For Finite Element Analysis

For f i ni t e el ement anal ysi s
1
Q: What are constitutive models and why are they
needed?*
A: They for m a central component of most, if not all,
the pr edictive models we develop of ground
response. From the simplest conceptual models, to
the most sophisticated mathematical model, we
need to idealise the behaviour of small elements
of soil.
A constitutive model can be defined as a set of
mathematical relationships between for example,
components of stress and components of strain.
2 *Carter (2006)
Soil compression
Odemeter Test
3
p
s
h
Shear ing of soil - Contractant
tr iaxial
4
3 1
3 1
q : stress Deviator
3
) 2 (
p' : stress Mean
o o
o o
=
+
=
o
3
o
1
Shear ing of soil - Dilatant
5
AV
6
Strain
Stress
Work hardening
Peak
Work softening
Yield point
o
t
t
o
Perfectly plastic
7
Strain
Stress
Strain
Stress
Strain
Stress
Strain
Stress
(a) Elastic
(d) Elastic-plastic softening
(b) Rigid plastic
(c) Elastic-Plastic
8
Elastic
Ther e is a one- t o- one
r elat ionship bet ween st r ess
and st r ain.
i.e. o = c E
I t can be linear or non linear
Af t er loading and unloading,
mat er ial r et ur n t o it s or iginal
condit ion.
Soil f low (def or mat ion)
mechanism depends on t he
st r ess incr ement .
Strain
Stress
9
Strain
Stress
Elastic-Plastic
The st r ess-st r ain r elat ionship
is not unique.
I t is non linear .
Af t er loading and unloading,
mat er ial doesnt r et ur n t o it s
or i ginal condit ion.
Soil f low (def or mat ion)
mechanism depends on t he
st r ess level.
Conditions of application of elastic model:
For problems where loads are at the working load
limit (i.e. much less than the ultimate load limit)
For initial displacement estimation of a structure
under loading.
Implications of adopting elastic model:
No-one can pretend that soil behaves as an elastic
mater ial except under str ictest conditions.
Choice of elastic modulus and Poissons ratio are
cr itical !!
10
Yield sur face is a boundar y in soil element
stress field.
It defines the state of stress at which soil
response changes from elastic to plastic.
11
allowed Not 0 ) f(
behaviour inelastic of Onset 0 ) f(
behaviour Elastic 0 ) f(
ij
ij
ij
>
=
<
o
o
o
Strain
Stress
Yield stress-
y
o
xy
o
yz
o
yz
o
xz
o
zz
o
xz
o
xy
12
x
z
y
o
xx
o
xy
o
xz
o
xx
o
yz
o
xz
o
zz
o
xy
o
zz
o
yz
o
zz
(
(
(
=
zz yz xz
yzx yy xy
xz xy xx
o o o
o o o
o o o
o
ij
: tensor Stress
o
h
13
q
h v
h
o o
o o
=
+
=
q : stress Deviator
(
p' : stress Mean
v
3
) 2
p
Yield Surface
Yield point
x
z
y
o
h
o
h
o
h
o
v
o
v
14
Plastic strain increment arise if:
1) The stress state is located on the yield surface, AND
2) The stress state remains on the yield surface after a
stress increment
Yield function f() tells us whether plastic strain is
occurring or not, however, we would like to know direction
and magnitude of plastic strain
For that we need flow rule
15
q
p
Yield Surface [f(o)]
Potential Surface [g(o)]
Critical State Line (CSL)
c
v
p
c
q
p
oc
q
p
oc
v
p
, )
o
o
c
c
c
=
g
p
16
Associated flow rule f() =g()
Non associated flow rule f() g()
It would be great advantage that f() =g() , only 1 function is to be
generated to describe plastic response
Advantages:
1) The solutions of the equations that emerge in the analyses is faster
2) The validity of the numerical predictions can be more easily
guaranteed
For metals f() =g() For soil f() g()
The assumption of normality of plastic strain vectors to the yield locus
would result in much greater plastic volumetric dilation than actually
observed.
17
c , st r ain or displacement )
o (st r ess) Real soil response
Idealised soil model MC model
18
c
o
First order approximation
Basic Law: c
i
= c
i
e
+ c
i
p
c
i
e
= r ever sible (elast ic) st r ain
c
i
p
= ir r ever sible (plast ic) st r ain
c
i
p
= 0 f or f < 0
f = f (o
xx
, o
yy
, o
zz
, o
xy
, o
xz
, o
yz
)
Yield f unct ion
c
i
p
c
i
e
c
i
19
o
t
o
o
o
o
z
o
x
t
xz
t
xz
t
o
o
o
c
s r
|
Failure criterion: t
f
= c + o
a
t an|
t
o
t
f
Or r c cos| + s sin|
s : cent r e of Mohr s st r ess cir cle
r : r adius of Mohr s st r ess cir cle
s sin|
c cos|
Hexagonal shape
20
21
strains plastic of direction the determines
strains plastic of magnitude the determines that multiplier a
etc , : means This
: strains plastic for rule Flow
p
y
p
x
p
i
=
c
c
=
c
c
=
c
c
=
c
c
=
i
y x
i
g
g g
g
o
c
o
c
o
c
22
g = 0
f = 0
o
n
o
n
oc
s
p
oc
v
p
The assumption of normality of plastic strain increment vectors to
the yield laws would result in much greater plastic volumetric
dilation than actually observed.
f = g
M-C model: f r s sin| c cos|
g r s sin - c cos, : dilatancy angle
23
c
t
Dense sand
Loose sand
c
v
Dense sand
Loose sand
c
v
= c
x
+ c
y
+ c
z
= 2c
x
+ c
y
24
Dense packing of grains
AV
Interlocking saw blades
t
o
|
|
i
Sliding takes place NOT on horizontal planes, but on planes inclined at an
angle to the horizontal.
| = + |
i
(strength = dilatancy + friction)
The apparent externally mobilised angle of friction on the horizontal
planes (|) is larger than the angle of friction resisting sliding of the inclined
planes (|
i
).
= | |
i
25
o
yy
o
xy
xy 1
c
yy
xy
o
xy
G
Only plastic strains
f = 0
f < 0
c
v
tan
) 1 ( 2
=
A
A
+
=
xy
yy
E
G
xy
c
yy
tan
A
xy
Ac
yy
26
o
y
o
x
= o
z
o
z
|o
y
- o
x
|
c
y
E = 2G(1+v)
c
y
c
v
c
v
= c
x
+ c
y
+ c
z
= 2c
x
+ c
y
o
sin 1
sin 2
tan
=
1
1-2v
o
27
28
c
o
E
0
E
50
o
y
2
y
o
Select E
0
when soil behaves
linear elast ic f or a lar ge
r ange.
Select E
50
f or gener al soils.
Select Eur f or unloading
pr oblems (t unnelling,
excavat ion)
However , E is r elat ed t o
conf ining pr essur e,
loading/ unloading/ r eloading
E
ur
29
For one-dimensional compression:
v
v
o
o
= =
1
0
v
h
K
Select v to match K
0
.
Under loading: v = 0.3 ~ 0.4
Under unloading: v = 0.15 ~ 0.25
For undr ained clay analysis, c is the
undrained shear str ength of the soil.
For dr ained sand analysis, c is nor mally zero.
t
f
= c + o tan|
However, in FE analysis, a small value
c > 0.2 kPa to avoid numer ical dif ficulties.
In advanced MC model, c can increase with
soil depth.
30
| = 0 can be used f or undr ained analysis.
A high | value is f or dense sand and can make
analysis t ake long t ime.
A not t oo high value is r ecommended f or init ial
pr eliminar y analysis.
31
o
c
t
u
Ef f ect ive st r ess = Tot al st r ess por e pr essur e
o = o u
o c = 0
Undrained Clay
Sand
t
|
Except for ver y over-consolidated clays, we
can use = 0 for clay layers.
The dilatancy of sand depends on the sand
density and its fr iction angle |.
For quar tz sand, | 30.
For sand with | < 30, = 0.
32

For Finite Element Analysis

Uploaded by

Copyright:

Available Formats

For Finite Element Analysis

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

For Finite Element Analysis

Uploaded by

Copyright:

Available Formats

For f i ni t e el ement anal ysi s

You might also like