Sie sind auf Seite 1von 54

1

1. Finite element analysis procedure


Part II: Finite Element Method
2
Governing equation
Problem Domain of Interest
Problem Domain
Prescribed Displacement
Prescribed Force
Divide problem domain into many finite elements
3
System Static Equilibrium (F vs o)
0 b
T
= + o V
Stress tensor
gravitational (body) force vector
Forces and stresses must equilibrium in the system
Governing equation
4
Compatibility (u vs c)
u
T
V = c
Strain vector Displacement vector
Displacement and strain must agree in the system
Positive for compression
Governing equation
5
Stress-Strain Relationships (o vs c)
c A = o A D
Stress Tensor
Strain Tensor
Constitutive model to characterize material behavior
Material Constitutive (Stiffness) Matrix
Governing equation
6
Internal work
The principal of virtual work
q
S
T
V
T
V
T
dS q u dV b u dV
q
} } }
o + o = o oc
External work
by body force
External work by
prescribed force
ext int
F F =
Check convergence,
If not converged, change(increase) displacement u
Governing equation
{ } | |{ }
n G n
d K R =
Gaussian Elimination method or Frontal solver method:
{ } | | { }
n G n
R K d
1
=
{ } | |{ }
n
d B = c
{ } | |{ } | || |{ }
n
d B D D = = c o
8
CACULATE
Displacement Increment Au
CONVERT to
Strain Increment Ac
INTEGRATE
Stress Increment Ao
Iteration Scheme
e.g. Newton-Raphson
START at
a Given Boundary Conditions
e.g. Prescribed force or displacement
CACULATE
External Force Increment AF
ext
CHECK
Convergence
AF
ext
AF
int
<tolerance
If NOT
If Yes
Next loading increment
Constitutive Model
Stress Integration Algorithm
e.g. forward Euler scheme
Compatibility
CACULATE
Internal Force Increment AF
int
Global/Nodal
(Force vs. Displacement)
Local/Gaussian
(Stress vs. Strain)
Computation framework
9
Calculation
Convergence Criteria
% 1
F
F F
ext
ext int
s

3%~5% is more practical


error default tolerance
Calculation until system equilibrium is reached
Convergence
10
1. Define Problem Type/Dimension
2. Define Geometry and Material
3. Set Boundary Condition
4. Generate Mesh
5. Generate Initial Condition
6. Simulate Construction Procedure
7. Calculate until Convergence is reached
8. Validate the Finite Element Model
Analysis procedure
2. Types of analysis
Effective stress analysis:
All of the calculation in the computer program are based on the
effective stress. Soil and water are treated as different material.
Input parameters: v |
' ' ' '
, , , E c
Total stress analysis:
All of the calculation in the computer program are based on the
total stress. Only one material, soil-water mixture, exists.
No excess porewater is generated in the program.
Input parameters:
u u u
E s
v |
, , 0 ,
=
Effective stress analysis: soil and water are treated as two-phase material.
The stress of water (porewater pressure) and soil (effective stress) are
computed separately
Total stress analysis: soil and water are treated as a single material.
The total stress of the single-phase material is computed.
CLAY
h
o' u
Effective stress analysis
h
o
Total stress analysis
sat

0 , = |
sat
s
'
|' ', c
Sand (drained behavior)--------Effective stress analysis
Clay (undrained behavior)-----Effective stress undrained analysis
-----Total stress undrained analysis
14
Linear Elastic Model:
Hooks Elasticity Law
Nonlinear Elastic Model:
Duncan Hyperbolic Model
Linear Elastic- Perfectly Plastic Model:
Mohr-Coulomb Model
Elasto-Plastic Model:
Hardening Soil Model
Hardening Soil Model w/ Small Strain
Cam-Clay Model
o
c
o linearly
increases w/ c
o
c
o nonlinearly
increases w/ c
o
c
o linearly increases
w/ c and reaches to
a peak strength
3. Constitutive model
15
Hardening
Peak
Softening
x
x
Residual
Compression
Dilatancy
Constant
Volume
Real Soil Stress-Strain-Volume Relationships
Comparison of constitutive model
16
Elastic Model Elasto-Plastic Model
Simple, require less input
parameters (E, u)
Complicated, many
parameters
Efficient computation Great computation effort
Less numerical problems
(convergence)
Many numerical problems
(integration, convergence)
Results are ok for working
stress conditions
(away from failure)
However, plastic model can
describe more realistically
plastic deformation of soil
(plastic strain, dilatancy,
stress history)
Close form solutions are
available
Comparison of constitutive model
17
Plastic Deformation:
Unloading (e.g., Excavation)
Dilatancy
Dense sand or OC clay under
drained condition
(e.g., Retaining Structure)
Stress History
Ac=f (o, Ao)
current stress state, stress increment
o
c
Elastic
Elaso-Plastic
c
e
c
p
c
v
c
Elastic
(No dilation, Volume change is
controlled only by Poissons
ratio)
Elaso-Plastic
(by dilation angle)
Comparison of constitutive model
18
Set Boundary Conditions
Prescribed Force: External force
Prescribed Displacement: u 0, v0
Horizontal fixity u=0, v0
Vertical fixity u0, v=0
Total fixity u=0, v=0
standard fixity
Prescribed force or displacement
Boundary condition
19
Generate Initial Condition
Initial water pressure
Initial effective stress
Phreatic level (for knowing GWT)
Pore pressure distribution
(for excess pore water pressure or seepage conditions)
K
o
procedure (for horizontal ground and soil layers)
Gravity loading (e.g., slopes)
Initial stress
20
(A) Linear elastic elastoplastic model
o
c
E
0
o
c
E
0
E
E
Linear elastic elastoplastic model Linear elastic perfectly plastic model
21
Linear elastic elastoplastic model
c
plasticity
o
Test results
Linear elastic
Yielding stress
Simulation of the real soil
E
Defined by c, |
Disturbed soils
3 1
o o
1
c
3
o
1
o
Undisturbed soils
Problems for the simulation of real soil
c
|
E is underestimated
may not be affected
For sand-like material:
502 . 0
58 . 65 N V
s
=
Wave velocity from relationship of V
s
and N
(Taipei silty clay)
Shear modulus from wave equation
2
s
V G =
(=density of soil)
) 1 ( 2
s
s
E
G
v +
=
Elastic theory
3 . 0 25 . 0 =
s
v
Sand
Youngs modulus of the soil
) 1 ( 2
s s
G E v | + =
| is a parameter which take into account of the high stiffness at
small strain, |=0.3-0.5
For clay-like material (total stress analysis or non-porous mode):
E
u
/s
u
can be assumed to be 450-500, or obtained from back
analysis of well-documented case histories
We can establish the relationship between E and N directly
E=2000N (kPa), for example

u
=0.495 (saturated clay)
M
p
'
q
p
'
ln
e
1
= '
p
a
e
cs
e
k
e
k

Unloading /reloading
Isotropic virgin consolidation line (IVCL
One dimensional consolidation line
Critical state line (CSL)
(a)
(b)
(B). Cam-clay and other high order models (MITE3, MITS1)
F
p M q
' =
Critical state line ( CSL )
Elastic wall
Isotropic virgin consolidation line
Yielding surface
q
e
p
'
E
'
E Y
X
X
. State boundary surface
Unloading / reloading
X
'
Y
3
o
1
o
F
a
b
c
d
e
f
. Yielding surface and stress paths
The equations for the state boundary:
The yielding equation
) / 1 (
2 2 2
2
/
k
.
|

\
|
+
=
p q M
M
p
p
e

.
|

\
|

=

e e
p
a
e
exp

.
|

\
|
+
=
2 2 2
2
0
/ p q M
M
p p

The Cam-clay model requires the following
parameters: , , , and .
v M
E k
|
|
'

'

=
sin 3
sin 6
M
|
|
'

'

=
sin 3
sin 6
M
303 . 2
c
C
=
303 . 2
s
C
= k
c
o
Failure stress
E
E
According to the definition of bulk modulus, we can derive the
effective bulk modulus (under the drained condition) as:
p
p d
de
'
'
= k
p e e
cs
'
=
ln
'
+
=
'
c
'
=
'
p e
e
de
p d
d
p
K
v
) 1 (
) 1 /(
k
'
+
=
+
'
=
c
'
=
'
p e
p
d
v
) 1 (
) 1 /(
Yield function-I
Yield function-II
Definitions of cap yield surface :
~
32
(C) Hardening soil model (HS model)effective stress model
Main features:
(1)Use the theory of plasticity rather than theory of elasticity
(2) For some cases, the relationship between strain and deviatoric
stress can be simulated by a hyperbola (Duncan-Chang model)
(3) Including the properties of dilatancy
Parameters:
| v , , , , , , ,
50
c E E E m
ur
ref
ur
ref
oed
ref
All of them can be determined in a reasonable way (see PlAXIS manual)
(D) HS small modelevolve from HS model
Typical field measuring result
Typical analysis result using the finite element
method
Wall deformation curve
35
Wang (1997):
Numerical program:
FLAC
Soil model:
Creep model
hyperbolic model
Excavation case:
TNEC
Yielding surfaces for clayey soils
Common material??
3
o
1
o
Area 1
Area 2
Area 3
Area 4
Initial yielding surface Y3
Y2
Y1
Variation of normalized secant modulus with strain
=209kPa

s
uc
= 63kPa
=430kPa

s
uc
= 139kPa
=310kPa

s
uc
= 102kPa
c
1
c
Ao
0
Ao
1
o
1
-
o
3
E
sec
=
-
E
sec
s
uc
0.001 0.01 0.1 1
10
100
1000
London clay
Jrdine et al . (1984)
5000
Strain
c
(%)
v
o'
v
o'
v
o'
c
1
Ao
1
Ao
0
For HS Small model, two additional parameters are required
0.7
7 . 0

7 . 0

0
G
G
0
: Initial or very small strain shear; which can be obtained
from small strain test, bender element test, cross hole field
test or seismic survey

0.7
: the strain level at which the secant shear modulus is
reduced to 70% of G
0
; This value can be determined from
the small strain test, or use the following correlation
] 2 sin ) 1 ( ) 2 cos 1 ( 2 [
9
1
0 1
0
7 . 0
| o |
'
+
'
+
'
+
'
~ K c
G
(PLAXIS manual suggest)
A good case
history, TNEC is
used in evaluation
of the performance
of the above models
GL.-2.8
GL.-4.9
GL.-8.6
GL.-11.8
GL.-15.2
GL.-17.3
GL.-19.7
1FL GL.+0.0
B1F GL.-3.5
B2F GL.-7.1
B3F GL.-10.3
B4F GL.-13.7
B5F GL.-17.1
H300x300x10x15 GL.-2.3
H400x400x13x21 GL.-16.5
unit: meter
GL.-35.0
CL =30
=33%-38%
LL=33-36
PI=13-16
Gravel
GP N>100
SM N=4-11
=31
CL
=32%-40%
LL=29-39
PI=9-23
=29
SM N=22-24 =31
CL N=9-11
=29
SM
N=14-37
=32
GL.-35.0
GL.-46.0
GL.+0.0
GL.-2.0
GL.-8.0
GL.-5.6
GL.-33.0
GL.-37.5
|'
e
|'
|'
e
|'
|'
|'
Sungshan I
Sungshan II
Sungshan III
Sungshan IV
Sungshan V
Sungshan VI
Stage 1
Stage 2
Stage 3
Stage 4
Stage 5
Stage 6
Stage 7
4. Evaluation of the commonly used soil models
Comparison
MCC model
D
e
p
t
h

(
m
)
50
45
40
35
30
25
20
15
10
5
0
Displacement (cm)
141210 8 6 4 2 0
Distance from the wall (m)
0 10 20 30 40 50 60 70 80 90 100
S
e
t
t
l
e
m
e
n
t

(
c
m
)
-8
-6
-4
-2
0
excavation stage
Field measurement
MCC model
3
4
6
7
5
2
3
4
6
7
5
2
1
real soil parameter
D
e
p
t
h

(
m
)
50
45
40
35
30
25
20
15
10
5
0
Displacement (cm)
141210 8 6 4 2 0
Distance from the wall (m)
0 10 20 30 40 50 60 70 80 90 100
S
e
t
t
l
e
m
e
n
t

(
c
m
)
-8
-6
-4
-2
0
excavation stage
Field measurement
MCC model
3
4
6
7
5
2
7
6
5
4
3
2
1
adjusted parameter,k/ = 0.25
real soil parameter
adjusted parameter,k/ = 0.25
HS model
D
e
p
t
h

(
m
)
50
45
40
35
30
25
20
15
10
5
0
Displacement (cm)
141210 8 6 4 2 0
Distance from the wall (m)
0 10 20 30 40 50 60 70 80 90 100
S
e
t
t
l
e
m
e
n
t

(
c
m
)
-8
-6
-4
-2
0
excavation stage
Field measurement
HS model
3
4
6
7
5
2
7
6
5
4
3
2
1
Comparison
Real soil parameter
q
p
Real soil yield surface
Modified Cam-Clay yield surface
Critical state line
K
0
line
A
B
C
D
Relationship of stress path in modified Cam-Clay yield surface
and real soil yield surface
E
Comparison
HS small
model
D
e
p
t
h

(
m
)
50
45
40
35
30
25
20
15
10
5
0
Displacement (cm)
141210 8 6 4 2 0
Distance from the wall (m)
0 10 20 30 40 50 60 70 80 90 100
S
e
t
t
l
e
m
e
n
t

(
c
m
)
-8
-6
-4
-2
0
excavation stage
Field measurement
HS small model
3
4
6
7
5
2
7
6
5
4
3
2
1
D
e
p
t
h

(
m
)
50
45
40
35
30
25
20
15
10
5
0
Displacement (cm)
141210 8 6 4 2 0
Distance from the wall (m)
0 10 20 30 40 50 60 70 80 90 100
S
e
t
t
l
e
m
e
n
t

(
c
m
)
-8
-6
-4
-2
0
excavation stage
Field measurement
HS small model
3
4
6
7
5
2
7
6
5
4
3
2
1
real soil parameter
adjusted parameter,
0.7
= 10
-5
Comparison
Mohr-Coulomb model (| = 0 )
D
e
p
t
h

(
m
)
50
45
40
35
30
25
20
15
10
5
0
Displacement (cm)
141210 8 6 4 2 0
Distance from the wall (m)
0 10 20 30 40 50 60 70 80 90 100
S
e
t
t
l
e
m
e
n
t

(
c
m
)
-8
-6
-4
-2
0
excavation stage
Field measurement
HS small model
3
4
6
7
5
2
7
6
5
4
3
2
1
back analysis (E/Su=500)
Comparison
Duncan-Chang model
Displacement (cm)
D
e
p
t
h

(
m
)
Distance from the wall (m)
S
e
t
t
l
e
m
e
n
t

(
c
m
)
Field observation
3
4
6
7
3
4
6
7
excavation stage
Duncan-Chang model
5
2
1
5
2
1
141210 8 6 4 2 0
50
45
40
35
30
25
20
15
10
5
0
0 10 20 30 40 50 60 70 80 90 100
-8
-6
-4
-2
0
back analysis
q
MCC yield surface
K
0
-line
K-line
Advanced effective stress model
Gyration
p
This will introduce a gyration rate related parameter
Wall displacement (cm) Distance from wall (m)
S
e
t
t
l
e
m
e
n
t
(
c
m
)
D
e
p
t
h
(
m
)
0 20 40 60 80 100
-10
-8
-6
-4
-2
0
12 10 8 6 4 2 0
50
40
30
20
10
0
Measurement
Analysis with small strain
Conventional elastoplastic analysis
.
Comparison
3KS model
Excav. surface
FIGURE 8.25 The finite element meshes used in the analysis of excavation
(a) bad mesh (b) good mesh
Excav. surface
(a)
(b)
Retaining wall
Retaining wall
5. Mesh generation for excavation
Boundary conditions
(a)
Retaining wall
(b)
Retaining wall
FIGURE8.26 Boundary conditions of the finite element mesh
(a) boundary outside the excavation zone is allocated with rollers
(b) boundary outside excavation zone is allocated with hinges
final depth
FIGURE 8.27 Distance of the boundary required for the analysis of wall deflection or ground settlement
ground settlement :
wall deflection :
e
H
D
e
H D 4
>
e
H D 3
>
6. Corner effect on deformation behavior
10
20
30
40
2 4 6 0 2 4 6 0 2 4 6 0
Lateral wall deformation (cm)
STAGE 4: STAGE 6: STAGE 7:
FEM 2D FEM 3D Measured values
0
D
e
p
t
h
(
m
)
8
FIGURE 8.29 Comparisons of the wall deflection fromplane strain analysis, three
dimensional analysis and field measurement respectively in a corner
of the Haihaw Financial Center excavation
}struts
B
B
L
L
FIGURE 6.30 Relationship between the plane strain ratio and the aspect ratio of an excavation
(a) PSR, the length-width ratio, and the distance from the corner
(b) symbol explanation
Distance to the corner (m)
0 5 10 15 20 25 30 35 40 45 50
0.2
0.4
0.6
0.8
1.0
1.2
1.4
1.6
1.8
2.0
2.2
2.4
2.6
0.2
0.3
0.5
0.6
0.7
0.8
PSR
=0.9
(a)
PSR=
Plane strain ratio B=
Width
d=
Distance to the corner L=
Length
Section to be
evaluated
Section to be
evaluated
(b)
0.4
d
d
PSR
=0.1

Das könnte Ihnen auch gefallen