from the above Eq. 4.50.
By the integration over the area of the pile base which has a radius ro, while
fiib can be gotten from the following form.
2 ro
w(i ) = q f ij r dr d (4.51)
0 0
The integrations for fiish and fiisb are illustrated in details in Annex1.
67
CHAPTER 4 ANALYSIS OF PILED RAFT IN LAYERED SOIL
n
wi = I i , j Q j (4.52)
j =1
Where:
Qj Contact force on node j, [kN]
68
CHAPTER 4 ANALYSIS OF PILED RAFT IN LAYERED SOIL
Let the matrix [Is] represents the flexibility coefficients Ii,j and the matrix
[ks] represents the inverse of it, which its coefficients are ki,j. then, contact
forces can be written as a function in terms ki,j of the inverted matrix as
follows:
n n n
Q
i =1
i = wo
i =1
k
j =1
i, j (4.55)
69
CHAPTER 4 ANALYSIS OF PILED RAFT IN LAYERED SOIL
Ph= ks wo (4.56)
In Eq. (4.56), the force on the pile head is the sum of all contact forces,
n
ph= Qi , while the composed coefficient is the sum of all stiffness matrix
i =1
n n
coefficients of the soil, ks = k i, j .
i =1 j =1
Equation (4.56) gives the linear relation between the applied load on the pile
head and the uniform settlement wo, which is analogous to Hooks low.
Therefore, the composed coefficient ks may be used to determine the soil
stiffness adjacent to the pile. In case of analysis of a single pile, it is easy to
determine the contact forces Qi. Substituting the value of wo from Eq. (4.56)
in Eq. (4.54) gives Eq. (4.57) in n unknown contact forces Qi as:
n
ph k i , j
j =1
Qi = (4.57)
ks
70
CHAPTER 4 ANALYSIS OF PILED RAFT IN LAYERED SOIL
wn
ph = (4.58)
1 wn
+
ks Ql
where:
wn Nonlinear settlement of the pile, [m]
Ql Limit pile load, [kN].
71
CHAPTER 4 ANALYSIS OF PILED RAFT IN LAYERED SOIL
In Fig. 4.8 and Eq. 4.58, the initial tangent modulus for single pile is easily
obtained from linear analysis of the pile, which is equal to the modulus of
soil stiffness ks. The limit pile load Ql is a geometrical parameter of the
hyperbolic relation. In some cases the value of Ql is different from the actual
ultimate pile load. For a single pile, the force on the pile head ph is known.
Therefore, Eq. 4.58 gives the nonlinear settlement of the pile wn.
4.3.7. Soil stiffness for freestanding raft (pile groups under vertical
loads)
Deriving equations for piled raft foundation requires taking into account the
interaction effect among the pile groups. For doing that, the simple
freestanding raft on pile groups shown in Fig. 4.8 as an example is
considered, which having np = 4 piles and total nodes of n=23. The relation
between pile settlement and contact force on pile groups shown in Fig. 4.8
can be expressed in matrix form as:
72
CHAPTER 4 ANALYSIS OF PILED RAFT IN LAYERED SOIL
I1,1 ... I1,6 I1,7 ... I1,12 I1,13 ... ... I1,17 I1,18 ... I1, 20 ... ... I1, 23
w1 ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... Q1
....
... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ....
.... I 6,1 ....
... I 6,6 I 6,7 ... I 6,12 I 6,13 ... ... I 6,17 I 6,18 ... I 6, 20 ... ... I 6, 23
w6 1 Q
w I 7,1 ... I 7,6 I 7,7 ... I 7, 23 6 1
7 .... Q7
... ... ... ... ....
....
.... ... ... ... ...
....
w I ... ... ... ... I12,13 I12, 23 Q12 2
12 2 = 12,1
w13 I13,1 ... ... ... ... I13, 23 Q13
....
... ... ... ... ....
.... ....
w17 3
... ... ... ... Q
w ... ... ... ... 17 3
Q18
18 I18,1 ... ... ... ... I18, 23 ...
....
.... ... ... ... ...
...
w ... ... ... ... Q23 4
23 4
I ... ... ... ... I 23, 23
23,1
(4.59)
The total flexibility matrix in Eq. (4.59) can be inverted to give the
relationship between contact forces and nodal settlements as follows:
k1,1 ... k1,6 k1,7 ... k1,12 I1,13 ... ... k1,17 k1,18 ... k1,20 ... ... k1, 23
Q1 ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... w1
....
... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ....
.... k 6,1 ... k 6,6 I 6,7 ... k 6,12 k 6,13 ... ... k 6,17 k 6,18 ... k 6, 20 ... ... k 6, 23 ....
Q6 1 w
Q k 7,1 ... k 7,6 k 7,7 ... k 7, 23 6 1
7 .... ... ... ... ...
w7
....
....
.... ... ... ... ...
....
Q k ... ... ... ... k12,13 k12, 23 w12 2
12 2 = 12,1
Q13 k13,1 ... ... ... ... k13, 23 w13
....
... ... ... ... ....
.... ....
Q17 3
... ... ... ... w
Q ... ... ... ... 17 3
w18
18 k18,1 ... ... ... ... k18, 23 ....
...
... ... ... ... ...
....
Q ... ... ... ... w23 4
23 4
k ... ... ... ... k 23, 23
23,1
73
CHAPTER 4 ANALYSIS OF PILED RAFT IN LAYERED SOIL
(4.60)
where, ki,j is the stiffness coefficient of the soil stiffness matrix,[kN/m]. Due
to the high rigidity of the pile in its length direction, the settlement in every
pile itself is considered as a uniform. This assumption can establish the
relationship between the uniform pile settlement and the force on the pile
head in the pile groups. It can be done by equating all settlements in each
pile by a uniform settlement. Carrying out the summation of rows and
columns corresponding to the pile i in Eq. (4.60) leads to:
6 6 6
6 12 6 17 6 23 6
Q i k i , j k i, j k i, j k i, j wo1
i =1 1 i =1 j =1 i =1 j = 7 i =1 j =13 i =1 j =18
12 i =1 1
12 12 6 12 12 12 17 12 23
Q i k i , j k i, j k i, j k i , j wo 2
i17=7 2 = i = 7 j =1
i =7 j =7 i = 7 j =13 i = 7 j =18
17
i =7 2 (4.61)
17 6 17 12 17 17 17 23
Qi i k i, j k k k i , j wo 3
=13 j =1 i =13 j = 7
i, j
i =13 j =13
i, j
i =13 j =18 i =13 3
23 23 6 23
i =13 3
23 12 23 17 23 23
Q
k i, j k k k i , j wo 4
i i
i, j i, j
i =18 4
=18 4
i =18 j =1 i =18 j = 7 i =18 j =13 i =18 j =18
Accordingly, Eq. (4.61) can be rewritten for the simple pile groups in Fig.
4.8 in composed coefficients as:
where:
woi Settlement in pile i, [m]
74
CHAPTER 4 ANALYSIS OF PILED RAFT IN LAYERED SOIL
n2 m2
Ki,j Composed coefficient, [kN/m]. In general Ki,j= k
n = n1 m = m1
n,m
Phi Force on the head of pile i, which is equal to the summation of all
n2
contact forces in that pile, [kN]. In general Phi= Qn
n = n1
i 1 i j 1 j
n1= 1+ nn(l ), n2 =
l =1
nn(l ) , m1= 1+
l =1
nn(l ), m2 =
l =1
nn(l ) ,
l =1
75
CHAPTER 4 ANALYSIS OF PILED RAFT IN LAYERED SOIL
w1 I 1,1 I 1, 2 I 1,3 ... I 1,8 I 1,9 I 1,10 I 1,11 ... I 1,16 ... I 1,33 Q
I 2 ,1 I 2,33 Q
1
w I 2, 2 I 2,3 ... I 2 ,8 I 2,9 I 2 ,10 I 2,11 ... I 2,16 ...
2 2
3 w I 3,1 I 3, 2 I 3, 3 ... I 3, 8 I 3, 9 I 3,10 I 3,11 ... I 3,16 ... I 3,33 Q3
... ... ... ... ... ... ... ... ... ... ... ...
... ...
w8 p1 I 8,1 I 8, 2 I 8, 3 ... I 8,8 I 8,9 I 8,10 I 8,11 ... I 8,16 ... I 8,33 Q8 p1
w I I 9, 2 I 9,3 ... I 9 ,8 I 9,9 I 9 ,10 I 9,11 ... I 9,16 ...
I 9,33 Q9
9 = 9,1
w10 I 10 ,1 I 10 , 2 I 10 ,3 ... I 10 ,8 I 10 ,9 I 10 ,10 I 10 ,11 ... I 10 ,16 ... I 10 ,33 Q10
w
11 I 11,1 I 11, 2 I 11,3 ... I 11,8 I 11,9 I 11,10 I 11,11 ... I 11,16 ... I 11,33 Q11
... ... ... ... ... ... ... ... ... ... ... ... ... ...
w
16 p 2 I 16 ,1 I 16 , 2 I 16 ,3 ... I 16 ,8 I 16 ,9 I 16 ,10 I 16 ,11 ... I 16 ,16 ... I 16 ,33 Q16 p 2
... ... ... ... ... ... ... ... ... ... ... ...
... ...
w33 I 33,1 I 33, 2 I 33,3 ... I 33,8 I 33,9 I 33,10 I 33,11 ... I 33,16 ... I 33,33 Q33
(4.63)
Q1 k1,1 k1, 2 k1,3 ... k1,8 k1,9 k1,10 k1,11 ... k1,16 ... k1,33 w
k 2 ,1 k 2,33 w
1
Q k 2, 2 k 2,3 ... k 2 ,8 k 2,9 k 2,10 k 2 ,11 ... k 2,16 ...
2 2
Q3 k 3,1 k 3, 2 k 3, 3 ... k 3, 8 k 3, 9 k 3,10 k 3,11 ... k 3,16 ... k 3,33 w3
... ... ... ... ... ... ... ... ... ... ... ...
... ...
Q8 p1 k 8,1 k 8, 2 k 8,3 ... k 8,8 k 8,9 k 8,10 k 8,11 ... k 8,16 ... k 8,33 w8 p1
Q k k 9, 2 k 9,3 ... k 9 ,8 k 9 ,9 k 9,10 k 9,11 ... k 9 ,16 ...
k 9 ,33 w9
9 = 9 ,1
Q10 k10 ,1 k10 , 2 k10 ,3 ... k10 ,8 k10 ,9 k10 ,10 k10 ,11 ... k10 ,16 ... k10 ,33 w10
Q11 k11,1 k11, 2 k11,3 ... k11,8 k11,9 k11,10 k11,11 ... k11,16 ... k11,33 w11
... ... ... ... ... ... ... ... ... ... ... ... ... ...
Q16 p 2 k16 ,1 k16 , 2 k16 ,3 ... k16 ,8 k16 ,9 k16 ,10 k16 ,11 ... k16 ,16 ... k16 ,33 w16 p 2
... ... ... ... ... ... ... ... ... ... ... ...
... ...
Q33 k 33,1 k 33, 2 k 33,3 ... k 33,8 k 33,9 k 33,10 k 33,11 ... k 33,16 ... k 33,33 w33
(4.64)
76
CHAPTER 4 ANALYSIS OF PILED RAFT IN LAYERED SOIL
8 16
k1,1 k1,2 k1, j k1,9 k1,10 k 1, j k1,33
j =3
8
j =11
k k2,33
2,1
k2,2 k
j =3
1, j k2,9 k2,10 k2,11
8 8 8 8 8 8 8 16 8
Q1 k ki,33
k k k k k w1
Q i =3
i ,1
i =3
i,2
i =3 i =3
i, j
i =3
i ,9
i =3
i ,10
i =3 i =11
i, j
i =3
8 2 w2
8
Qi k9,1 k8,33
8
woi
i =3 p1
k9,2 k
j =3
1, j k9,9 k9,10 k8,11
i =3 p1
Q9 8 w9
Q = k10,1 k10,2 k 1, j k10,9 k10,10 k9,11 k9,33 w
10
j =3 16 10
16 16
ki,33 j woi
16 8 16 16 16 16 16
Qi ki,1 j k10,2 k k k k
i =11 p 2
i, j i ,9 j i ,10 j i, j
i =11 p 2 j =11 i =11 i =3 j =11 j =11 i =11i =11 j =11
... ...
Q w
33 33
8 16
k33,1 k33,2 k 33, jj k33,9 k33,10 k 33, jj k33,33
j =3 j =11
(4.65)
Accordingly, the total stiffness matrix of the piled raft of size (npr * npr) is
reduced to the size (nr * nr). It means that the composed coefficients
technique makes the size of the soil stiffness matrix of the piled raft problem
equivalent to that of the raft problem alone without piles. Now, Eq. (4.65)
can be rewritten in composed coefficients as:
77
CHAPTER 4 ANALYSIS OF PILED RAFT IN LAYERED SOIL
(4.66)
Where Kpi,pj, Ki,pj and Kpi,j are composed coefficients of the piled raft,
[kN/m].
Based on Eq. (4.66), the relationship between settlements and contact forces
of the piled raft can be written in general compacted matrix form as:
78
CHAPTER 4 ANALYSIS OF PILED RAFT IN LAYERED SOIL
The linear system of equations of the piled rigid raft may be expressed as:
{N }= [X ][kb][X ]T {} (4.69)
where:
{} 3 vector of translation wc and rotations tan y and tan x
79
CHAPTER 4 ANALYSIS OF PILED RAFT IN LAYERED SOIL
where:
{p} 3*nr vector of applied loads and moments on the raft nodes.
[kr] 3 nr*3 nr plate stiffness matrix.
{} 3*nr deformation vector of the raft.
In the case of analyzing an elastic raft on pile groups, the elastic shortening
of the pile may be added to the pile settlement in Eq. (4.70). The elastic
shortening of the pile i is expressed as:
Phi l i
i = (4.72)
Epi Api
where:
i Elastic shortening of pile i, [m].
li Length of pile i, [m].
80
CHAPTER 4 ANALYSIS OF PILED RAFT IN LAYERED SOIL
Equation (4.72) is written for the entire piled raft in matrix form as:
where:
{wp} Elastic shortening vector.
[Cp] Elastic pile matrix, which is a diagonal matrix.
{Ph} Vector of forces on pile heads.
To take the effect of pile shortening into account, the elastic coefficient of
the pile i in the matrix [Cp] is added to the flexibility coefficient of that pile
in the matrix [Cb] in Eq. (4.72) as follows:
Inverting the total flexibility matrix [[Cb]+[Cp]] gives also the total stiffness
matrix of piled raft [kp] with the effect of pile stiffness due to its elastic
material.
where [kp] is nr*nr stiffness matrix of the piled raft with the effect of pile
elastic material, [kp]=[[Cb]+[Cp]]-1.
displacement and substituting Eq. (4.75) in Eq. (4.71) gives the following
linear system of equations of the piled elastic raft:
Once settlements on piles woi are determined in the above three cases of
piled rafts, the individual forces along the pile shaft and on the pile base can
be obtained from Eq. (4.64).
Phio
ksi = (4.77)
woio
where:
Phio Force on the pile head obtained from the linear analysis, [kN].
woio Pile settlement obtained from the linear analysis, [m].
82
CHAPTER 4 ANALYSIS OF PILED RAFT IN LAYERED SOIL
and its corresponding settlement. Only the pile stiffness is modified at each
cycle from the iteration process. Using the equivalent diagonal matrix,
equations of the piled raft are solved for each iteration cycle until the
compatibility between raft, piles and soil is achieved.
Figure 5 shows the iteration cycle and the flow chart of the iteration process.
The iteration process can be described in the following steps:
1 Carry out the linear analysis of the piled raft by solving system of
linear Eqns (4.69) or (4.76) whichever is applicable, to get the
settlements {w}.
2 Find the nodal contact forces {Q} due to the computed settlements
from Eq. (4.70) for rigid raft and from Eq. (4.75) for elastic raft.
3 From the computed settlements and contact forces, determine the
nodal stiffness at all nodes on the raft and on pile heads from:
Qi
kei = (4.78)
wi
1
kei = (4.79)
1 w
+ i
ksi Qu i
5 Convert the full soil stiffness matrix (matrix [kb] or matrix [kp]) to
equivalent diagonal stiffness matrix [ke]. This matrix can be generated
from nodal raft stiffness computed in step 3 and pile stiffness
83
CHAPTER 4 ANALYSIS OF PILED RAFT IN LAYERED SOIL
computed in step 4.
6 Replace the full matrix by diagonal matrix [ke]. Then, carry out the
nonlinear analysis of the piled raft to get the settlements {w}.
7 Compute the contact force under the raft and force on pile head by:
Qi = kei wi (4.80)
8 Compare the settlement from cycle i with that of cycle i-1 to find the
accuracy of the solution.
The steps 3 to 8 are repeated until the accuracy reaches a specified tolerance
, which means that a sufficient compatibility between settlements of piles,
raft and soil are achieved at the piles-raft-soil interface. However, in this
analysis the nonlinear response is applied on the raft and piles as indicated
by El Gendy (1999) to the piled raft system.
84
CHAPTER 4 ANALYSIS OF PILED RAFT IN LAYERED SOIL
determined according to Eqns. 4.81 and 4.82, while the other one fhii is
gotten by integrating Eq. 4.82 over the surface Sp of the pile exposed to the
passive pressure of the soil (d*lp), where lp is the length of the pile element
in Eq. 4.81.
w p (i ) = t p ( j )
Sp
G (i, j ) dS p ( j ) (4.81)
1 3 4 1 x2 x 2 ( 3 4 )
G (i , j ) = + + +
16 G (1 )
x2 + y2 x2 + y2 ( x2 + y2 ) (
3
)
x 2 + y 2
3
1 4 (1 )( 1 2 ) 4 (1 )( 1 2 ) x 2 (4.82)
+
16 G (1 )
x2 + y2 x2 + y2
3
( )
( Es (i ) + Es ( i +1) ) + ( Es ( j ) + Es ( j +1) )
Es (ij ) = (4.83)
4
85
CHAPTER 4 ANALYSIS OF PILED RAFT IN LAYERED SOIL
2 Es (1) + ( Es ( j ) + Es ( j +1) )
Es (1 j ) = (4.84)
4
where Es(ij)
is the averaged Youngs modulus of the soil at the depths Zi and
Zj; Es(i) and Es(j) are the Youngs modulus for soil layer numbers i and j.
86
CHAPTER 4 ANALYSIS OF PILED RAFT IN LAYERED SOIL
and
where:
{ph} 4*np vector of applied loads and moments on the piles nodes in x-
87
CHAPTER 4 ANALYSIS OF PILED RAFT IN LAYERED SOIL
and y-directions.
[kph] 4 np*4 np pile stiffness matrix.
{h} 4*np deformation vector of the pile in x- and y-directions.
To express the displacement due to the elasticity of piles Eq. (4.90) can be
written as follows.
where:
{wph} Elastic deformation vector.
[Cph] Elastic pile matrix.
{Phh} Vector of forces on pile heads.
To take the effect of pile deformation due to the elasticity into account, the
elastic coefficient of the pile i in the matrix [Cph] is added to the flexibility
coefficient of that pile in the matrix [Cbh] in Eq. (4.90) as follows:
Inverting the total flexibility matrix [[Cbh]+[Cph]] gives also the total
stiffness matrix of piled raft [kph] with the effect of pile stiffness due to its
elastic material.
where [kph] is 2np*2np stiffness matrix of the piled raft with the effect of pile
88
CHAPTER 4 ANALYSIS OF PILED RAFT IN LAYERED SOIL
Once displacements on piles whi are determined in the above three cases of
piles, the individual forces along the pile shaft can be obtained from Eq.
(4.93).
Phj = Khij whi (4.93)
89
CHAPTER 4 ANALYSIS OF PILED RAFT IN LAYERED SOIL
tss = Nc Cu (4.95)
where is the effective vertical stress and Kp is the passive earth pressure
coefficient, equal to (1+sin) / (1-sin), where is the friction angle of the
soil.
90
CHAPTER 4 ANALYSIS OF PILED RAFT IN LAYERED SOIL
that of the foundations by the structural and geotechnical engineers, thus the
stiffness of the superstructure is neglected. However, the structural stiffness
can have effects on the distribution of column loads and bending moments
transmitted from the structure to the foundation, therefore interaction
analysis which accounts for the structural stiffness has to be considered.
Piled rafts are composite structures which are comprised of three elements:
the piles, raft and the supporting soil. Loads applied to the raft are
transferred to the soil through the piles; therefore, it is necessary to take into
account the interaction among the three elements. In the next sections, the
interaction mechanism for piled raft against vertical and lateral loads is
presented.
91
CHAPTER 4 ANALYSIS OF PILED RAFT IN LAYERED SOIL
92
CHAPTER 4 ANALYSIS OF PILED RAFT IN LAYERED SOIL
stiff elements on the raft to represent the extremely stiff structural elements
such as shear walls or solid cores. For very flexible structures, the effect of
the stiffness of the superstructure on the behavior of the foundation is not
significant, therefore, it can be neglected in the analysis of the raft. However,
if the superstructure is stiff, it needs to be included into the analysis in some
way.
93
CHAPTER 4 ANALYSIS OF PILED RAFT IN LAYERED SOIL
where:
{Pv} 3*nst vector of applied vertical loads and moments on the 3D-space
structure nodes.
[kprv] npr*npr piled plate stiffness matrix.
{v} 3*npr deformation vector of the piled raft.
where:
[Xv] 3*npr matrix xi, yi and zi are coordinates of node i.
{v} 3*npr vector of translations wc and rotations tan y and tan x.
while, the relation between the piles group against horizontal loads and the
soil in the 3D-space structure can be expressed as follows:
[kph] {v} = {Ph}-{Qh} (4.98.a)
where:
{Ph} 3*nst vector of applied lateral loads and torsion on the 3D-space
structure nodes.
[kph] 2npp*2npp piles stiffness matrix.
{h} 3*npp deformation vector of the pile group.
94
CHAPTER 4 ANALYSIS OF PILED RAFT IN LAYERED SOIL
where:
[Xh] 3*npp matrix xi, yi and zi are coordinates of node i.
{h} 3*npp vector of translations x, y and the rotations tan z.
The linear system of equations of the whole superstructure and the piled raft
stiffness due to the elasticity of system can be expressed as:
Considering the compatibility between the soil settlement and the piled raft
displacement, the following equations of 3D- space structure with elastic
piled raft against vertical and horizontal loads is given as follows.
where:
{P} 6*nst vector of applied vertical and horizontal loads, moments and
the torsion on the 3D-space structure nodes.
{} 6*nst deformation vector of the whole system in 3D-space.
95
CHAPTER 4 ANALYSIS OF PILED RAFT IN LAYERED SOIL
diagonal beam
P1
side beam nodal point
P2
Z
Y
framework cell
K.a K.a
3 4
3 Ic 4
Is Id Is
Id
a
a
r.a
1 2 Z
1 2
(a) (b)
X
1 2 6 7
aM1/2 aM1/2
M1 M1 M1
aM1/2 aM1/2
(c) (d)
M2
3
kaM2/2
8 KaM2/2
KaM2/2 KaM2/2
M2 4 9
(e) (f)
H KaH/2 KaH/2
H aH/2
aH/2
5 10
H
aH/2 aH/2
KaH/2 KaH/2
H (g) (h)
y y 2 8
4 10
5 1
11
x x 7
z i z j i4 j
3
6 9 12
Y\
member
j x axis
yj - yi
i X\
xj - xi zj - zi
Z\
Y\ y
Z\
z
97
CHAPTER 4 ANALYSIS OF PILED RAFT IN LAYERED SOIL
y
z
n Raft-soil-raft
interaction
x
Layer 1 j i
Pile-soil-raft
l interaction
k
Layer 2
Pile-soil-pile
Layer n interaction
a Ground surface
a
1
1 qs1 Q1 1
1 2
zj-1 2 Q2 2
zj 3
2
i
j-1
4
j qsj lj
j (lj-1+lj)/2
j
Qj j
b 2ro 6
n
qb
Qn
a) Pile shaft elements b) Pile line elements
98
CHAPTER 4 ANALYSIS OF PILED RAFT IN LAYERED SOIL
P2 P1
p
Ground surface
1 7 13 18
2
8 14 19
3 20
9 Q9 15
pile 1 pile np=4
4 21
10 16
5
17 22
11
6
23
Q6 12
a) Free-standing raft on piles y
N
ex
Ground surface
Ph1 y Phnp
b) Equivalent statical system
pile 1 pile np
C.L.
Ground surface
wo1 wonp
2y wc
c) Soil settlement Pile displacement
99
CHAPTER 4 ANALYSIS OF PILED RAFT IN LAYERED SOIL
P2 Ground surface
P1
p
1 2 9 10 17 18 24 25 32 npr = 33
3 11 19 26
Q1 Q32
4
12 20 27
5 28
13 Q13 21
6 29
14 22
7
23 30
15
8
31
Q8 16
pile 4
pile 1
a) Piled raft with node numbering of piled raft
P2 Ground surface
P1
p
1 2 3 4 5 6 7 8 9 10 11 12 13 nr = 14
Q1 Q32
Ph1 Ph4
pile 1 pile 4
b) Equivalent statical system with node numbering of raft alone
Ground surface
w1 w32
wo1 wo4
c) Soil settlement
100
CHAPTER 4 ANALYSIS OF PILED RAFT IN LAYERED SOIL
Linear analysis
Nonlinear analysis
ks
Pile-soil-pile
Py1 interaction in x-
Px1 direction
1
Pyn+2
Pxn+2
2
Py3
lp Px3
3
Pyn+4 Pile-soil-pile
Pxn+4 interaction in y-
Pyn direction
Pxn
101
CHAPTER 5 COMPUTER PROGRAM AND VALIDITY OF PROCEDURE
CHAPTER 5
COMPUTER PROGRAM AND VALIDITY OF PROCEDURE
5.1. Introduction
In this chapter a designed algorithm is introduced to handle the proposed
analysis. A computer program is edited to study the piled raft behavior
against own weight, live loads and seismic loads with 3D-space
superstructure considering the nonlinearity and the non-homogeneity of the
supporting soil medium using the proposed method of analysis. The program
is called ASTNII (Analysis of Piled Raft with Space Structure Resting on
Nonlinear Soil using Composed coefficient technique) and ASTNIII
(Analysis of Piled Raft with Space Structure Resting on Nonlinear Soil using
against vertical and horizontal loads). This program includes mainly the
effect of the continuity of the soil through a continuum soil model and the
nonlinear relation between contact pressures and soil deformations.
Moreover, ASTNII&ASTNIII can deal with a mat of variable thickness, a
mat with openings, and a mat with inverted beams. Besides, it can deal with
many types of loads such as concentrated vertical and horizontal forces
and/or moments. In addition, partial uniform vertical loads within the frame
element length could be analyzed. ASTNII&ASTNIII can deal with slabs of
variable thickness and variable properties. The effect of the variety of piles
length, diameter and geometry arrangement has been considered. Also, they
can deal with soil bores considering different horizontal layers of the soil
with different geological characteristics. They can also analyze the 3D-space
structure with its boundary conditions taking into account a linear soil or a
nonlinear soil. To include the effect of seismic loads on construction
buildings ASTNIII has been developed to resist the said forces by the pile
102
CHAPTER 5 COMPUTER PROGRAM AND VALIDITY OF PROCEDURE
group. It has the ability to reflect the effect of earthquake forces on the both
superstructure and substructure elements. Composed coefficient technique
employed in ASTNII & ASTNIII by some way to save in memory during
solving equations and consequently to reduce the run time. In spite of taking
the time factor in the consideration, the accuracy is considered also to be in
accepted limits. However, the interface plays an important role in such
analysis codes. It can be considered the gate which controls input and output
data especially huge problems. To confirm the validity of problems analyzed
by ASTNII&ASTNIII, they have been supported by a strong interface using
data base tables. Moreover, a graphic show has been added to the code in
order to enable the user to check up input data as 2D slights and 3D
isometric. ASTNII&ASTNIII are written by Visual Basic which has the
ability to use the extend memory on the hard disk to be promoted easy and
subsequently introduces a more comfortable program. Fig. 5.1 illustrates the
relation between the master program and subprograms. Brief details and
clarifying of master and subprograms are evidently explained in the
following sections.
103
CHAPTER 5 COMPUTER PROGRAM AND VALIDITY OF PROCEDURE
104
CHAPTER 5 COMPUTER PROGRAM AND VALIDITY OF PROCEDURE
105
CHAPTER 5 COMPUTER PROGRAM AND VALIDITY OF PROCEDURE
Each element should be defined by its number, number of nodes of the two
ends and the kind of its material.
106
CHAPTER 5 COMPUTER PROGRAM AND VALIDITY OF PROCEDURE
by selecting directions from a check list has a six degrees as icons. The load
value can be inserted in a text box.
107
CHAPTER 5 COMPUTER PROGRAM AND VALIDITY OF PROCEDURE
108
CHAPTER 5 COMPUTER PROGRAM AND VALIDITY OF PROCEDURE
109
CHAPTER 5 COMPUTER PROGRAM AND VALIDITY OF PROCEDURE
110
CHAPTER 5 COMPUTER PROGRAM AND VALIDITY OF PROCEDURE
If the maximum percentage of error exceeds the allowable limit (5%), the
modules of subgrade reaction of all nodes are modified to be the slopes of
111
CHAPTER 5 COMPUTER PROGRAM AND VALIDITY OF PROCEDURE
112
CHAPTER 5 COMPUTER PROGRAM AND VALIDITY OF PROCEDURE
START
END
113
CHAPTER 5 COMPUTER PROGRAM AND VALIDITY OF PROCEDURE
START
114
CHAPTER 5 COMPUTER PROGRAM AND VALIDITY OF PROCEDURE
115
CHAPTER 5 COMPUTER PROGRAM AND VALIDITY OF PROCEDURE
RETURN
116
CHAPTER 5 COMPUTER PROGRAM AND VALIDITY OF PROCEDURE
l No. of Iterations
START
l=0
l= l+1
X
Y
117
CHAPTER 5 COMPUTER PROGRAM AND VALIDITY OF PROCEDURE
Y
X
IF NO
ACCURACY
IS ACCEPTED
YES
RETURN
118
CHAPTER 5 COMPUTER PROGRAM AND VALIDITY OF PROCEDURE
START
NEXT
NMEMB= Number of total
RETURN
119
CHAPTER 5 COMPUTER PROGRAM AND VALIDITY OF PROCEDURE
l=0
1
kei =
1 w
+ i
ksi Qli
X Y
120
CHAPTER 5 COMPUTER PROGRAM AND VALIDITY OF PROCEDURE
YES Y
Rigid raft
Compute contact
forces Qi=kei wi
l=l+1
Convergence NO
satisfied
{}=|{w}l-{w}l-1|
YES
END
121
CHAPTER 5 COMPUTER PROGRAM AND VALIDITY OF PROCEDURE
l No. of Iterations
START
l=0
l= l+1
X
Y
122
CHAPTER 5 COMPUTER PROGRAM AND VALIDITY OF PROCEDURE
Y
X
IF NO
ACCURACY
IS ACCEPTED
YES
RETURN
123
CHAPTER 5 COMPUTER PROGRAM AND VALIDITY OF PROCEDURE
l=0
nodes
1
kei =
1 w
+ i
ksi Qli
X Y
124
CHAPTER 5 COMPUTER PROGRAM AND VALIDITY OF PROCEDURE
Compute contact
forces Qi=kei wi
l=l+1
NO
Convergence
satisfied
YES
END
125
CHAPTER 5 COMPUTER PROGRAM AND VALIDITY OF PROCEDURE
126
CHAPTER 5 COMPUTER PROGRAM AND VALIDITY OF PROCEDURE
P2 = P1
Piles are located below P1 & P2
1m
P1 P2
P1
2
P1
P2 P1
2
P1
P2 P1
1
1m 2 2 2 2
Fig. 5.8-a. Load capacity of each pile = 875 [kN] Pile and load configuration
P2
Ep = ER = tR = 0.5 m
P1 P1
30000 MPa
P = P =0.2
l = 10 m
E = 20 MPa
= 0.3
H = 20 m
2 2 2 2
d = 0.5 m
127
CHAPTER 5 COMPUTER PROGRAM AND VALIDITY OF PROCEDURE
20
10 ASTNII
ELPLA
5 Van Imp(2001)
GARP
0
0 20 40 60 80 100 120 140
Central settlement [mm]
1200
Max Moment [kNm/m]
900
600
ASTNII
300
ELPLA
Van Impe (2001)
0
0 0.5 1 1.5 2 2.5
Raft thickness [m]
128
CHAPTER 5 COMPUTER PROGRAM AND VALIDITY OF PROCEDURE
8
7
6
Max. settlement [cm]
5
4
3
2 ASTNII
1 ELPLA
Van Impe (2001)
0
0 0.5 1 1.5 2 2.5
Raft thickness [m]
2
Diffrential settlement [cm
ASTNII
ELPLA
1.5 Van Impe (2001)
0.5
0
0 0.5 1 1.5 2 2.5
Raft thickness [m]
129
CHAPTER 5 COMPUTER PROGRAM AND VALIDITY OF PROCEDURE
3.5
l /d =25, K=50
3 poisson=0
2.5
2
z /l
1.5
ASTNII
1 Polus
0.5
0
0 0.2 0.4 0.6 0.8 1
p (pai) dl /P
130
CHAPTER 5 COMPUTER PROGRAM AND VALIDITY OF PROCEDURE
l /d
2 4 6 8 10 12 14 16 18 20
0
0.05
0.1
Ip=settlement *Es *l /P
0.15
0.2
0.25
0.3
0.35
0.4 Polus FEM
Polus Elastic approach
0.45
FEM by ASTNII
0.5
131
CHAPTER 5 COMPUTER PROGRAM AND VALIDITY OF PROCEDURE
20
15
Pile head load [MN]
10
ASTNII
Jarden et al (1986)
PGROUPN RF shaft=0, base=0
5 Poulos 1986 curve a
Poulos 1986 curve b
PGROUPN RF shaft=.9, base=.9
0
0 1 2 3 4 5 6 7 8 9 10
Head settlement s [mm]
Fig. 5.11 Comparison of load-settlement response for single pile (EP =
30,000GPa)
132
CHAPTER 5 COMPUTER PROGRAM AND VALIDITY OF PROCEDURE
where s is the soil Poissons ratio and rg, for rectangular pile group
configurations, may be taken as the radius of the circle of equivalent area to
that covered by the pile group. In the analyses of ASTNII, DEFPIG,
PGROUP and PGROUPN, no limiting value for axial interaction effects has
been adopted.
15
ASTNII
L/d=25 DEFPIG
)
GROUPPALO
0.5
Ep/G=1000
Normalized group stiffnessKp /s (Gn
s/d=2.5 PGROUP
PIGLET
10 PGROUPN
0
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
Square root of number of piles in group
Fig. 5.12 Comparison of different pile group analysis methods for s/d=2.5
133
CHAPTER 5 COMPUTER PROGRAM AND VALIDITY OF PROCEDURE
5.3.5. Test problem 5 Case study: Piled raft of Torhaus (El Gendy,2007)
Torhaus is the first building in Germany with a foundation designed as a
piled raft. The building lies in Frankfurt city in Germany. It is 130 [m] high
and rests on two separate piled rafts, where a street passes under the
building. Measured instruments were installed inside the foundation to
record piled raft settlement and stress. Many authors studied the foundation
of the Torhaus and applied their analysis methods on piled raft. Some of
them are Sommer et al. (1985), Sommer (1989) and Reul & Randolf (2003).
Figure 5.13 shows a layout of Torhaus with piled rafts. The building has no
underground floors. The foundation is two separate equal piled rafts with
rectangular shape areas, each of 17.5 [m] * 24.5 [m] sides. The distance
between the two rafts is 10 [m]. Rafts are founded at a depth 3.0 [m] under
the ground surface. The estimated total load on each raft is 200 [MN]. Raft
thickness is 2.5 [m]. A total of 42 bored piles with a length of l = 20 [m] and
diameter of D = 0.9 [m] are located under each raft. The pile spacing varies
from 3.5 D to 3.0 D. The subsoil at the location of the building consists of
gravel and sand up to 2.5 [m] below the ground surface, followed by layers
of Frankfurt clay extending to great depth. The groundwater level lies below
rafts.
The building is constructed between 1983 and 1986, the recorded maximum
settlement at the raft meddle in 1988 was about 12 [cm] according to
Sommer (1989). If Torhaus stands on a raft only, the expected settlement
would be about 26 [cm] based on geotechnical studies according to Sommer
et al. (1985). Therefore, to reduce the settlement, piled rafts were
considered. Using available data and results of Torhaus piled rafts, which
have been discussed in details in the previous references, the present piled
raft analysis is evaluated and verified for analyzing a piled raft.
134
CHAPTER 5 COMPUTER PROGRAM AND VALIDITY OF PROCEDURE
[
W s = 70 MN/ m
2
] (5.3)
where:
Es Modulus of compressibility for loading, [MN/m2].
Ws Modulus of compressibility for reloading, [MN/m2].
Eso Initial modulus of compressibility, Eso=7 [MN/m2].
z Depth measured from the clay surface, [m].
135
CHAPTER 5 COMPUTER PROGRAM AND VALIDITY OF PROCEDURE
136
CHAPTER 5 COMPUTER PROGRAM AND VALIDITY OF PROCEDURE
load geometry about both x- and y-axes to carry out the analysis for a quarter
of a piled raft. In the present analysis, seven cycles in few minutes are
required to obtain the nonlinear analysis of the piled rafts together. This is
related to that using composed coefficient technique reduced the size of soil
stiffness matrix from [1314*1314] to [390*390]. Accordingly, the total
number of equations was reduced to 1170, where npr =1314, nr = 390 and
number of unknown per node is 3 (3 nr = 1170).
Table 5.1 lists results of central settlement and bearing factor of piled raft
obtained by the present analysis, El Gendy (2007) and those obtained by
Reul & Randolf (2003) using three-dimensional finite element analysis.
Also, the table includes the measured results presented by Sommer (1989).
Figures 5.15 and 5.16 compare loads on piles 1 to 6 (Fig. 5.14) obtained by
the present analysis with those obtained by Reul & Randolf (2003) using
three-dimensional finite element analysis, El Gendy Analysis and with
measured pile loads presented by Sommer (1989).
This case study shows that the present analysis is not only an acceptable
method to analyze piled raft but also a practical one for analyzing large piled
raft problems. Besides the analysis gives good agreement with measured
137
CHAPTER 5 COMPUTER PROGRAM AND VALIDITY OF PROCEDURE
results, it takes less computational time and less effort for generating input
data compared with other complicated models using three dimensional finite
element analysis.
- For the four analysis types, Table 5.2 shows central settlement and
bearing factor of piled raft.
138
CHAPTER 5 COMPUTER PROGRAM AND VALIDITY OF PROCEDURE
Bearing 88 77 75 64 88 77
factor kpp
[%]
Applying different analysis types on piled raft of Torhous shows that the
nonlinear analysis of piled elastic raft is the acceptable analysis type, which
its results are agreement with measured values.
139
CHAPTER 5 COMPUTER PROGRAM AND VALIDITY OF PROCEDURE
(130m)
(100m)
(0.0m)
Sand with gravel
Frankfurt clay
6 1 2 24.5 [m]
5 4 3
G1.dwg
Fig. 5.14 Mesh of piled raft with piles
140
CHAPTER 5 COMPUTER PROGRAM AND VALIDITY OF PROCEDURE
8
Measurement
Present study
3D-FE analysis
El Gendy analysis
6
Pile load Ph [MN]
0
Pile 5 Pile 4 Pile 3
8
Measurement
Present study
3D-FE analysis
El Gendy analysis
6
Pile load Ph [MN]
0
Pile 6 Pile 1 Pile 2
141
CHAPTER 5 COMPUTER PROGRAM AND VALIDITY OF PROCEDURE
142
CHAPTER 5 COMPUTER PROGRAM AND VALIDITY OF PROCEDURE
V
M
H
.
1 2 3
d lp
s s
400
300
Equivelant-bent analysis
DEFPIG
Pile load V 1 [kN]
ASTNIII
200 GEPAN
PGROUPN
PIGLET
100
143
CHAPTER 5 COMPUTER PROGRAM AND VALIDITY OF PROCEDURE
400
Equivelant-bent analysis
DEFPIG
300 ASTNIII
GEPAN
Pile load V 2 [kN]
PGROUPN
PIGLET
200
100
400
300
Equivelant-
Pile load V 3 [kN]
bent analysis
DEFPIG
200
ASTNIII
GEPAN
100 PGROUPN
PIGLET
144
CHAPTER 5 COMPUTER PROGRAM AND VALIDITY OF PROCEDURE
400
300
Equivelant-bent analysis
DEFPIG
Pile load H1 [kN]
ASTNIII
200 GEPAN
PGROUPN
PIGLET
100
400
300
Equivelant-bent analysis
DEFPIG
Pile load H2 [kN]
ASTNIII
200 GEPAN
PGROUPN
PIGLET
100
145
CHAPTER 5 COMPUTER PROGRAM AND VALIDITY OF PROCEDURE
400
300
Equivelant-bent analysis
DEFPIG
Pile load H3 [kN]
ASTNIII
200 GEPAN
PGROUPN
PIGLET
100
-50
Equivelant -bent
analysis
-40 DEFPIG
Pile momentM1 [kN.m]
ASTNIII
-30 GEPAN
PGROUPN
-20 PIGLET
-10
146
CHAPTER 5 COMPUTER PROGRAM AND VALIDITY OF PROCEDURE
-50
Equivelant -bent
analysis
-40 DEFPIG
Pile momentM2 [kN.m]
ASTNIII
-30 GEPAN
PGROUPN
-20 PIGLET
-10
-50
Equivelant -bent
analysis
-40 DEFPIG
Pile moment M3 [kN.m]
ASTNIII
-30 GEPAN
PGROUPN
-20 PIGLET
-10
147
CHAPTER 5 COMPUTER PROGRAM AND VALIDITY OF PROCEDURE
30
Equivelant-bent
analysis
25 DEFPIG
ASTNIII
20
GEPAN
Settlement w [mm]
PGROUPN
15
PIGLET
10
30
Equivelant-bent
analysis
25 DEFPIG
ASTNIII
20
Lateral displacement u [mm]
GEPAN
PGROUPN
15
PIGLET
10
148
CHAPTER 5 COMPUTER PROGRAM AND VALIDITY OF PROCEDURE
10
Equivelant-bent analysis
8 DEFPIG
Pile rotation * 10 -3 [rad]
ASTNIII
6 GEPAN
PGROUPN
PIGLET
4
149
CHAPTER 5 COMPUTER PROGRAM AND VALIDITY OF PROCEDURE
2
1.8 Corner pile curve
1.6
MPILE
1.4
1.2 DEFPIG
H/Hav
1
PGROUPN(nonlin)
0.8
0.6 PGROUPN(lin)
0.4
Center pile curve ASTNIII
0.2
0
3 6 9 12 15
s/d
5.4. conclusion
In this chapter, a new two codes edited by the author have been introduced
using flow charts. ASTNII and ASTNIII which presented in this chapter have
the ability to analyze 3D buildings resting on piled raft considering a
continuum model for the soil. This model takes in the account the interaction
effect between different elements pile-soil-pile, raft-soil-raft and pile-soil-
raft. Moreover, loads in the lateral directions besides the vertical direction
are considered and non-homogenuity and non-linearity of the soil are
considered also. The valadity of those codes are examined in some test
problems which are presented in this chapter also. Curves which show the
comparison between results carried out by ASTNII and ASTNIII and those
results analyzed by other approved programs reflected a good agreement
between ASTNII and ASTNIII from a side and other approved programs from
another side.
150
REFERENCES
REFERENCES
References
240
REFERENCES
241
REFERENCES
242
REFERENCES
243
REFERENCES
244
REFERENCES
245
REFERENCES
246
REFERENCES
247
REFERENCES
248
REFERENCES
249
REFERENCES
250
REFERENCES
251
REFERENCES
252
APPENDIX I DISPLACEMENT AT THE CENTER DUE TO A CYLINDRICAL SURFACE STRESS
APPENDIX I
The derivation of the integrated form of the singular Mindlins solution over
cylinder surface is presented.
R2
ZB
A
ZA ro = x2 + y2
R1
ZB
B
R1 = ro2 + ( z A z B ) 2
ro
R2 = ro2 + ( z A + z B ) 2
P
y
x
Z
w( A) = G ( A, B) P( B) (I.1)
252
APPENDIX I DISPLACEMENT AT THE CENTER DUE TO A CYLINDRICAL SURFACE STRESS
3 4 s 8 (1 s ) 2 ( 3 4 s ) ( z A z B ) 2
+ +
1 R1 R2 R1 3
G ( A, B ) = 2
16 G s (1 s ) ( 3 4 s )( z A + z B ) 2 z A z B
2
6z A zB (z A + zB )
+ 3
+
R 25
R2
(I.2)
For distributed (constant ) traction t over a surface S, the following equation
can be obtained:
w( A) = t ( B ) ( A, B )dS ( B ) (I.3)
s
where the axial traction t are distributed over a cylindrical shaft area of the
pile-soil interface. The singular part G*(A,B) of Mindlins solution may be
expressed as:
3 4 s ( z A z B )
G * ( A, B ) = C + (I.4)
R1 R13
1
Where C is a constant equal to
16G s (1 s ) . In order to integrate Eq.
(I.4), it is convenient to choose the origin of coordinates at the nodal point A,
ie zA =0, and, for simplicity, zB=z. Therefore, R1 may be expressed as
253
APPENDIX I DISPLACEMENT AT THE CENTER DUE TO A CYLINDRICAL SURFACE STRESS
3 4
z2
G ( A, B ) = C
* s
+ 2
r +z
2 2 (ro + z )
2 3/ 2 (I.5)
o
3 4 2
z
s G ( A, B)dS ( B) = C s r 2 + z 2 + (ro2 + z 2 ) 3 / 2 dS ( B)
* s
(I.6)
o
where S is the cylindrical surface of the pile shaft element. The two integrals
in Eq. (I.6) may be evaluated in polar coordinates) separately. The first
integral yields:
2 h / 2
1 1
(3 4 s ) dS ( B ) = (3 4 s ) ro dzd
s r +z
o
2 2
0 h / 2 r +z
o
2 2
2
h/2 h + h 2 + 4ro2
= 2ro (3 4 s ) ln z + z + ro
2
= 2ro (3 4 s ) ln
h / 2 h + h 2 + 4r02
h + h2 + d 2
= d (3 4 s ) ln
h + h2 + d 2
where h is the height of the pile shaft element and d is the pile shaft
diameter. The second integral yields:
2 h / 2
z2 z2
s (ro2 + z 2 ) 3 / 2 dS ( B) = 0 h/ 2 (ro2 + z 2 ) 3 / 2 ro dzd
254
APPENDIX I DISPLACEMENT AT THE CENTER DUE TO A CYLINDRICAL SURFACE STRESS
h/2
z2
= 2 ro dz
h / 2 ( ro
2
+ z 2 3/ 2
)
This integral may be evaluated by substitution-let z = ro tan , then:
dz = ro sec 2 d
Thus:
ro2 tan 2
h/2 arctan h / 2 r
z2 o
h / 2 ( ro + z )
2 2 3/ 2
dz =
arctan h / 2 ro ( ro tan + ro )
2 2
r sec 2 d
2 3/ 2 o
arctan h / d o
=
arctan h / d o
tan sin d
h h
arctan h / d + sec arctan
= [ sin ]arctan h / d
h
+ sec d = 2 sin arctan + ln d d
arctan h / d
d h h
arctan h / d sec arctan
d d
255
APPENDIX I DISPLACEMENT AT THE CENTER DUE TO A CYLINDRICAL SURFACE STRESS
G ( A, B )dS ( B ) =
*
h h
sec arctan +
h+ h +d
2 2
h d d
= Cd (3 4 s ) ln 2 sin arctan + ln
+ + h h
sec arctan
h h 2
d 2 d
d d
256
APPENDIX II DISPLACEMENT AT THE CENTER DUE TO A CIRCULAR UNIFORM LOAD
APPENDIX II
2 ro
wi = q f ij rdrd (II.1)
0 0
Substituting Eq. (I.2) into Eq. (II.1) and carrying out the integration:
[
(3 4 ) r 2 + ( z c) 2 z c ]
][ ]
s o
[
+ 8(1 ) 2 (3 4 ) r 2 + ( z c) 2 ( z + c)
s s o
1
wi =
q + ( z c) 2
1
8G s (1 s ) z c ro + ( z c)
2 2
[
+ (3 4 )( z + c) 2 2cz 1 ] 1
z + c
s
ro2 + ( z + c) 2
+ 2cz ( z + c) 2 1
1
( z + c ) 3
(
ro + ( z + c)
2
)
2 3/ 2
(II.2)
257
APPENDIX II DISPLACEMENT AT THE CENTER DUE TO A CIRCULAR UNIFORM LOAD
Fig. II.1 Geometry of circular loaded area for finding displacements at center
2cz ( z + c) 2
8([
1 ) 2
(3 4 ) ]
( z + c )
( z + c)
s s
3
q
wi =
16G s (1 s ) (II.3)
(2 4 ) z c + (3 4 s )( z + c) 2cz
2
z+c
s
258
APPENDIX III ELASTIC EQUATIONS USED FOR LATERALLY-LOADED PILE ANALYSIS
APPENDIX III
The basic equation from which the required solution is derived that obtained
by Mindlin (1936), for the horizontal displacements caused by a horizontal
point load embedded in the interior of a semi-infinite elastic-isotropic
homogeneous mass. Referring to Fig. III.1, this solution is as follows:
(3 4 ) 1 x 2 (3 4 ) x 2 2cz 3 x 2
+ + 3
+ 3
+ 3 1 2 +
Q R R R R R2 R2
x =
1 2 1 2
(III.1)
16G (1 ) 4(1 )(1 2 ) x2
1
R2 + z + c R2 ( R2 + z + c )
Douglas and Davis (1964) integrated this equation over rectangular area, and
obtained the following solution.
At the upper corners A and B as shown in Fig. III.2, for a uniform horizontal
pressure p,
259
APPENDIX III ELASTIC EQUATIONS USED FOR LATERALLY-LOADED PILE ANALYSIS
c
Surface
r = (x2+y2)1/2
(Loading in x direction)
Fig. III.1. Definition of Point-load problem.
x =
pb
[(3 4 ) F1 + F4 + 4(1 2 ) (1 ) F5 ] (III.3)
32G (1 )
x =
pb
[(3 4 ) F1 + F2 + 4(1 2 ) (1 ) F3 ]
32G (1 )
where
2C1
K1 =
b
2C2
K2 =
b
( K1 K 2 ) 2
F1 = ( K1 + K 2 ) ln 2 ln
2 + 4 + ( K1 K 2 ) 2 ( K1 K 2 ) + 4 + ( K1 K 2 ) 2
260
APPENDIX III ELASTIC EQUATIONS USED FOR LATERALLY-LOADED PILE ANALYSIS
2( K1 + 1 + K12 )
F2 = 2 ln + ( K1 K 2 )
( K1 + K 2 ) + 4 + ( K1 + K 2 ) 2
4 + ( K1 + K 2 ) 2 4 + ( K + K )2 1 + K12
ln 2 + K1
2 1 2
( K1 + K 2 ) ( K1 + K 2 ) K1
( K1 + K 2 )
F3 = 2 K1 ln + ( K + K ) ln
K1
1+ 1+ K 2 1 2
2 + 4 + ( K + K ) 2
1 1 2
( K + K ) + 4 + ( K + K )2 ( K + K )
ln 1
2 1
2( K1 + 1 + K1 )
2
2
+ 1
4
2
[ 4 + (K + K )
1 2
2
]
( K1 + K 2 )
K1 ( 1 + K12 K1 )
2( K 2 + 1 + K 22 ) 2 + 4 + ( K + K )2
F4 = 2 ln + ( K1 K 2 ) ln 1 2
+
( K1 + K 2 ) + 4 + ( K1 + K 2 )
2
( K1 + K 2 )
4 + ( K + K )2 1 + K 22
K
2 1 2
( K1 + K 2 )
2
K2
K1 + K 2
F5 = 2 K 2 ln ( K + K ) ln
K2
+
1+ 1+ K 2 1 2
2 + 4 + ( K + K ) 2
2 1 2
( K + K ) + 4 + ( K + K )2 K + K
ln 1
2 1
2( K 2 + 1 + K 2 )
2
2
1
2
4
[ 4 + (K + K )
1 2
2
( K1 + K 2 ) ]
(
K 2 K 2 1 + K 22 )
For the displacement at other points in the same plane, the principle of
superposition may be employed.
261
APPENDIX III ELASTIC EQUATIONS USED FOR LATERALLY-LOADED PILE ANALYSIS
Surface z = 0
C2
A B
C1
D b C
262