Plaxis Introductory Course, Danang, Vietnam

PLAXIS INTRODUCTORY COURSE
10-12 JULY 2013

Danang, Vietnam

Plaxis Introductory Course, Danang, Vietnam
PLAXIS DANANG 2013
Day 1: Wednesday 10.7.2013

Time Module Subject Lecturer
9:00 9:30 Lecture 1 Introduction to Geotechnical Finite Element Analysis Dr Phung
9:30 10:15 Lecture 2 Mohr-Coulomb Soil Model Dr Cheang
10:15 10:30 Break
10:30 12:00 Exercise 1 Simple Foundation on Elastoplastic Soil Dr Cheang

12:00 1:30 Lunch
1:30 2:15 Lecture 3 Non-linear Computation Dr Phung

2:15 3:00 Lecture 4 Hardening Soil Model Dr Cheang
3:00 3:15 Break
3:15 4:45 Exercise 2 Simulation of Laboratory Tests Dr Cheang

4:45 5:00 Q/A Session
Day 2: Thursday 11.7.2013

9:00 9:30 Lecture 5 Geometry, Elements & Descretization Dr Cheang
9:30 10:15 Lecture 6 Structural Elements Dr Phung
10:15 10:30 Break
10:30 12:00 Exercise 3 Modelling of Anchored Excavation Dr Phung

12:00 1:30 Lunch
1:30 2:15 Lecture 7 Drained and Undrained Analysis Dr Cheang

2:15 3:00 Lecture 8 Modelling of Groundwater Dr Cheang
3:00 3:15 Break
3:15 4:45 Exercise 4 Dewatering of Excavation Dr Phung

Day 3: Friday 12.7.2013

9:00 9:30 Lecture 9 Initial Geo-static Stresses Dr Phung
9:30 10:15 Lecture 10 Safety of Factor Analysis through Phi'-C reduction Dr Cheang
10:15 10:30 Break
10:30 12:00 Exercise 5 Stability of A Slope Reinforced with Soil Nails Dr Cheang
12:00 1:30 Lunch
1:30 2:15 Lecture 11 Overview of Soil Models Dr Phung

2:15 3:00 Lecture 12 Consolidation Analysis Dr Cheang
3:00 3:15 Break
3:15 4:45 Exercise 6 Geotextile Reinforced Embankment & Consolidation Dr Phung

4 of 375
LECTURERS
Dr. William Cheang Wai Lum Dr. Phung Duc Long
Plaxis AsiaPac, Singapore VSSMGE, Hanoi, Vietnam
William obtained his PhD from the Dr. Phung got PhD degree at Chalmers
National University of Singapore. His University of Technology, Sweden. He
interest is in Computational has more than 30 years of international
Geotechnics. He has worked as a experience, including more than 20 years
Geotechnical Engineer in Malaysia, with Plaxis. His expertise areas are: deep
Singapore and Thailand. He is involved foundations, deep excavations, soil
with many seminars and workshops improvement, pile dynamics, tunnelling,
around Asia for the promotion of good and numerical analysis. He has worked
and effective usage of Plaxis Finite with projects in many countries, among
Element Codes. other, Sweden, Norway, Denmark, USA,
England, Russia, Germany, India, Hong
Kong, China and Vietnam, etc.
ORGANIZERS
Construction Informatics and Consultancy
Plaxis AsiaPac, Singapore
JSC (CIC)
ADD: 37 Le Dai Hanh, Hai Ba Trung
16 Jalan Kilang Timor,
Dist., Hanoi, Vietnam,
05-08 Redhill Forum, TEL#: (84.4) 39746798
Singapore FAX#: (84-4) 38216793
Contact: Mr. Luong Thanh Hung
Mobile: (84) 988 922 884
Email: Hunglt@cic.com.vn
Skype : Hunglt07
University of Transport and
Communications, Fecon
Contact: Mr. Le Quang Hanh
Mobile: (84) 948 171 135
Email: hanhquangle@fecon.com.vn

CONTENTS
LECTURES&EXERCISES PAGE
Lecture1 GeotechnicalFiniteElementAnalysis 5
Lecture2 MohrCoulombModel 21
Exercise Exercise1:SimpleFoundationwithMohrCoulombModel 43
Lecture3 NonlinearCalculations 74
Lecture4 HardeningSoil&HSsmallModel 88
Exercise Exercise2:SimulationofTriaxialandOedometerTests 124
Lecture5 Geometry&MeshSelection 169
Lecture6 StructuralElementsinPlaxis2D 186
Exercise Exercise3:AnchoredExcavation 199
Lecture7 UndrainedAnalysisinPlaxis 222
Lecture8 ModellingofGroundwaterinPlaxis 238
Exercise Exercise4:ExcavationandDewatering 264
Lecture9 InitialStresses 276
Lecture10 FactorofSafetyAnalysisviaPhi=C'Reduction 282
Exercise Exercise5:StabilityAnalysisofASlopeStabilisedbySoilNails 310
Lecture11 OverviewofSoilModels 323
Lecture12 ConsolidationAnalysis 341
Exercise Exercise6:GeotextileReinforcedEmbankmentwithConsolidation 357
CG1: GEOTECHNICAL FINITE ELEMENT ANALYSIS
Antonio Gens
Technical University of Catalunya, Barcelona
some off the

h slides
lid were originally
i i ll createdd by:
b
Cino Viggiani (Laboratoire 3S, Grenoble, France)
Outline
Introduction
Finite Elements displacement analysis
Elements for two-dimensional analysis
Displacement interpolation
Strains
Constitutive equation
Element stiffness matrix
Global stiffness matrix
Solution of the global stiffness equations
Elasticity as applied to soils
Fundamentals, and elastic parameters
Two-dimensional elastic analysis
5 of 375
design requirements in geotechnical engineering
Stability (local and general)
Admissible deformation and displacements

p
geotechnical analysis: basic solution requirements
Equilibrium (3 equations)
Unknowns: 15
Compatibility (6 equations)
((6 stresses,, 6 strains,, 3 displacements)
p )
Constitutive
C tit ti equation
ti (6 equations)
ti )
Potts & Zdravkovic

(1999)
6 of 375
geotechnical numerical analysis
methods for numerical analysis

Finite
Fi it difference
diff method
th d
Boundary element method (BEM)
Discrete element method (DEM)
Finite element method (FEM)
Others (meshless methods, particle methods)
while the FEM has been used in many fields of engineering practice for
over 30 years, it is only recently that it has begun to be widely used for
analyzing geotechnical problems.
when properly used, this method can produce realistic results which are
of value to practical soil engineering problems
geotechnical finite element analysis
Objectives of the numerical (finite element) analysis

Selection of design alternatives
Quantitative predictions
Backcalculations
Understanding!
Identification of critical mechanisms
Identification
d f off kkey parameters
7 of 375
Advantages of numerical (finite element) analysis

Simulation of complete construction history
Interaction with water can be considered rigorously
Complex geometries (2D-3D) can be modeled
Structural elements can be introduced
No failure mechanism needs to be postulated (it is an outcome of the
analysis)
(Nea l ) unavoidable
(Nearly) na oidable uncertainties
nce tainties
Ground profile
Initial conditions (initial stresses
stresses, pore water pressure
pressure))
Boundary conditions (mechanical, hydraulic)
pp op ate model
Appropriate ode for
o soil
so behaviour
be a ou
Model parameters
Some requirements for successful numerical modelling

q
Construction of an adequate p
conceptual model that includes the basic
features of the model. The model should be as simple as possible but
not simpler
Selection of an appropriate constitutive model
model. It depends on:
type of soil or rock
goal of the analysis
quality and quantity of available information
Payy attention to p
patterns of behaviour and mechanisms rather than
just to quantitative predictions
Perform sensitivity analyses. Check robustness of solution
Model calibration (using field results) should be a priority, especially of
quantitative predictions are sought
Check against alternative computations if available (even if simplified)
8 of 375
three final remarks
1. ggeotechnical engineering
g g is complex.
p y
It is not because youre
using the FEM that it becomes simpler
2 the quality of a tool is important

2. important, yet the quality of a result
also (mainly) depends on the users understanding of both
the problem and the tool
3. the design process involves considerably more than analysis
Borrowed from C. Viggiani, with thanks
introduction: the Finite Element Method
the FEM is a computational procedure that may be used to obtain an

approximate
i t solution
l ti to
t a boundary
b d value
l problem
bl
the governing mathematical equations are approximated by a series of

algebraic equations involving quantities that are evaluated at discrete points
within the region of interest. The FE equations are formulated and solved in
such a wayy as to minimize the error in the approximate
pp solution
this lecture presents only a basic outline of the method

the discussion is restricted to:
linear elasticity
two-dimensional plane strain
attention is focused on the "displacement based" FE approach
9 of 375
introduction: the Finite Element Method
the first stage in any FE analysis is to generate a FE mesh
Footing
width = B
Node
Gauss point
a mesh consists of elements connected together at nodes
examples: embankment
10 of 375
examples: multi-anchored diaphragm wall
introduction: the Finite Elements Method
the nodes are the points where values Footing

width = B
of the primary variables (displacements)
are calculated
Node
the values of nodal displacements are

interpolated within the elements to give
algebraic expressions for displacement
Gauss point
and strain throughout the complete mesh
a constitutive law is then used to relate strains to stresses and this

leads to the calculation of forces acting
g at the element nodes
the nodal forces are related to the nodal displacements by equations

which
hi h are set up and
d solved
l d to find
fi d values
l off the
h nodal
d l displacements
di l
11 of 375
displacement interpolation
two-dimensional analysis of continua is generally based on the use of

either triangular or quadrilateral elements
the most used elements are based on an iso-parametric approach
displacement interpolation
primary unknowns: values of the nodal displacements

displacement within the element: expressed in terms of the nodal values
how does it work? p

polynomial
y interpolation
p
shape functions
= nodal
d l displacements
di l t
12 of 375
illustration for thePlaxis
six-noded triangular element
Introductory Course, Danang, Vietnam
y quadratic interpolation
3
v x
6 5
1 2 12 coefficients, depending on the values of the 12 nodal

4 p
displacements
strains may be derived within the element using the standard definitions
B matrix
B-matrix
constitutive relation (elasticity)
elasticity: one-to-one relationship between stress and strain

in a FE context,
context stresses and strains are written in vector form
the stress-strain relationship is then expressed as: = D
material stiffness matrix

linear isotropic elasticity in plane strain

1 v v 0
D
E v 1 v 0
(1 2v)(1 v) 1 2v
0 0
2
in this case the coefficients of the matrix are constants,, which means
that (for linear kinematics) the resulting F.E. equations are linear
13 of 375
what happens with inelastic constitutive relations?
advantage with elasticity: the coefficients of the matrix are constants,

the resulting F.E.
F E equations are linear
the problem may be solved by applying all of the external loads

in a single calculation step
soils usually do not behave elastically
D

with D depending on the current and past stress history
it is necessaryy to apply
pp y the external load in separate
p increments
and to adopt a suitable non-linear solution scheme
element stiffness matrix
body forces and surface tractions applied to the element may be

generalized into a set of forces acting at the nodes (vector of nodal forces)
P1x
nodal forces may be related
to the nodal displacements by: 3 1y
P

P2 x
K e Ue Pe 6 5 P2 y
P e
P1x 1
Ke element stiffness matrix 2
4
P1y
P6 x
Ke
B T DBdv P
6y
D material stiffness matrix

recall B matrix relating nodal displacements to strains
14 of 375
Gauss points Plaxis Introductory Course, Danang, Vietnam
Ke
B T DBdv
to evaluate Ke, integration must be performed for each element
a numerical integration scheme must be employed (Gaussian integration)
essentially, the integral of a function is replaced by a weighted sum

of the function evaluated at a number of integration points
global stiffness matrix (1)
the stiffness matrix for the complete mesh is evaluated by combining the
individual
i di id l element
l t stiffness
tiff matrixes
ti (assembly)
( bl )
this produces a square matrix K of dimension equal to the number of degrees-

of-freedom
ff d i the
in h meshh
the global vector of nodal forces P is obtained in a similar way by assembling

the element nodal force vectors
the assembled stiffness matrix and force vector are related by:
KUe P
where vector Ue contains the displacements at all the nodes in the mesh
15 of 375
global stiffness matrix (2)
if D is symmetric (elasticity), then Ke and hence K will be symmetric
the global stiffness matrix generally contains many terms that are zero
if the node numbering scheme is efficient then all of the non
non-zero
zero
terms are clustered in a band along the leading diagonal
assembly
schemes for storage
solution
take into account its sym and

banded structure
number of dofs
solution of the global stiffness equations
once the global stiffness equations have been established

(and
( d the
h boundary
b d conditions
d added),
dd d) they
h mathematically
h ll
form a large system of symultaneous (algebraic) equations
KU e P
these have to be solved to give values for the nodal displacements
it is advantageous to adopt special techniques to reduce

computation time (e.g. bandwidth and frontal techniques)
detailed discussion of such techniques is beyond the scope of

this lecture
16 of 375
Computation of other variables
once the nodal displacements have been obtained from the inversion
off the
h matrix
i K
KU P
e
The complete displacement field can be obtained:
Strains and stresses are computed at the Gauss points:
= D
elasticity as applied to soils
simple example: uniaxial tensile stress applied to an elastic bar
a xx 1
b yy
xx xx E 0
E
xx 2b xx
yy
v 1 v 0.5
xx
2a
1
xx ( xx v yy v zz )
stress-strain equations E
in three dimensions
1
yy ( yy v xx v zz )
(linear isotropic elasticity) E
1
zz ( zz v xx v yy )
E
17 of 375
elasticity as applied to soils
yx
shear stress shear strain
xy
xy xy xy
xy
G
yx
four elastic parameters (Young's modulus E , shear modulus G ,

Poisson's ratio and bulk modulus K ) are commonly used to
specify elastic materials
an (isotropic) elastic material is fully specified when values of two of

these parameters are given
E E
set of relations, e.g.: G K
2(1 v) 3(1 2v)
Iff v 0.5 then
h K
two-dimensional elastic analysis
a number of geotechnical engineering

problems can be studied in plane strain
retaining walls, continuous footings,

slopes
xx
E
(1 v) xx v yy
(1 2v)(1 v)
yy
E
(1 v) yy v xx
(1 2v)(1 v)
xy G
G xy
18 of 375
two-dimensional elastic analysis
Matrix D
E E
In terms of K and G
3(1 2 ) 2(1 )
4 2 2
K G K G K G 0
xx
3 3 3
xx
e
D
yy
K 2 G K 4 G K 2 G 0 e
3 3 3 eyy
zz zz
K G K G K G 0 e
2 2 4
xy 3 3 3 xy
0 G
0 0
drained and undrained analysis
in soil mechanics it is conventional to use two limiting types of analysis:
if the load is applied so quickly and/or the permeability is so small

that there is no time for outflow of pore water, the loading is said
to be undrained
if the loading is sufficiently "slow" for the pore water to escape
f
from ((or tto enter
t iinto)
t ) an element
l t off soil,
il th
then th
the loading
l di is i said
id
to be drained
in undrained conditions, excess pore pressures are built up

0,
u 0
'
'
in drained conditions, no excess pore pressures are built up

u = 0, = '
19 of 375
elastic effective stress analysis
in practice,
practice elastic soil parameters obtained from laboratory or
field testing may be specified in terms of either drained or
undrained values. If undrained values are specified, then it
i necessary to
is t convertt these
th t drained
to d i d values
l i order
in d tot
carry out undrained calculations using effective stress analysis
drained and undrained shear moduli are identical
Gu G G
drained and undrained Young's moduli are related by
2 ( in the range 0.3 to 0.35)

E (1 v) Eu
3
20 of 375
MOHR-COULOMB MODEL
Elasticity, Plasticity & Yielding of Soils
William Cheang Wai Lum
PhD MSc PGDip BEng (Hons)
Acknowledgement: Some slides from Dr Ronald Brinkgreve
Plaxis Introductory Course, 10- 12 July, Danang, Viet Nam
Mohr-Coulomb
Mohr Coulomb model and soil stiffness
Objectives:
1. To indicate features of soil behaviour

2. To formulate Hookes law of isotropic linear elasticity
3. To formulate the Mohr-Coulomb criterion in a plasticity
framework
4. To identify the parameters in the LEPP Mohr-Coulomb model
5. To g
give suggestions
gg on the selection of p parameters
6. To indicate the possibilities and limitations of the MC model
Plaxis Introductory Course, Da Nang 10 - 12 July 2013
21 of 375
Typical results from soil lab tests

Triaxial test ((axial loading)
g)
F
1-3
strength
t th
P
stiffness
tiff
-1
1
v
1 dilatancy
-1
3 v 3

Oedometer test ((one-dimensional compression)
p )
P
Pre-consolidation
lid ti stress
t
1 reloading 1
primary
i lloading
di
1
1
unloading
22 of 375

Oedometer test ((constant load;; secondaryy compression)
p )
1 time
1 creep
Typical results for soil stiffness

Stiffness at different levels of strain
Modulus reduction curve after Benz (2007)
23 of 375
Features of soil behaviour
1. Elasticity (reversible deformation; limited) > stiffness

2. Plasticityy (irreversible
( deformation)) > stiffness,, strength
g
3. Failure (ultimate limit state or critical state) > strength
4. Presence and role of pore water
5. Undrained behaviour and consolidation
6. Stress dependency of stiffness
7
7. Strain dependency stiffness
8. Time dependent behaviour (creep, relaxation)
9. Compaction
p en dilatancy
y
10. Memory of pre-consolidation pressure
11. Anisotropy (directional strength and/or stiffness)
Concepts of soil modelling

yy
yx
yz xy
1. Relationship between stresses (stress rates)
zy
and strains ((strain rates)) xx
xz
zx
1. Elasticity (reversible deformations) d=f (d) zz
1. Example: Hookes law

2 Plasticity
2. Pl ti it (i
(irreversible
ibl d ) d=f
deformations)
f ti d f (d,,h)
(d h)
1. Perfect plasticity, strain hardening, strain softening
g, yield
2. Yielding, y function,, plastic
p potential,
p , hardening/softening
g g rule
3. Example: Mohr-Coulomb yielding
3. Time dependent behaviour (time dependent deformations)
1. Biots (coupled) consolidation d f (d,,t)
d=f (d )
2. Creep, stress relaxation
y, visco p
3. Visco elasticity, plasticity
y
24 of 375
Types of stress
stress-strain
strain behaviour
Linear-elastic Non-linear elastic Elastoplastic

Lin. elast. perfectly-plast. EP strain-hardening EP strain-softening

Hookes
Hooke s law
xx 1 0 0 0 xx
1 0 0 0
yy yy
zz E 1 0 0 0 zz

(1 )(1 2 ) 0 0 0 1 0 0
xy 2 xy
yz 0 0 0 0 1
2
0 yz
1
zx 0 0 0 0 0 2 zx
Inverse:
xx 1 0 0 0 xx
1 0 0 0
yy yy
zz 1 1 0 0 0 zz
E 0 0 0 2 2 0 0
xy xy
yz 0 0 0 0 2 2 0 yz

zx 0 0 0 0 0 2 2 zx
25 of 375
Hookes
Hooke s law
In principal stress / strain components:

1 1 1
E
2 (1 )(1 2 ) 1 2
3 1 3
In isotropic and deviatoric stress / strain components:

p K 0 v
q 0 3G s

p 1
3 1 2 3
1
q ( 1 2 ) 2 ( 2 3 ) 2 ( 3 1 ) 2
2
Model parameters in Hookes

Hooke s law:
Two parameters:
- d1
- Youngs modulus E
- Poissons ratio
d3
- 1
Meaning (axial compr.):
d1 E
E
d1 1
d - 1
3
d1 1
3
26 of 375
Alternative parameters in Hookes

Hooke s law:
dxy
Shear modulus:
d xy E dxy
G
d xy 21
Bulk modulus:
dpp
dp E
K dv
d v 31 2
Oedometer modulus:
- d1
d E 1
- d1
Eoed 1
d1 1 1 2
Stress definitions
1. In general
1 general, soil cannot sustain tension
tension, only compression
2. PLAXIS adopts the general mechanics definition of stress and
strain: Tension/extension is positive; Pressure/compression is
negative
yy yy
xx xx xx xx
yy yy
3. In general, soil deformation is based on stress changes in the

grain skeleton (effective stresses)
4. According to Terzaghis principle: = - pw
27 of 375
Hookes
Hooke s law for effective stress rates
The modeling
Th d li off non-linear
li soilil behaviour
b h i requires
i a relationship
l ti hi
between effective stress rates (d ) and strain rates (d)
d 'xx 1 ' ' ' 0 0 0 d xx

d ' ' 1 ' ' 0 0 0 d
yy yy
d 'zz E' ' ' 1 ' 0 0 0 d zz

d 'xy (1 ')(1 2 ') 0 2 ' d xy
1
0 0 0 0
d ' yz 0 0 0 0 2 '
1
0 d yz

d 'zx 0 0 0 0 0 1
2 '
d zx
Symbolic: d ' D d
e
d D e 1
d '
Plasticity
Basic principle of elasto-plasticity:
ij ije ijp (total strains)
d ij d ije d ijp (strain

(st a rates)
ates)
Elastic strain rates:
d ije D e ijkl d 'kl

1
28 of 375
Plasticity
Basic principle of elasto-plasticity:
ij ije ijp (total strains)
d ij d ije d ijp (strain rates)
Plastic strain rates:

g
d ijp d
'ij
d
d = scalar;
l magnitude
i d off plastic
l i strainsi
dg/d = vector; direction of plastic strains
g = plastic potential function
When do plastic strains occur?
Determination based on yield function f = f (,)
1. If f<0 Pure elastic behaviour

2. If f=0 and df<0 Unloading from a plastic state (= elastic behaviour)
3. If f=0 and df=0 Elastoplastic behaviour
29 of 375
When do plastic strains occur?

Yield function f is (a.o.) a function of the stress state
f=0 0 can
ca be represented
ep ese ed as a bo border
de in thee
stress space (yield contour) f=0
f<0
Within the yield contour: f<0 f>0
On the yield contour: f=0
Outside the yield contour: f>0 (impossible stress state)
Condition: Yield contour must be convex
The Mohr-Coulomb failure criterion

Origin: F n
Coulomb: T A + F tan c - n tan
A c
F n
30 of 375
The Mohr-Coulomb failure criterion

n
In general:

The condition c - n tan must hold for arbitrary angle
The Mohr
Mohr-Coulomb
Coulomb failure criterion

c cos MC criterion:
-s* sin t* c cos - s* sin

t*
c
3
- 1
- n
-
-s*
31 of 375
The Mohr
Mohr-Coulomb
Coulomb failure criterion
MC criterion: t*c cos - s* sin
( 3 - 1)
t* = (
s* = (3+1)
1
2 '3 '1 c' cos ' 12 '3 '1 sin '
2c' cos ' 1 sin '

'1 '3
i ' 1 sin
1 sin i '
Note: Compression is negative and 1 2 3
Visualisation of the M-C failure criterion

c
n

-1
2c' cos '
a
b 1 sin '
1
1 sin '
a b
1 sin '
3
-
32 of 375
Full Mohr-Coulomb criterion
1 1
2 '3 '2 c' cos ' 12 '3 '2 sin '
2 '2 '3 c ' cos ' 2 '2 '3 sin '
1 1
2 '3 '1 c ' cos ' 2 '3 '1 sin '

1 1
2 '1 '3 c ' cos ' 2 '1 '3 sin '

1 1
2 '2 '1 c ' cos ' 2 '2 '1 sin '

1 1
3
2 '1 '2 c ' cos ' 2 '1 '2 sin '
2 1 1
Reformulation into yield functions
1
2 '3 '1 c' cos ' 12 '3 '1 sin '
f 2b 12 '3 '1 12 '3 '1 sin 'c' cos '
33 of 375
Reformulation into yield functions
1 f1a 12 '3 '2 12 '3 '2 sin 'c' cos '

f1b 12 '2 '3 12 '2 '3 sin 'c' cos '
f 2 a 12 '1 '3 12 '1 '3 sin
i 'c' cos '
f 2b 12 '3 '1 12 '3 '1 sin 'c' cos '
f 3a 12 '2 '1 12 '2 '1 sin 'c' cos '
3
2 f 3b 12 '1 '2 12 '1 '2 sin 'c' cos '
Parameters: Effective cohesion (c) and effective friction angle ()
Plastic potentials of the M-C model

g1a 12 '3 '2 12 '3 '2 sin c' cos
g1b 12 '2 '3 12 '2 '3 sin c' cos
g 2 a 12 '1 '3 12 '1 '3 sin c' cos
g 2b 12 '3 '1 12 '3 '1 sin
i c' cos
g 3a 12 '2 '1 12 '2 '1 sin c' cos
g 3b 12 '1 '2 12 '1 '2 sin c' cos
Dilatancy angle instead of friction angle
Motivation based on simple shear test

34 of 375
Failure in a simple

shear test: xx
t*
yy
g
d ije D e ijkl d 'kl 0
1
d ijp d d xx 0
'ij
g ' ' yy 1
d xxp d d xx 2 sin 0
' xx 4 t*
g ' ' xx 1
d yyp d d yy 2 sin d sin
' yy 4 t*
g '
d xyp d d xy d cos
' xy t*
d yy d yyp
t
tan
d xy d xyp
Failure in a simple shear test:
d yy d yyp
tan xy
d xy d xyp
xy
yy

dilatancy xy
35 of 375
The LEPP Mohr-Coulomb model
Linear-elastic perfectly-plastic stress-strain relationship
- Elasticity: Hookes law

- Plasticity:
y Mohr-Coulomb failure criterion
The LEPP model with Mohr-Coulomb failure contour is in PLAXIS

called the Mohr-Coulomb model
For this model: Plasticityy = Failure
This does NOT apply to all models!!!
The LEPP Mohr-Coulomb

Mohr Coulomb model
Model parameters:
- Youngs modulus (stiffness) E

- Poissons ratio
- Cohesion c
- Friction angle
- Dilatancy angle
Model
M d l parameters must beb determined
d i d such h that
h
real soil behaviour is approximated in the best possible way
36 of 375
Parameter determination
Parameter determination from:
1. Laboratory tests (triaxial test (CD, CU), oedometer test or CRS,

simple shear test, )
2. Field tests (SPT, CPT, pressure meter (Menard, CPM, SBP),
dilatometer ))
dilatometer,
3. Correlations with qc , PI , RD and other index parameters
4. Rules-of-thumb,, norms,, charts,, tables
5. Engineering judgement
MC approximation of a CD triax. test

1-3 E 50 3 = confining pressure
2c 'cos ' 2 '3 sin '

1 sin '
-1
v
2 sin
1 sin
-1
1-2
1 2
37 of 375
MC approximation of a compr. test

-1
Eoed
-1
(1 )(1 2 )
Eoed E
(1 )
Stiffness parameter suggestions
Order of magnitude for E50:
E50 3'
- Sand: 150..500 ref
p ref p
Loose Dense
15000 cu 5000 cu
- Clay:
u
E50 or G50
I p [%] I p [%]
Ip = plasticity index
38 of 375

-1 Order of magnitude Eoed (sand):
Eoedd 1'
150..500 ref
p ref p
pref Loose Dense

Eoed
Eoed 1..3 qc (correlation)
-1
d 1 1 E (1 )(1 2 )
Eoedd E Eoedd
d1 1 1 2 (1 )
Thiss Evalue
a ue applies
app es to
o primary
p a y compression
co p ess o

-1 Order of magnitude Eoed (clay):
500
Eoed '1 (correlation)
Ip
pref Eoed
Eoed 3..5 qc (correlation)
-1
d 1 1 E (1 )(1 2 )
Eoedd E Eoedd
d1 1 1 2 (1 )
Thiss Evalue
a ue applies
app es to
o primary
p a y compression
co p ess o
39 of 375

-1 Secant oedometer stiffness:
1
Eoed
1
1(1)
Eoed
1 (0)
-1
G
cu
40 of 375

Duncan & Buchignani
Possibilities and limitations of the LEPP

Mohr-Coulomb
Mohr Coulomb model
Possibilities and advantages

g
1. Simple and clear model

2. First order approach of soil behaviour in
general
3. Suitable for many practical applications 1
4 Limited number and clear parameters
4.
5. Good representation of failure behaviour
(drained)
6 Dilatancy
6. Dilatanc can be incl
included
ded
3
2
41 of 375
Possibilities and limitations of the LEPP

M h C l b model
Mohr-Coulomb d l
1
Limitations and disadvantages
1. Isotropic and homogeneous behaviour

2. Until failure linear elastic behaviour
3
3. No stress/stress-path/strain-dependent
stress/stress path/strain dependent stiffness
4. No distinction between primary loading and 3
unloading or reloading 2
5 Dilatancy
5. Dil t continues
ti for
f ever (no
( critical
iti l state)
t t )
6. Be careful with undrained behaviour
7. No time-dependency (creep)
42 of 375
Elastoplastic analysis of a footing
ELASTOPLASTIC ANALYSIS OF A
FOOTING
Computational Geotechnics 1
43 of 375
INTRODUCTION
One of the simplest forms of a foundation is the shallow foundation. In this exercise we will
model such a shallow foundation with a width of 2 meters and a length that is sufficiently long
in order to assume the model to be a plane strain model. The foundation is put on top of a 4m
thick clay layer. The clay layer has a saturated weight of 18 kN/m3 and an angle of internal
friction of 20.
Figure 1: Geometry of the shallow foundation.
The foundation carries a small building that is being modelled with a vertical point force.
Additionally a horizontal point force is introduced in order to simulate any horizontal loads
acting on the building, for instance wind loads. Taking into account that in future additional
floors may be added to the building the maximum vertical load (failure load) is assessed. For
the determination of the failure load of a strip footing analytical solutions are available from for
instance Vesic, Brinch Hansen and Meyerhof:
Qf
B
= c Nc + 12 0 B N
0
Nq = e tan tan2 (45 + 12 0 )
0
q 1) cot
Nc = (N
0
2(Nq + 1) tan
(V esic)
N = 1.5(Nq 1) tan 0 (Brinch Hansen)

(Nq 1) tan(1.4 0 ) (M eyerhof )

This leads to a failure load of 117 kN/ m2 (Vesic), 98 kN/m2 (Brinch Hansen) or 97 kN/m2
(Meyerhof) respectively.
2 Computational Geotechnics
44 of 375
SCHEME OF OPERATIONS
This exercise illustrates the basic idea of a finite element deformation analysis. In order to
keep the problem as simple as possible, only elastic perfectly-plastic behaviour is considered.
Besides the procedure to generate the finite element mesh, attention is paid to the input of
boundary conditions, material properties, the actual calculation and inspection of some output
results.
Aims
Geometry input
Initial stresses and parameters
Calculation of vertical load representing the building weight
Calculation of vertical and horizontal load representing building weight and wind force
Calculation of vertical failure load.
A) Geometry input
General settings
Input of geometry lines
Input of boundary conditions
Input of material properties
Mesh generation
B) Calculations
Initial pore pressures and stresses
Construct footing
Apply vertical force
Apply horizontal force
Increase vertical force until failure occurs
C) Inspect output
45 of 375
GEOMETRY INPUT
Start PLAXIS by double-clicking the icon of the PLAXIS Input program. The Quick select
dialog box will appear in which you can select to start an new project or open an existing
one. Choose Start a new project (see Figure 2). Now the Project properties window appears,
consisting of the two tabsheets Project and Model (see Figure 3 and Figure 4).
Figure 2: Quick select dialog
Project properties
The first step in every analysis is to set the basic parameters of the finite element model.
This is done in the Project properties window. These settings include the description of the
problem, the type of analysis, the basic type of elements, the basic units and the size of the
drawing area.
The Project tabsheet
Figure 3: Project tabsheet of the Project Properties window
In order to enter the proper settings for the footing project, follow these steps:
46 of 375
In the Project tabsheet, enter Exercise 1 in the Title box and type Elasto-plastic
analysis of drained footing or any other text in the Comments box.
In the General options box the type of the analysis (Model) and the basic element type
(Elements) are specified. As this exercise concerns a strip footing, choose Plane strain
from the Model combo box. Select 15-node from the Elements combo box.
The Acceleration box indicates a fixed gravity angle of -90, which is in the vertical
direction (downward). Independent acceleration components may be entered for pseudo-
dynamic analyses. Leave these values zero and click on the Next button below the
tabsheets or click on the Model tabsheet.
The Model tabsheet
Figure 4: Model tabsheet of the Project properties window
In the Model tabsheet, keep the default units in the Units box (Length = m; Force = kN;
Time = day).
In the Geometry dimensions box the size of the considered geometry must be entered.
The values entered here determine the size of the draw area in the Input window.
PLAXIS will automatically add a small margin so that the geometry will fit well within
the draw area. Enter Xmin =0.00, Xmax =14.00, Ymin =0.00 and Ymax =4.25.
The Grid box contains values to set the grid spacing. The grid provides a matrix of dots
on the screen that can be used as reference points. It may also be used for snapping to
regularly spaced points during the creation of the geometry. The distance of the dots is
determined by the Spacing value. The spacing of snapping points can further be divided
into smaller intervals by the Number of snap intervals value. Enter 1.0 for the spacing
and 4 for the intervals.
Click on the Ok button to confirm the settings. Now the draw area appears in which the
geometry model can be drawn.
47 of 375
Hint: In the case of a mistake or for any other reason that the project properties
should be changed, you can access the Project properties window by
selecting the Project properties option from the File menu.
Creating the geometry

Once setting the project properties have been completed, the draw area appears with an
indication of the origin and direction of the system of axes.
The cursor is automatically switched in the Geometry line drawing mode. If not, the user can
change the drawing mode to Geometry line by clicking the geometry line button .
In order to construct the contour of the proposed geometry as shown in Figure 5, follow these
steps. (Use Figure 5 for orientation, it represents the completed geometry).
Figure 5: Geometry model
Create sub-soil
Position the cursor (now appearing as a pen) at the origin (point 0) of the axes (0.0; 0.0).
Click the left mouse button once to start the geometry contour.
Move along the x-axis to (14.0; 0.0). Click the left mouse button to generate the second
point (number 1). At the same time the first geometry line is created from point 0 to point
1.
Move upward to point 2 (14.0; 4.0) and click again.
Move to the left to point 3 (0.0; 4.0) and click again.
Finally, move back to the origin (0.0; 0.0) and click the left mouse button again. Since
the latter point already exists, no new point is created, but only an additional geometry
line is created from point 3 to point 0. PLAXIS will also automatically detect a cluster
(area that is fully enclosed by geometry lines) and will give it a light colour.
48 of 375
Click the right mouse button to stop drawing.

This action created the sub-soil cluster. The next step is to introduce the footing.
Create footing
Position the cursor at point 4, (6.0, 4.0) and click the left mouse button once.
Move vertical to point 5, (6.0; 4.25). Click the left mouse button to generate a vertical
line.
Move horizontal to point 6, (8.0; 4.25). Click the left mouse button to generate a horizontal
line.
Generate a second cluster by clicking the left mouse button on coordinate (8.0; 4.0).
This action created the footing.
The proposed geometry does not include plates, hinges, geogrids, interfaces, anchors or
tunnels. Hence, you can skip the corresponding buttons in the second toolbar.
Hints: Mispositioned points and lines can be modified or deleted by first choosing the
Selection button from the toolbar. To move a point of line, select the point or
the line and drag it to the desired position. To delete a point or a line, select the
point or the line and press the Delete key on the keyboard.
> Undesired drawing operations can be restored by pressing the Undo button
from the toolbar or by selecting the Undo option from the Edit menu or by
pressing <Ctrl><Z> on the keyboard.
Hint: The full geometry model has to be completed before a finite element mesh can be
generated. This means that boundary conditions and model parameters must be
entered and applied to the geometry model first.
Hint: During the input of geometry lines by mouse, holding down the Shift key will
assist the user to create perfect horizontal and vertical lines.
Input of boundary conditions

Boundary conditions can be found in the second block of the toolbar and in the Loads menu.
For deformation problems two types of boundary conditions exist: Prescribed displacements
and prescribed forces (loads). In principle, all boundaries must have one boundary condition
in each direction. That is to say, when no explicit boundary condition is given to a certain
boundary (a free boundary), the so-called natural condition applies, which is a prescribed
force equal to zero and a free displacement. In order to avoid the situation where the displacements
of the geometry are undetermined, some points of the geometry must have prescribed
displacements. The simplest form of a prescribed displacement is a fixity (zero displacement),
but non-zero prescribed displacements may also be given.
To create the boundary conditions for this exercise, follow the steps below.
49 of 375
Prescribed displacements
Click on the Standard fixities button on the toolbar or choose the Standard fixities option
from the Loads menu to set the standard boundary conditions. As a result PLAXIS will
automatically generate a full fixity at the base of the geometry and roller conditions at the
vertical sides (ux =0; uy =free). A fixity in a certain direction is presented as two parallel lines
perpendicular to the fixed direction. Hence, the rollers appear as two vertical parallel lines and
the full fixity appears as cross-hatched lines.
Hint: The Standard fixities option is suitable for most geotechnical applications. It is
a fast and convenient way to input standard boundary conditions.
Vertical load
Click on the Point load - load system A button on the toolbar or choose the Point load
- static load system A option from the Loads menu to enter another point force. Click on the
coordinate (7.0, 4.25) to enter a point force. As a result PLAXIS will automatically generate a
vertical point force on the indicated point with a unity force (f = 1).
Horizontal load (see also next step "Changing direction .....")
Click on the Point load - load system B button on the toolbar or choose the Point load -
static load system B option from the Loads menu to enter a point force. Click on the coordinate
(7.0, 4.25) to enter a point force. As a result PLAXIS will automatically generate a vertical point
force on the indicated point. As a horizontal force is needed, the direction of load B needs to
be changed.
Changing direction and magnitude of loads
Choose the Selection button from the toolbar. Double click on the geometry point 8 with
coordinate (7.0, 4.25) which will display a box as indicated in Figure 6. Select Point Load -
load system B, click OK and enter 1.0 as x-value and 0.0 as y-value. These values are the
input load of point force B. Click OK to close the window.
Input of material properties

In order to simulate the behaviour of the soil, a proper soil model and corresponding parameters
must be applied to the geometry. In PLAXIS, soil properties are collected in material data sets
and the various data sets are stored in a material database. From the database, a data set
can be assigned to one or more clusters. For structures (like walls, plates, anchors, geogrids,
etc.) the system is similar, but obviously different types of structures have different parameters
and thus different types of data sets. PLAXIS distinguishes between material data sets for Soil
50 of 375
Figure 6: Select window and Point load window
& Interfaces, Plates, Anchors and Geogrids. The creation of material data sets is generally
done after the input of boundary conditions. Before the mesh is generated, all material data
sets should have been defined and all clusters and structures must have their appropriate data
set.
Table 1: Material properties of the clay layer and the concrete footing.
Parameter Symbol Clay Concrete Unit
Material model Model Mohr-Coulomb Linear elastic
Type of behaviour Type Drained Non-porous
Weight above phreatic level unsat 16.0 24.0 kN/m3
Weight below phreatic level sat 18.0 kN/m3
Youngs modulus Eref 5.0103 2.0107 kN/m2
Poissons ratio 0.35 0.15
Cohesion c 5.0 kN/m2
Friction angle 20
Dilatancy angle 0
The input of material data sets can be selected by means of the Material Sets button on
the toolbar or from the options available in the Materials menu.
Create material data sets
To create a material set for the clay layer, follow these steps:
Select the Material Sets button on the toolbar.
Click on the <New> button at the lower side of the Material Sets window. A new dialog
box will appear with five tabsheets: General, Parameters, Flow parameters, Interfaces
and Initial (see figure 7).
In the Material Set box of the General tabsheet, write Clay in the Identification box.
Select Mohr-Coulomb from the Material model combo box and Drained from the Material
type combo box.
51 of 375
Enter the proper values for the weights in the General properties box according to the
material properties listed in table 1
See also figure 8 and figure 9. In these figures the Advanced parameters part has been
collapsed.
Figure 7: General tabsheet of the soil and interface data set window for Clay
Click on the Next button or click on the Parameters tabsheet to proceed with the input of
model parameters. The parameters appearing on the Parameters tabsheet depend on
the selected material model (in this case the Mohr-Coulomb model).
Enter the model parameters of table 1 in the corresponding edit boxes of the Parameters
tabsheet. The parameters in the Alternatives and Velocities group are automatically
calculated from the parameters entered earlier.
Since the geometry model does not include groundwater flow or interfaces, the third and
fourth tabsheet can be skipped. Click on the OK button to confirm the input of the current
material data set.
Now the created data set will appear in the tree view of the Material Sets window.
For the concrete of the footing repeat the former procedure, but choose a Linear Elastic
material behaviour and enter the properties for concrete as shown in table 1 (see also
figures 9 and 10).
52 of 375
Figure 8: Parameters tabsheet of the soil and interface data set window for Clay
Figure 9: General tabsheet of the soil and interface data set window for Concrete
Figure 10: Parameters tabsheet of the soil and interface data set window for Concrete
53 of 375
Assigning material data sets to soil clusters
Drag the data set Clay from the Material Sets window (select it and keep the left mouse
button down while moving) to the soil cluster in the draw area and drop it there (release
the left mouse button). Notice that the cursor changes shape to indicate whether or not
it is possible to drop the data set. When a data set is properly assigned to a cluster, the
cluster gets the corresponding colour. Drag the concrete material set to the footing and
drop it there.
Click on the OK button in the Material Sets window to close the database.
Hint: PLAXIS distinguishes between a project database and a global database of
material sets. Data sets may be exchanged from one project to another using
the global database. In order to copy such an existing data set, click on the
Show global button of the Material Sets window. Drag the appropriate data set
(in this case Clay) from the tree view of the global database to the project
database and drop it there. Now the global data set is available for the current
project. Similarly, data sets created in the project database may be dragged
and dropped in the global database.
Hints: Existing data sets may be changed by opening the material sets window,
selecting the data set to be changed from the tree view and clicking on the Edit
button. As an alternative, the material sets window can be opened by double
clicking a cluster and clicking on the Change button behind the Material set box
in the properties window. A data set can now be assigned to the corresponding
cluster by selecting it from the project database tree view and clicking on the
OK button.
> The program performs a consistency check on the material parameters and will
give a warning message in the case of a detected inconsistency in the data
Mesh generation
When the geometry model is complete, the finite element model (mesh) can be generated.
PLAXIS includes a fully automatic mesh generation procedure, in which the geometry is
automatically divided into elements of the basic element type and compatible structural elements,
if applicable. The mesh generation takes full account of the position of points and lines in the
geometry model, so that the exact position of layers, loads and structures is reflected by
the finite element mesh. The generation process is based on a robust triangulation principle
that searches for optimised triangles, which results in an unstructured mesh. This may look
disorderly, but the numerical performance of such a mesh is usually better than for regular
(structured) meshes. In addition to the mesh generation itself, a transformation of input data
(properties, boundary conditions, material sets, etc.) from the geometry model (points, lines
and clusters) to the finite element mesh (elements, nodes and stress points) is made.
In order to generate the mesh, follow these steps:
Click on the Generate mesh button in the toolbar or select the Generate option from
the Mesh menu. After the generation of the mesh a new window is opened (PLAXIS
Output window) in which the generated mesh is presented (see Figure 11).
54 of 375
Click on the Close button to return to the geometry input mode.
Figure 11: Generated finite element mesh of the geometry around the footing
If necessary, the mesh can be optimised by performing global or local refinements. Mesh
refinements are considered in some of the other exercises. Here it is suggested to accept the
current finite element mesh.
Hints: By default, the Global coarseness of the mesh is set to M edium, which is
adequate as a first approach in most cases. The Global coarseness setting
can be changed in the M esh menu. In addition, there are options available to
refine the mesh globally or locally.
> At this stage of input it is still possible to modify parts of the geometry or to add
geometry objects. In that case, obviously, the finite element mesh has to be
regenerated.
Press the close button to close the output program and return to PLAXIS input.
Creating the input for this project now finished. Press the green Calculation button on the
toolbar to continue with the definition of the calculation phases.
55 of 375
CALCULATION
After the finite element model has been created, the calculation phases need to be defined.
This analysis consists of four phases. In the initial phase the initial pore pressures and
stresses are generated, in the first phase the footing is constructed, during the second phase
the vertical load is applied and in the third phase the horizontal load is applied.
When starting the PLAXIS Calculation program the Calculation mode window appears. In
this window the user can choose how he wants PLAXIS to handle pore pressures during the
calculation. This is important when calculating with undrained behaviour and/or groundwater
flow. In this first exercise this is not important and so the default setting of Classical mode is
chosen. Press <OK> to close the Calculation mode window. PLAXIS now shows the General
tabsheet of the initial phase (see Figure 12).
Figure 12: General tabsheet of the initial calculation phase
Initial phase (generation of initial conditions)

Before starting the construction of the footing the initial conditions must be generated. In
general, the initial conditions comprise the initial groundwater conditions, the initial geometry
configuration and the initial effective stress state. The clay layer in the current footing project is
fully saturated with water, so groundwater conditions must be specified. On the other hand, the
situation requires the generation of initial effective stresses. As we want to include the footing
construction in the simulation process, the footing should not be present in the initial situation
(prior to construction). In PLAXIS it is possible to switch off clusters in order to calculate
correct initial effective stresses. The initial stresses in this example case are generated using
the K0 -procedure. The initial conditions are entered in separate modes of the Input program.
In order to generate the initial conditions properly, follow these steps:
In the phase list select the initial phase
56 of 375
Make sure the Calculation type is set to K0 -procedure on the General tabsheet. This is
the default setting.
Go to the Parameters tabsheet by clicking the Parameters button or by directly selecting

the tabsheet.
On the Parameters tabsheet press the Define button located in the Loading input box.
This will start a window presenting the problem in Staged construction mode. In Staged
construction mode it is possible to switch on and off various parts of the geometry,
change loads, apply strains etc.
In the initial condition of this exercise, that is the situation before we start constructing
our project, the footing is not present. Therefore the footing has to be deactivated. In
order to do so, click on the area that represents the footing so that it will change color
from the material set color to white. The footing is now disabled.
Click on Water conditions in the button bar in order to move to the Water conditions
mode of the program.
Select the Phreatic level button .
Position the cursor (appearing as a pen) at coordinate (0.0, 4.0) and click the left mouse
button to start the phreatic level.
Move along the x-axis to position (14.0, 4.0). Click the left mouse button to enter the
second point of the phreatic level.
Press the Water pressures button to view the pore pressures.
The pore pressures are generated from the specified phreatic level and the water weight.
Directly after the generation, a PLAXIS Output window is opened, showing the pore pressure
as presented in Figure 13. The colors indicate the magnitude of pore pressure. The pore
pressures vary hydrostatically, ranging from 0 kN/m2 at the top to -40 kN/m2 at the bottom.
Close the output program in order to return to the input program.
Click on Update in order to save the changes made and return to the PLAXIS Calculations
program. This completes the definition of the initial conditions.
Hints: For the generation of initial stresses based on the K0 procedure it is necessary
to specify the coefficient of lateral earth pressure, K0 . This K0 value is defined
per material set and therefore has to be set when entering material set data. If
the K0 value is not explicitly set PLAXIS uses a value according to Jakys
formula (K0 = 1-sin()).
> The K0 procedure may only be used for horizontally layered geometries with a
horizontal ground surface and, if applicable, a horizontal phreatic level.
57 of 375
Figure 13: Initial pore pressures
First calculation phase (construction of footing)
Click on the Next button . This will introduce a new calculation phase and
present the corresponding tabsheets for the first calculation stage. Enter a suitable name
in the Number/ ID box (e.g. Construction of footing).
Select the second tabsheet called Parameters. On this sheet Staged construction is
selected by default in the Loading input combo box. Click the Define button. This will
open the window presenting the problem in Staged construction mode.
Click on the cluster that represents the strip footing, in order to switch on the footing
(original colour should reappear).
Click on Update to conclude the definition of the first calculation phase. Updating will
automatically present the calculation window.
Second calculation phase (apply vertical load)

Click on the Next button . This will introduce a new calculation phase and present the
corresponding tabsheets for the second calculation stage. Enter a suitable name in the
Number/ ID box (e.g. apply vertical load).
Select the Parameters tabsheet. On this tabsheet accept the selection Staged construction
in the Loading input combo box. Click on the Define button. This will open the window
presenting the problem in Staged construction mode.
Click on the point forces in the middle of the footing, a Select items window comes up.
Select the Point load - Load System A to activate point load A and press the Change
button to change the load value. Change the y-value to -50 kN/m and press the Ok
button.
58 of 375
Figure 14: Parameters tabsheet of the first calculation phase
The point load A is now active (blue) and has a load value of 50 kN/m.
Press Update.
Figure 15: Select items window
Third calculation phase (add horizontal load)

Click on the Next button to add another phase. This will present the tabsheets for
the third calculation stage. Enter a suitable name in the Number/ID box (e.g. apply
horizontal load).
Select the second tabsheet called Parameters. On this sheet accept the selection
Staged construction in the Loading input combo box. Click on the Define button.
Click on the point forces in the middle of the footing, select the Point load - load system B
to activate point load B and press the Change button to change the load value. Change
the load x-value to 20 kN/m2 and press the Ok button.
Press the Ok button to closed the Select items window.
Press Update.
59 of 375
Fourth calculation phase (vertical load to failure)

Click on the Next button . This will present the tabsheets for the fourth calculation stage.
Enter a suitable name in the Number/ID box (e.g. vertical load failure).
Directly below the Number/ID box select from the Start from phase dropdown list the
second calculation phase. By selecting this the 4th phase will be a continuation of the
2nd phase, hence we will continue to apply the vertical load without having the additional
horizontal load that was applied in phase 3.
Select the second tabsheet called Parameters. On this sheet choose the selection Total
multipliers in the Loading input group box. Select the third tabsheet called Multipliers by
either clicking on the Define button or directly selecting the tabsheet.
Enter a MloadA of 10. In this way the working force is increased to a maximum load of
10 x 50 = 500 kN/m.
In PLAXIS two methods exist to increase an active load. The magnitude of the
activated load is the input load multiplied by the total load multiplier. Hence, in
this excersise MloadA x (input load of point load A) = Active load A
The value of the input load A can be changed using Staged construction as
Loading input while using Total multipliers as Loading input may be used to
change the load multiplier.
Define load displacement points

After the calculation it is possible to create load-displacement curves. These can be used
to inspect the behaviour in a node during the calculation steps. In order to create load-
displacement curves it is first necessary to indicate for which node(s) the displacements
should be traced.
Click on the Select points for curves button in the toolbar. This will result in a plot of
the mesh, showing all generated nodes. Click on the node, located in the centre directly
underneath the footing. For a correct selection of this node it may be necessary to use
the zoom option . After selection of the node it will be indicated as point A. Press
the Update button to proceed to calculations.
Start the calculation

After definition of the last calculation phase, the calculation process is started by clicking the
Calculation button . This will start the calculation. During the calculation a calculation
window appears showing the status and some parameters of the current calculation phase.
60 of 375
INSPECT OUTPUT
After each successful execution of a calculation phase PLAXIS will indicate the phase with
a green check mark ( ). This indicates a successful calculation phase. If during execution
either failure or an error occurs, PLAXIS marks the stage with a red cross ( ).
Figure 16: Calculation window with all phases calculated
While phase 3 is highlighted, press the View calculation results button that will start
the output program, showing the deformed mesh for the situation with both horizontal
and vertical load applied, as presented in figure 17.
Figure 17: Deformed mesh at the end of phase 3
61 of 375
Check the various types of output, such as the deformed mesh, displacement contours,
effective (principal) stresses etc. These can be found from the Deformations and
Stresses menus.
Still in the Output program, select from the dropdown list at the right of the toolbar the
output step belonging to phase 4.
From the Displacements menu in the Output program now select Incremental
displacements and then the option |u|. Display the incremental displacements as
contours or shadings. The plot clearly shows a failure mechanism (see Figure 18).
Figure 18: Shadings of displacement increments after phase 4
Load displacement curves

In the Output program, select the Curves manager from the Tools menu. The Curves
manager has 2 tabsheets, one for the curves defined in this project (currently none) and
one for the points selected to make load-displacement curves (currently 1 node that was
pre-selected, that is before the calculation).
In the Curves manager select the button New to define a new curve. Now the Curve
generation window opens.
On the x-axis we want to plot the settlement of our chosen point in the middle of the
footing. In the x-axis box choose point A from the dropdown list and then below in
Deformations and then Total displacements choose |u|.
On the y-axis we want to plot the force applied on the footing, which is a global value
not connected to a specific node or stress point. In y-axis box choose Project from the
dropdown list to indicate we want to plot a global value, and then in Multipliers choose
MLoadA.
62 of 375
Figure 19 shows the Curve generation window after applying the steps mentioned.
Press OK to show the resulting curve. See also figure 20.
Figure 19: Curves generation window
Figure 20: Load displacement curve for the footing
The input value of point load A is 50 kN/m and the load multiplier MloadA reaches approximately
4.6. Therefore the failure load is equal to 50 kN/m x 4.6 = 230 kN/m. You can inspect the load
multiplier by moving the mouse cursor over the plotted line. A tooltip box will show up with the
data of the current location.
63 of 375
RESULTS DRAINED BEHAVIOUR

In addition to the mesh used in this exercise calculations were performed using a very coarse
mesh with a local refinement at the bottom of the footing and a very fine mesh. Fine meshes
will normally give more accurate results than coarse meshes. In stead of refining the whole
mesh, it is generally better to refine the most important parts of the mesh, in order to reduce
computing time. Here we see that the differences are small (when considering 15-noded
elements), which means that we are close to the exact solution. The accuracy of the 15-
noded element is superior to the 6-noded element, especially for the calculation of failure
loads.
Hint: In plane strain calculations, but even more significant in axi-symmetric
calculations, for failure loads, the use of 15-noded elements is recommended.
The 6-noded elements are known to overestimate the failure load, but are ok
for deformations at serviceability states.
The results of fine/coarse and 6-noded/15-noded analyses are given below.
Table 2: Results for the maximum load reached on a strip footing on the drained sub-soil for
different 2D and 3D meshes
Mesh size Element Nr. of Max. Failure

type elements load load
[kN/m] [kN/m2 ]
very coarse mesh with local refinements 6-noded 79 281 146
under footing
coarse mesh 6-noded 121 270 141
very fine mesh 6-noded 1090 229 121
very coarse mesh with local refinements 15-noded 79 236 124
under footing
coarse mesh 15-noded 121 248 130
very fine mesh 15-noded 1090 220 116
Analytical solutions of:
- Vesic 117
- Brinch Hansen 98
- Meyerhof 97
In this table the failure load has been calculated as:
Qu M aximum f orce M aximum f orce

B
= B
+ concrete d = 2
+6
From the above results it is clear that fine FE meshes give more accurate results. On the other
hand the performance of the 15-noded elements is superior over the performance of the lower
order 6-noded elements. Needless to say that computation times are also influenced by the
number and type of elements.
64 of 375
ADDITIONAL EXERCISE:
UNDRAINED FOOTING
65 of 375
66 of 375
INTRODUCTION
When saturated soils are loaded rapidly, the soil body will behave in an undrained manner, i.e.
excess pore pressures are being generated. In this exercise the special PLAXIS feature for
the treatment of undrained soils is demonstrated.
In PLAXIS, one generally enters effective soil properties and this is retained in an undrained
analysis. In order to make the behaviour undrained one has to select undrained as the Type
of drainage. Please note that this is a special PLAXIS option as most other FE-codes require
the input of undrained parameters e.g. Eu and u .
Aims
The understanding and application of undrained soil behaviour
How to deal with excess pore pressures.
A) Geometry input
Use previous input file
Save as new data file
Change material properties, undrained behaviour for clay
Mesh generation, global mesh refinement
B) Calculations
Re-run existing calculation phases
Construct footing
Apply vertical force
Apply horizontal force
C) Inspect output
Inspect excess pore pressures
67 of 375
GEOMETRY INPUT
Use previous input file

Start PLAXIS by clicking on the icon of the Input program.
Select the existing project file from the last exercise (drained footing).
From the File menu select Save As and save the existing project under a new file name
(e.g. exercise 1b)
Change material properties

Change material properties by selecting the item Soils & Interfaces from the Materials menu
or click on the Material sets button . Select the clay from the Material sets tree view and click
on the Edit button. On the first tab sheet, General, change the Drainage type to "Undrained
A" and close the data set.
Figure 21: Set drainage type to "Undrained A"
Mesh generation
The mesh generator in PLAXIS allows for several degrees of refinement. In this example
we use the Refine global option from the Mesh menu, which will re-generate the mesh,
resulting in an increased number of finite elements to be distributed along the geometry lines.
Notice the message that appears about staged being reconstructed: the program will take into
account the newly generated mesh for the previously generated initial conditions and staged
construction phases. From the output window, in which the mesh is shown, press the continue
button to return to the Input program.
Hint: After generation of a finer mesh, the geometry may be refined until a
satisfactory result appears. Besides the option Refine global several other
methods of refinement can be used.
68 of 375
Hint: After re-generation of the finite element mesh new nodes and stress points
exists. Therefore PLAXIS has to regenerate pore water pressures and initial
stresses. This is done automatically in the background when regenerating the
mesh. Also, the new mesh is taken into account for any change to calculation
phases with the exception of ground water flow analysis.
After generating the mesh one can now continue to the calculation program. Click on the
Caculations button to proceed to the calculations program. Click yes to save the data.
CALCULATIONS
Re-run existing calculation list

The calculation list from example 1 appears, as indicated below. All phases are indicated
by (blue arrows). After mesh (re)generation, staged construction settings remain and phase
information has been rewritten automatically for the newly generated mesh. However, this is
not the case for points for load displacement curves due to the new numbering of the mesh
nodes.
Click on the Select points for curves button in the toolbar. Reselect the node located
in the centre directly underneath
Click on the Calculate button to recalculate the analysis. Due to undrained behaviour
of the soil there will be failure in the 3rd and 4th calculation phase.
INSPECT OUTPUT
As mentioned in the introduction of this example, the compressibility of water is taken into
account by assigning undrained behaviour to the clay layer. This results normally, after
loading, in excess pore pressures. The excess pore pressures may be viewed in the output
window by selecting:
Select in the calculation program the phase for which you would like to see output results.
Start the output program from the calculation program by clicking the View output button .
Select from the Stresses menu the option Pore pressures and then pexcess , this results in
Figure 22 .
The excess pore pressures may be viewed as stress crosses ( ), contour lines ( ),
shadings ( ) or as tabulated output ( ). If, in general, stresses are tensile stresses
the principal directions are drawn with arrow points. It can be seen that after phase 3 on the
left side of the footing there are excess pore tensions due to the horizontal movement of the
footing. The total pore pressures are visualised using the option of active pore pressures.
These are the sum of the steady state pore pressures as generated from the phreatic level
and the excess pore pressures as generated from undrained loading.
69 of 375
Figure 22: Excess pore pressures at the end of the 3rd phase
Select from the Stresses menu the option Pore pressures and then pactive . The results
are given in Figure 23.
From the load displacement curve it can be seen that the failure load in the last phase is
considerably lower for this undrained case compared to the drained situation, as expected.
For the undrained case the failure load is approx. 70 kPa.
70 of 375
Figure 23: Active pore pressures at the end of phase 3
71 of 375
72 of 375
APPENDIX A: BEARING CAPACITY CALCULATION

Given the formula for bearing capacity of a strip footing:
Qf
B
= c Nc + 12 0 B N
0
Nq = e tan tan2 (45 + 12 0 )
0
q 1) cot
Nc = (N
0
2(Nq + 1) tan
(V esic)
N = 1.5(Nq 1) tan 0 (Brinch Hansen)

(Nq 1) tan(1.4 0 ) (M eyerhof )

Filling in given soil data:
Nq = e tan(20) tan2 (55) = 6.4

1) cot(20) = 14.84
Nc = (6.4
2(6.4 + 1) tan(20) = 5.39
(V esic)
N = 1.5(6.4 1) tan(20) = 2.95 (Brinch Hansen)

(6.4 1) tan(28) = 2.97 (M eyerhof )

The effective weight of the soil:
0 = w 10 kN/m3 = 18 10 = 8 kN/m3
For a strip foundation this gives:

1 2
5 14.83 + 2 8 2 5.39 117 kN/m
(V esic)
Qf
B
= c Nc + 12 0 B N = 5 14.83 + 12 8 2 2.95 98 kN/m2 (Brinch Hansen)

5 14.83 + 21 8 2 2.87 97 kN/m2 (M eyerhof )

Qf
L=
B
III
I
II
73 of 375
Non-linear calculations in PLAXIS

Dr William Cheang
Regional Technical Manager (Plaxis AsiaPac)
Acknowledgement: PowerPoint slides by Dr Paul Bonnier
Plaxis Introductory Course, Danang 10-12 July 2013 1 / 27
Content
1. Learning
g objectives
j
2. Introduction
p
3. Multipliers
4. Iteration process
5. Plastic points
p
6. Recommendations
Plaxis Introductory Course,74 of 375 10-12 July 2013

Danang 2 / 27
L
Learningobjectives
i bj i
1. To recognize the items in the calculation progress window
1. To be able to evaluate the progress of a calculation
2. To use the calculation control parameters appropriately
3. To understand and explain the calculation procedure
I
Introduction
d i
1 Load multipliers
1.

Danang 4 / 27
I
Introduction
d i
1 Load multipliers
1.
2. Miscellaneous parms
I
Introduction
d i
1 Load multipliers
1.
3. Load-displ. curve

Danang 6 / 27
I
Introduction
d i
1 Load multipliers
1.
I
Introduction
d i
1 Load multipliers
1.
5. Plastic points

Danang 8 / 27
L d
Loadmultipliers
l i li
Applied load = Load multiplier x Input load
Defaults:
1. Load multiplier = 1
2 Input
2. I t load
l d = 1 unit
it
Loading input:
1. Staged construction: Change Input load
2. Total multipliers: Change Load multiplier (M)
3. Incremental multipliers: Change Load multiplier (M)
T t l multiplier
Total lti li (phase)
( h ) = Sum
S off incremental
i t l multipliers
lti li (step)
( t )
L d
Loadmultipliers
l i li
MdispX
p : Tot. mult. p
prescribed x-displacements
p
MdispY : Tot. mult. prescribed y-displacements
MloadA : y
Tot. mult. loads system A
MloadB : Tot. mult. loads system B
Mweight
g : Tot. mult. soil & structural weights
g
Maccel : Tot. mult. pseudo-static acceleration
Msf : Tot. mult. Phi-c reduction process
Mstage : Tot. mult. staged-construction process

Danang 10 / 27
L d
Loadmultipliers
l i li I
Incrementalmultipliersinput
l l i li i
L d
Loadmultipliers
l i li T l
Totalmultipliersinput
l i li i

Danang 12 / 27
Mi ll
Miscellaneousparameters
PMax : Maximum ((excess)) p
pore pressure
p in the model
Marea : Relative part of the mesh area currently active
Force-X : Reaction force due to horizontal p
prescr. displ.
p
Force-Y : Reaction force due to vertical prescribed displ.
Stiffness : Current ((relative)) Stiffness Parameter
Time : Elapsed model time (usually in days)
y
Dynamic time : Elapsed model time for dynamics
y ((s))
L d di l
Loaddisplacementcurve
1. Evaluation of calculation progress:
p g
1. Multipliers
2. Stiffness (CSP)
( )
3. Pmax
p
4. Load-displacement curve
5. Iterations
6. Global error
7. Plastic points

Danang 14 / 27
I
Iterationprocess
i
Calculation phase
Load steps (q)
Equilibrium iterations
constitutive model
q displacement strain stress reaction
Equilibrium?
I
Iterationprocess
i
Load q
Elastic stiffness (K)
qex Non-linear
iterations behaviour
Unbalance
Load step q
qin
Settlement of Node A

Danang 16 / 27
I
Iterationprocess
i
Current step
p Max. step
p Additional steps
p
Iteration Max. iterations Maximum iterat.
Unbalance Global error Tolerance Tolerated error

Control parameters
I
Iterationprocess
i C
Controlparameters
l

Danang 18 / 27
I
Iterationprocess
i O
Overrelaxation
l i
Standard setting: 1.2

Absolute maximum: 2.0 q0+q
Low s (<20): 1.5 acceptable A
Recommendation: Do not change
q0 B
A B
Overrelaxation
A
uo u
I
Iterationprocess
i A l
Arclengthcontrol
h l
Standard setting: Active

p
Purpose: ULS situations
Problems: Sometimes
spontaneous
unloading
Solution: Switch off
(before ULS)

Danang 20 / 27
I
Iterationprocess
i A l
Arclengthcontrol
h l
qP qP
0 0+ +
1 P = I Pe Pc I = const.
const
K
Pq0
P0 0 Arc length control
uo u
I
Iterationprocess
i D i d i i
Desiredminimum/maximum
/ i
Standard setting: Des. min = 6

Des. max = 15
Purpose: Automatic load
advancement

Danang 22 / 27
I
Iterationprocess
i D i d i i
Desiredminimum/maximum
/ i
q
Converged within desired minimum Scaling up
number of iterations:
Scaling up load step by a
factor 2 Scaling down
Not converged within desired
maximum number of iterations:
Scaling down load step by a
Scaling up
factor 2
Pl i
Plasticpoints
i
-1 Cap (HS, SS and SS-Creep model)
Cap point
Mohr-Coulomb
failure surface
Mohr-Coulomb point
f<0
Shear h
Sh hardening
d i Cap & Hardening point
yield surface (HS model)
Hardening point
-3
Apex point -1
Tension point
-3
Tension cut-off: Principal tensile stress is set to zero
Danang 24 / 27
Pl i
Plasticpoints
i
Local error criterion: Constitutive stress c:
Stress that follows from the constitute

model (Mohr- Coulomb)
eqq
Equilibrium
E ilib i stress
t eq:
Stress that is in equilibrium with the
c external load
|| c eq ||
Local Error
|| c ||
Inaccurate point:
Local error > Tolerated error
Convergence requirement:
I
Inaccurate
t stress
t points
i t 3 + ((plastic
l ti soilil points)
i t ) /10
Inaccurate interface points 3 + (plastic interface points) /10
R
Recommendations
d i
1. Use mostly
y defaults
2. Monitor and evaluate calculation progress
g
3. In case of bad convergence or numerical failure, check input
4. Use output facilities to trace input errors
5. In case input is right, consider control parameters
6. Dont change control parameters without understanding
consequences!
7. Dont increase tolerated error to speed up convergence!

Danang 26 / 27
END
87 of 375
Hardening Soil Model (HSM)

HS with Small-strain Overlay
y ((HS-small))
Dr William W.L. Cheang
Plaxis AsiaPac
Some slides by:

Professor Helmut Schweiger-TU Graz
Professor Pieter Vermeer-
A/P f
A/Professor T
Tan Si
Siew Ann-NUS
A NUS
Dr Lee Siew Wei-Golder Associates(HK)
88 of 375
INTRODUCTION
INTRODUCTION
89 of 375
90 of 375
91 of 375
92 of 375
93 of 375
94 of 375
LINES OF EQUAL SHEAR STRAINS (Tatsuoka & Ishihara, 1974)
95 of 375
Do you need plasticity when unloading (back into the yield locus)?
Yes..if the accumulation plastic volumetric strains are important in cyclically loaded
soils..dynamic liquefaction related boundary value problems
96 of 375
97 of 375
98 of 375
99 of 375
100 of 375
101 of 375
102 of 375
103 of 375
104 of 375
105 of 375
106 of 375
107 of 375
108 of 375
109 of 375
110 of 375
111 of 375
112 of 375
113 of 375
114 of 375
115 of 375
116 of 375
Surface Heave in Initial Exc./Cantilever Wall

3 m deep excavation with cantilever wall
20kPa
5m
3m
7m
Dry sandy material
FSP III sheetpile
3 analyses with Mohr Coulomb, Hardening Soil & Hardening Soil-Small models using equivalent
soil input parameters
Compare ground movements, wall displacements & wall stability
Soil Input Parameters for 3 Analyses

Parameters for soil strength & initial stress state
Analyses Material c' ' Rinter
Model (or ur)
(kN/m3) (kPa) (Deg) [-] [-]
1 MC 20 5 35 0.3 0.426 0.67
2 HS 20 5 35 0.2 0.426 0.67
3 HSsmall 20 5 35 02
0.2 0 426
0.426 0 67
0.67
Parameters for soil stiffness prior to failure

Analyses
y Material Eref Eurref pref m G0 0.7
Model (or E50 or Eoedref)
ref
(MPa) (MPa) (kPa) [-] (MPa) [-]

1 MC 30 - - - - -
2 HS 30 90 100 0.5 - -
3 HSsmall 30 90 100 0.5 150 210-5
For derivation of soil stiffness parameters,

- HS model
d l ffrom standard
t d dd drained
i d ttriaxial
i i l compression
i ttests
t
- HSsmall model from small-strain triaxial tests or field tests (e.g.
downhole / crosshole seismic survey)
117 of 375
Pre-failure Stress-strain Behaviour

1: Mohr Coulomb
1: Linear elastic,, p
perfectly
y plastic
p
2: Hyperbolic stress-strain curve
(stiffness degradation for > 1E-4)
3: Non-linear stiffness from very
small strains (1E-6)
3:Hardening Soil + Small Strain Overlay

2: Hardening Soil
1e-6 1e-5 1e-4 1e-3 1e-2 1e-1
Predicted Surface Settlement Behind Wall

Distance behind wall (m)
0 5 10 15 20 25 30
0.006
0.004 Heave
0.002
Setttlement (m)
0.000
-0.002
-0.004
Settlement
-0.006
MC
-0.008 HS
HSsmall
-0.010
MC predicts unrealistic surface heave 4 mm

HS & HSsmall predict max. surface settlement 9 mm
118 of 375
Predicted Heave at Exc. Level in Cofferdam

Distance in front of wall (m)
-5 -4 -3 -2 -1 0 1 2 3
0.025
MC
HS
Wall
0.020 HSsmall
0.015
eave (m)
0.010
He
0.005
0 000
0.000
-0.005
MC predicts 20 mm heave at cofferdam centreline

HS & HSsmall predict 11 mm & 8 mm respectively
Predicted Wall Resultant Displacement
MC HS HSsmall
Ux=6mm
Ux=11mm Ux=10mm
Ux: wall horizontal

119 displacement
of 375
Predicted Stability of Wall

3 FOS=2.8
2.5
MC Rotation mechanism
2 with
ith FOS 2
2.8
8
1.5
3 FOS=2.8
2.5
2
HS
1.5
Phi-c'
Phi c reduction
reduction for predicting FOS
3 FOS=2.8
FSP III sheetpile properties:
2.5 EI=34440 kNm2/m; EA=3
EA=3.9210
92106kN/m
2 HSsmall
Mp=369 kNm/m; Np=3575 kN/m
1.5
Summary of Predictions
Analyses Surface settlement Heave at Wall horizontal FOS for wall
behind wall excavation level displacement
p stability
y
MC Heave 4 mm Heave 20 mm 6 mm 2.8
(not OK)
HS Settle 9 mm Heave 11 mm 11 mm 2.8
HS
HSsmall
ll S l 9 mm
Settle H
Heave 8 mm 10 mm 28
2.8
MC predicts incorrect surface heave behind wall

- related to soil stiffness (E) prior to failure different ways of
modelling E in 3 constitutive models
Stability of wall has FOS = 2.8 for 3 analyses
- related to soil shear strength all 3 constitutive models use
M h C
Mohr Coulomb
l b ffailure
il criterion
it i with
ith c'=5 kPa & '=35
' 5 kP ' 35
120 of 375
Variation of Soil Stiffness in Excavation

A. Soil stiffness is not constant and varies with
1 stress-level.
1. stress level Higher stress,
stress higher stiffness
2. strain-level. Higher strain (or displacement), lower stiffness
3 stress-path
3. t th (recent
( t soilil stress
t history).
hi t ) Rotation
R t ti off stress
t path,
th
higher soil stiffness
4 anisotropy,
4. anisotropy destructuration
B. During excavation, soil elements at different locations experience
different changes
g in
1. stress,
2. strain
3. stress-path direction
Soil Stress Paths Near Excavation

GCO No.1/90
A: unloading compression;
B: unloading extension
Rotation of stress paths at121Aof&375B
Soil Stress Paths Near Excavation

20kPa 25 20kPa
Failure line
20
3m A K0
15 A
Exc. A
B 10
B
Exc.
t (kPa)
7m K0 20kPa B
5
0
5m
-5
-10
Failure line
A: unloading compression
-15
B: unloading extension 0 10 20 30 40 50 60
s' (kPa)
path at A, A 90 w.r.t. K0 direction

Rotation of stress p
Rotation of stress path at B, B 160 w.r.t. K0 direction
Stress Path Dependent Soil Stiffness

Shear modulus, 3G (MPa)

Stress path rotation,
t
=0

=180
K0
=90
s'
Atkinson et al
al. (1990)
Triaxial tests on
London Clay
Shear strain (%)
-1 -0.1 -0.01 0.01 0.1 1
=0,, no change
g in stress path
p direction
=180, full reversal of stress path direction
122 of 375
Stress Path Dependent CDG Stiffness

St
Stress-level
l l Test series
Extension
Compress
Compression
Extension
=90
Wang & Ng (2005)

At s 0.01%, shear stiffness in extension 60% higher
g than in compression
p
Why MC Predicts Incorrect Surface Heave?

MC models a constant soil stiffness prior to failure not realistic
In reality
reality, stiffness of soil elements near excavation varies
according to
1. stress-level
2. strain-level
3 direction of stress
3. stress-path
path
Realistic prediction of wall deflections & ground settlements in all
excavation stages requires a constitutive model that considers
above factors, e.g. HS & HSsmall models
HS & HSsmall consider the interplay between factors (1), (2) & (3)
i d
in determining
t i i ththe operational
ti l soilil stiffness
tiff (E)
(E), ii.e. E iis changing
h i
during excavation
123 of 375
Simulation of laboratory tests
SIMULATION OF LABORATORY TESTS
124 of 375
125 of 375
INTRODUCTION
In daily engineering practice soil parameters are obtained from one or more laboratory tests. In order to perform
the best possible Plaxis calculation these soil parameters have to be translated into input parameters for the
constitutive model used, taking into account the possibilities and limitations of the constitutive model. Most
parameters for the constitutive models used in Plaxis can be determined directly from standard laboratory tests
as triaxial tests and oedometer tests. However, due to the complexity of the models it is recommended to
not simply accept the parameters determined from those tests, but to actually model the tests and see if the
parameters found actually give a proper representation of the real laboratory test results within the limits of the
constitutive models. For this purpose the SoilTest module is available in Plaxis with which in a simple manner
laboratory tests can be simulated without the need for making a finite element model.
In this exercise the SoilTest tool will be used for the simulation of both oedometer and triaxial tests on sand and
clay.
CONTENT
Laboratory tests on Sand

Laboratory tests on Clay
Appendix A: Parameter determination
Appendix B: Introduction to the SoilTest tool
How to model an oedometer test

How to model a triaxial test
SIMULATION OF LABORATORY TESTS

In this exercise results from oedometer and triaxial tests are presented for two different materials and the aim is
to determine the parameters for the Hardening Soil model such that a simulation of the tests within Plaxis gives
the best possible results compared to the original laboratory tests. In short:
1. Determine soil parameters based on given real laboratory tests results
2. Perform the laboratory tests using SoilTest with the parameters found
3. Match SoilTest results with the original laboratory results to find the best matching model parameters for
the Hardening Soil model.
Exercise 1: Laboratory tests on sand
Parameter determination
On a sample of dense sand both oedometer tests and triaxial tests have been performed. The results of those
tests are given in the figures below. Use these figures to determine the parameters for the Hardening Soil model
and collect the parameters in Table 1 (see below the figures). Note that it is possible that some parameters
cannot be determined with the given laboratory results, in which case these parameters have to be estimated.
126 of 375
Figure 1: Oedometer test results on sand
Figure 2: Development of horizontal and vertical stress in oedometer test
127 of 375
Figure 3: Triaxial test unloading-reloading (cell pressure = 100 kPa)
Figure 4: Axial vs. volume strain in drained triaxial test
128 of 375
Collect the soil parameters in table 1:
Table 1: Hardening Soil Parameters of the sand

Parameter Unit Value
ref
E50 [kPa]
ref
Eoed [kPa]
ref
Eur [kPa]
pref [kPa]
ur [-]
c [kPa]
0 [o ]
[o ]
m [-]
K0N C [-]
With these data perform a triaxial test in the SoilTest program.
129 of 375
Exercise 2: Laboratory tests on clay
Figure 5: Oedometer test on Clay
Figure 6: Undrained triaxial (CU) tests at cell pressures of 100 kPa and 400 kPa
130 of 375
Figure 7: Undrained triaxial (CU) test at cell pressure of 100 kPa
Collect the soil parameters in table 2:
Table 2: Hardening Soil Parameters of the clay

ref
E50 [kPa]
ref
Eoed [kPa]
ref
Eur [kPa]
pref [kPa]
ur [-]
c [kPa]
0 [o ]
[o ]
m [-]
K0N C [-]
With these data perform an oedeometer test in the SoilTest program.
131 of 375
APPENDIX A: PARAMETER DETERMINATION

SAND
First we determine parameters from the triaxial test data.
Figure 8: Determine stiffness parameters from drained triaxial test
Cohesion and friction angle
For a cell pressure 30 = 100 kPa a maximum value of approximately |10 30 | = 400 kPa is reached at failure.
The Mohr-Coulomb failure criterium is:
1 0
2 |1 30 | + 12 (10 + 30 ) sin c cos = 0
Considering it is sand we assume that the cohesion is zero and so the Mohr-Coulomb failure criterium reduces
to:
|10 30 |
(10 +30 ) = sin
Filling in 30 = 100 kPa and 10 = 500 kPa as obtained from the test we find for the
friction angle0 = 420
Reference stiffness from triaxial test
The triaxial test stiffness E50 is the secant stiffness over the first 50% of the failure value for | 10 30 |. This is
indicated in red in the triaxial test graph of figure 8.
0 =100 kP a 400
E503 = 0.013 = 30800 kP a
132 of 375
The triaxial test stiffness ,E 50 , is within the Hardening Soil model defined as:
m m
c cos30 sin 30

ref ref
E50 = E50 c cos+pref sin , c = 0 E50 = E50 pref
The reference stress pref is chosen equal to the cell pressure of this triaxial test then
ref 0 =100 kP a
E50 = E503 30000 kPa
Reference unloading-reloading stiffness
Similar to the determination of the reference stiffness for triaxial testing the reference unloading-reloading stiffness
can be determined. In the triaxial test results an unloading-reloading cycle is done for this. The Hardening Soil
model does not have unloading-reloading behaviour with hysteresis but simple non-linear elastic unloading-
reloading behaviour. Therefore a secant value is taken for the unloading-reloading behaviour, as given with the
green line in the triaxial test results.
0 =100 kP a 400
Eur3 = 0.0260.021 = 80000 kPa
Under the same assumptions as for the stiffness in triaxial testing counts:
ref 0 =100 kP a
Eur = Eur3
But this is a bit low value for the unloading reloading stiffness and so
ref
Eur = 90000 kPa
is chosen
Dilatancy angle
From the plot of axial strain versus volume strain the dilatancy angle can be determined according to
v
sin = 21 +v
See figure 9 for details.

With v = 0.048-0.004 = 0.044 and 1 = -0.09-(-0.03) = -0.06 the dilatancy can be calculated as =16o
Note: The Poissons ratio needed for the Hardening Soil model cannot be determined from this graph as this
graph represents an oedometer test in primary loading and the Poissons ratio needed is an unloading-reloading
Poissons ratio.
An acceptable value for the unloading-reloading Poissons ratio is ur = 0.2.
133 of 375
Figure 9: Determination of diltancy angle from drained triaxial test
Oedometer stiffness and power of stress dependent stiffness
From the oedeometer test results we determine the stiffness Eoed for vertical stresses y0 = 100 kPa en y0 =
200 kPa, see figure 10. Note that Eoed is a tangent stiffness. Make sure to use the primary loading part of the
oedometer test results.
0 =100 kP a
y 3200
Eoed = 1.4%0.33% = 29900 kPa
y0 =200 kP a 4000
Eoed = 1.4%0.47% = 43000 kPa
Within the Hardening Soil model the stress dependent oedometer stiffness is defined as:
m m
c cosy0 sin y0

ref ref
Eoed = Eoed c cos+pref sin , c = 0 Eoed = Eoed pref
Choosing the reference pressure pref = 100 kPa gives
ref 0 =100 kP a
Eoed = Eoed
3
30000 kPa
The power m for stress dependent stiffness can now be determined as:
0 =200 kP a m
y
y0

Eoed 43000 200 m

ref
Eoed
= pref 30000 = 100 m = 0.5
134 of 375
Figure 10: Determination of oedometer stiffness and power of stress dependency
K0 value for normal consolidation
The K0 value for normal consolidation (K0N C )can only be obtained if measurements for horizontal stresses have
been performed during the oedometer test. If so, results as given in figure 11 may be obtained. From the primary
loading line can be obtained that
0
x 30 100
K0N C = y0 = 10 = 300 = 0.33
Alternatively one can use Jakis formula
K0N C 1 sin = 1 sin(42o ) = 0.33
135 of 375
Figure 11: Horizontal/vertical stress ratio during oedometer test
Note on unloading-reloading stiffness
If no triaxial test with unloading-reloading is available the unloading-reloading stiffness can also be determined
from an oedometer test with unloading. However, the unloading-reloading stiffness required for the Hardening
Soil model is stress dependent on 3 while the oedometer test results presented in figure 10 give the strain vs
the vertical stress y (= 1 voor oedometer testing).
0 =100 kP a 0 =100/K0N C kP a 0 =300 kP a 400

Eur3 = Eur1 = Eur1 = 1.28%0.91% = 108000 kPa
With pref = 100 kPa (pref refers to 30 !) it follows that
ref 0 =100 kP a
Eur = Eur3 110000 kPa
ref
This is a bit high and so a value of Eur = 90000 kPa is chosen.
Table 3: Summary of Hardening Soil Parameters for the sand

ref
E50 [kPa] 30,000
ref
Eoed [kPa] 30,000
ref
Eur [kPa] 90,000
pref [kPa] 100
ur [-] 0.2
c [kPa] 0
0 [o ] 42
[o ] 16
m [-] 0.5
K0N C [-] 0.33
136 of 375
CLAY
Cohesion and friction angle
We start with the determination of the strength parameters based on the CU triaxial tests.
Figure 12: Determination of soil strength parameters for clay
The black dotted lines is the Mohr-Coulomb failure criterium in the p-q plane. In principal stresses the Mohr-
Coulomb failure criterium is defined as:
|1 3 | 1 +3

2 + 2 sin c cos = 0
With p0 = (10 + 230 )/3 and q = 10 30 under triaxial test conditions this can be rewritten as:
2p0 + 13 q

q 6sin 0 6c cos
2 = 2 sin c cos = 0 q = 3sin p + 3sin
Hence, the slope M of the Mohr-Coulomb line in p-q plane is defined as:
6sin 195
M= 3sin = 200 = 250
From the intersection between Mohr-Coulomb line and the vertical axis where p=0 the cohesion can be determined:
6c0 cos
q= 3sin = 0 c = 0 kPa
137 of 375
Reference oedometer and unloading-reloading stiffness
From the results of the oedometer test the oedometer stiffness as well as the unloading-reloading stiffness can
be determined. As the graph is given on logarithmic scale one cannot simply draw a tangent line as was done
for the oedometer test on sand.
Figure 13: Determination of oedometer and unloading/reloading stiffness
Considering that both primary loading and unloading/reloading paths are straight lines in the log(p)-v graph,
hence they have a relation of the form:
y = v = A log(y0 )
2 1 0.3700.270
A= log(2 )log(1 ) = log(120)log(30) =0.166
In order to determine the stiffness we calculate the derivative of the strain over the stress and change to natural
logarithm:
ln(y0 )
y = v = A ln(10)
dy dy0 ln(10)
dy0 =A 1
ln(10) 1
y0 E= dy = A y0
The E modulus found is the oedometer stiffness can be rewrittens as:
y0

ln(10)
E = Eoed = A pref pref
In the Hardening Soil model the oedometer stiffness is defined as (assuming c = 0) :
m
y0

ref
Eoed = Eoed pref
Hence:
ref ln(10)
Eoed = A pref and m=1
If we choose pref = 100 kPa and with the previously determined A = 0.166 we get:
138 of 375
ref ln(10) 2.3

Eoed = A pref = 0.166 100 = 1.4 MPa.
The determination of the unloading-reloading stiffness follows the same method:
y = v = B log(y0 )
2 1 0.4270418
B= log(2 )log(1 ) = log(120)log(30) =0.0149
dy0 ln(10)
Eur = dy = B y0
However, the Eur in the Hardening Soil model is dependent on the smallest principal stress, which is x0 in an
oedeometer test and not y0 .
During the unloading process there is no linear relation between horizontal and vertical stress, as in the beginning
of unloading y0 > x0 where as after much unloading y0 < x0 . Therefore the assumption is made that during
unloading on average x0 = y0 .
0

ln(10) ln(10) ln(10) x
Eur = B y0 = B x0 = B pref pref
With the definition of Eur in the Hardening Soil model of
0
m
ref x
Eur = Eur pref
ref
Follows, in a similar way as for the Eoed , that
ref ln(10) 2.3

Eur = B pref = 0.0149 100 =15 MPa and m = 1
Stiffness from triaxial test
As only undrained triaxial test data is available it is only possible to determine an undrained E50 and not an
effective E50 . Therefore the only solution is to estimate the E50 with several runs of the SoilTest program using
different input values for the reference E50 until the best fit for the undrained triaxial test data is found. Typically
for normally consolidated clays the effective reference E50 is in the range of 2-5 times the effective reference
ref
Eoed , hence this can be used as a start value for the estimation procedure. By doing so a value E50 3.5 MPa
of is found.
K0 value for normal consolidation
The K0-value for normal consolidation can only be obtained if measurements for horizontal stresses have been
performed during the oedometer test. As this is not the case here we can only use the estimation according to
Jakys rule:
K0N C 1 sin = 1 sin(250 ) =0.58
Poissons ratio
The Poissons ratio for unloading and reloading is again estimated as ur = 0.2
139 of 375
Table 4: Summary of Hardening Soil Parameters for the clay

ref
E50 [kPa] 3,500
ref
Eoed [kPa] 1,400
ref
Eur [kPa] 15,000
pref [kPa] 100
ur [-] 0.2
c [kPa] 0
0 [o ] 25
[o ] 0
m [-] 1.0
K0N C [-] 0.58
140 of 375
141 of 375
APPENDIX B: INTRODUCTION TO THE SOILTEST TOOL

For the simulation of laboratory tests Plaxis offers the SoilTest tool based on a single stress point calculation
that makes it possible to do fast simulations without the need for a finite element mesh. The SoilTest tool can be
called from within the material sets database or from within the definition of a material set. (see figure ).
Figure 14: The SoilTest tool
In the following paragraphs a step-by-step description is given on how to model both an oedometer test and a
triaxial test with the help of many screen shots of the SoilTest tool. Please note that any parameters given on
those screen shots have no relation with the actual exercise and are solely for illustrating the possibilities of the
SoilTest tool.
142 of 375
How to model an oedometer test
In order to model an oedometer test first the material data set has to be created. After doing so, press the
<SoilTest> button to start the SoilTest tool. The window that opens is show in figure .
Figure 15: Main window of the SoilTest tool
In the main window select the Oedometer tabsheet and set the parameters as indicated in Figure .
Figure 16: Setting the oedometer test parameters
After the the oedometer test has been calculating graphs with results appear at the bottom of the SoilTest window.
The user can double-click these graphs to view them in separate windows. Furthermore, custom charts can be
added, see figure 4.
143 of 375
Figure 17: Inspect oedometer test results
How to model a triaxial test

From the material database or the material set definition window press the <SoilTest> button to start the SoilTest
tool. In the main window choose the tabsheet Triaxial and set the type of test as well as the test parameters as
shown in figure
Figure 18: Defining a triaxial test
After the triaxial test has been calculated graphs with results appear at the bottom of the SoilTest window. As
described above for the oedometer test, the user can double-click this graphs to view them in separate windows
as well as add custom charts.
144 of 375
Modelling a triaxial test with unloading/reloading
The standard functionality in SoilTest for simulation of a triaxial test does not allow for an intermediate unloading-
reloading path. However, the SoilTest functionality contains a General option with which soil test can be defined
in terms of boundary stresses or strains on all sides of a soil test cube. Hereafter it will be shown how this can
be used for the simulation of a triaxial test with unloading/reloading path.
After opening the SoilTest option from the material set definition window the tabsheet General should be chosen.
On this tabsheet a list of calculation phases can be defined where stress or strain increments can be applied.
Initial phase
First of all we have to specify whether stresses or strains will be applied on the boundaries during the test. For
this exercise stresses will be applied. Now the values of the initial stresses on the soil sample have to specified.
For a triaxial test the initial stresses are the cell pressures acting on the soil, hence for xx , yy and zz the cell
pressure has to entered. The cell pressure is a water pressure and so there will be no shear stress acting on the
soil: xy = 0. See figure for details.
Figure 19: General option for simulation of laboratory tests used for triaxial test
Phase 1
Apply a stress increment in vertical direction (yy ) until the stress level where the unloading path should start.
Note that the horizontal stresses (xx and zz ) remain the same as they represent the cell pressure. Hence,
the horizontal stress increments are zero in this phase.
Phase 2
Press the Add button to add another phase to the phase list. This phase represents the unloading phase. See
figure for details.
Phase 3
Press the Add button once more in order to add the 3rd phase. This phase represents the reloading of the soil
as well as the continuation of primary loading until either failure or a higher stress level from where for instance
145 of 375
another unloading/reloading cycle is going to be made.
Figure 20: Unloading/reloading cycle in a triaxial test using the General option
146 of 375
DerivationofSoilParametersfrom
LabTestResults&Verificationin
PlaxisSoilTest
by
RFShen
23Nov2011
While engineers will use the c, , or Cu from SI

report, how many of them make use of the massive
stress-strain test data (which the client has spent a
lot of money for the lab to obtain such data) to
derive the soil stiffness parameters? Correlation with
SPT N values are too commonly used instead.
Singapore 2011
147 of 375
In this exercise, we are going to fully utilize the test

data to derive soil parameters for Hardening Soil
Parameters from most common stress-strain data
provided in a typical SI report, and subsequently
use Plaxis SoilTest to verify the derived parameters
Part 1: Sand
Singapore 2011
148 of 375
For sand, one of the most common lab tests is

Triaxial Isotropically Consolidated Drained (CID)Test
A Triaxial setup in NUS

Geotechnical Lab
For sand, one of the most common lab tests is

Triaxial Isotropically Consolidated Drained (CID)Test
Fa/A = q (deviatoric
stress)
Typical sample size 38 mm x 76 mm a = q + r
Singapore 2011
149 of 375
450
400
350
Deviatorstress(kPa)
300
250
200 3 = 100 kPa

150
100 Testdata
50
0
0 0.01 0.02 0.03 0.04 0.05 0.06 0.07
Axialstrain
Test data: Deviator stress ~ axial strain

curve (Triaxial)
0.06
0.05
0.04
Volumetricstrain
0.03
0.02
0.01
Testdata
0
0 0.02 0.04 0.06 0.08 0.1
0.01
Axialstrain
Test data: Volumetric strain ~ axial

strain curve (Triaxial)
Singapore 2011
150 of 375
Another common lab test is Oedometer Test
Oedometer setups in NUS

Geotechnical Lab
Settlementdialgauge
OedometerCell
Sample:dia.=75mm
Protruded
leverarm Height=20mm
Heavydeadweights
Singapore 2011
151 of 375
Typical sample size 75 mm x 20 mm
Boundary
conditions
0
0.1
0.2
0.3
0.4
Verticalstrain(%)
0.5
0.6
0.7 Testdata
0.8
0.9
1
1.1
1.2
1.3
1.4
0 100 200 300 400
Verticalpressure(kPa)
Vertical stress ~ vertical strain curve

(Oedometer)
Singapore 2011
152 of 375
400
300
200
100
Testdata
0
0 50 100 150 200
Lateralstress(kPa)
Vertical stress ~ lateral stress curve

(Oedometer)
Hardening Soil Parameters to be derived based on the

above typical lab test data
Singapore 2011
153 of 375
Part 1: Strength parameters
C=0 for sand
450
400
Since c = 0 for sand, it can be
350
simplified to:
Deviatorstress(kPa)
300
250
200 3 = 100 kPa

150
100 Testdata
50
0
0 0.01 0.02 0.03 0.04 0.05 0.06 0.07
500 100
sin ' 0.67
Axialstrain
500 100
' 42
Singapore 2011
154 of 375
0.06
0.05
So,
0.048
0.04
Volumetricstrain
1 sin 0.09 0.03

0.03
1-sin 1.36
2 sin 0.048 0.004
0.02
2sin
0.01
0.004
Testdata sin 0.27
0
0 0.02 0.04 0.06 0.08 0.1
0.03 0.09
16
0.01
Axialstrain
BTW, why there is an initial contraction

before the soil sample to dilate prominently
??
BTW, why there is an initial contraction

before the soil sample to dilate prominently
??
What contributes to the sample contraction? e dp '
(1) dp >0 elastic volumetric contraction! v K
(2) Isotropic hardening plastic volumetric contraction! 1 m
pc
vp ,cap
1 m p ref
What contributes to the sample dilation?
(1) As the stress path cut through series of shear yield line, plastic
shear strain d was generated.
p
(2) the plastic shear strain will be accompanied by plastic volumetric

strain by d vp , fric d p , fric sin m , and it is dilative!
q
MC line
pc p
Singapore 2011
155 of 375
Part 2: Stiffness parameters
450
400 400
350
Deviatorstress(kPa)
300
250
3 = 100 kPa
200
150
100 Testdata
50
0
0 0.01 0.02 0.03 0.04 0.05 0.06 0.07
0.013
Axialstrain
400
E50ref 30800 kPa 30000 kPa
0.013
Singapore 2011
156 of 375

450
400400
350
Deviatorstress(kPa)
300
3 = 100 kPa
250
200
150
100 Testdata
50
0
0 0.01 0.02 0.03 0.04 0.05 0.06 0.07
0.021 0.026
Axialstrain
400
Eurref 80000 kPa
0.026 0.021
As sand unload-reloading stiffness Eurref is generally

about 3~5 times of E50ref, we may set Eurref = 90000kPa
Singapore 2011
157 of 375
0
0.1
0.2
0.3
0.33
0.4
Verticalstrain(%)
0.5
0.6
0.7 Testdata
0.8
0.9
1
1.1
1.2
1.3
1.4
0 100 200 300 320 400
320
ref
Eoed 29900kPa 30000kPa
1.4% 0.33%

m
c cos ' ' 3 sin '
E50 E ref

50 ref
c cos ' p sin '
m
ref c cot ' '1
Eoed Eoed ref
c cot ' p
Singapore 2011
158 of 375
0
0.1
320
29900kPa 30000kPa
0.2 ref
Eoed
0.3
0.4
1.4% 0.33%
0.47
Verticalstrain(%)
0.5
400
0.6 200 kPa
Eoed 43000kPa
0.7
0.8
Testdata
1.4% 0.47%
0.9
1
1.1
1.2
1.3
1.4
0 100 200 300 400
Verticalpressure(kPa) 400
m
c cot ' '1 200
200 kPa m
Eoed 43000
m = 0.5
c cot ' p
ref ref
Eoed 100 30000
Part 3: Other parameters
Jakis formula:
K 0NC 1 sin ' 1 sin 42 0.33

400
300
200
100
Testdata
0
0 50 100 150 200
Lateralstress(kPa)
x ' 100
K 0NC 0.33
y ' 300
Singapore 2011
159 of 375
Summary of Hardening Soil Parameters
FEM simulation using Plaxis SoilTest Facility
(1) Change of dilation angle and see its effects
(2) How to simulate unload-reload step?
(3) Oedometer test simulation
Singapore 2011
160 of 375
Part 2: Clay
For Clay, one of the most common lab tests is Triaxial

Isotropically Consolidated UnDrained (CIU) Test
A Triaxial setup in NUS

Geotechnical Lab
Singapore 2011
161 of 375
For Clay, one of the most common lab tests is Triaxial

Isotropically Consolidated UnDrained (CIU)Test
Fa/A = q (deviatoric
stress)
Close the valve = Undrained test =
a = q + r
Excess will accumulate with shearing
350
Testdata
300
250
q(kPa)
200
195
150
100
50
0
0 50 100 150 200 250 300 350 400 450
p'(kPa)
Test data: stress path p~q
Singapore 2011
162 of 375
CIU stress path
Gradient:
350
Testdata
300
250 6 sin ' 195

3 sin ' 200
q(kPa)
200
195
150
100 = 25
50
0 Intercept:
0 50 100 150 200 250 300 350 400 450
p'(kPa)
6c ' cos '

0
3 sin '
c = 0
Oedometer setups in NUS

Geotechnical Lab
Singapore 2011
163 of 375
Typically less test points are available

due to long consolidation period for
each loading stage
Boundary
conditions
Testdata
0.1
Verticalstrain(%)
0.2
0.3
0.4
0.5
1 10 100 1000
Typically oedometer test results are presented in SI

report as logv ~ yy which is linear (unlike sand) which
must be dealt with cautions!
Singapore 2011
164 of 375
Oedometer test for clay

0
d y '
Testdata
Eoed
0.1
d y
Verticalstrain(%)
0.2
d (log y ' )
Gradient _ k
d y
0.3
Obviously, Eoed Gradient _ k

0.4
0.5
1 10 100 1000
ln y ' 1
d ( y ' )
d (log y ' ) d( )
2.3 1 y' 1 d ( y ' ) 1
Gradient _ k Eoed
d yy d yy 2.3 d yy 2.3 y ' d yy 2.3 y '
So, Eoed 2.3 y ' gradient _ k

0
Testdata
0.1
ref
Eoed 2.3 100 6.02 1350 kPa
Verticalstrain(%)
0.2
Eoed y '
0.27 ref

0.3
Eoed pref
0.37 m
0.4
ref c cot ' '1
Eoed Eoed ref
0.5
c cot ' p
1 10 30 100 1000
120 m
Eoed '1
ref
Eoed 2.3 y ' gradient _ k
ref
Eoed p
gradient _ k
log(120) log(30)
6.02
m=0
0.37 0.27
So, Eoed 2.3 y '6.02
Singapore 2011
165 of 375

Eur refers to when 3 =
0
100kPa
Testdata
0.1
During oedometer loading,

Verticalstrain(%)
when y =100kPa,
0.2
0.3 x<100kPa;
When y loaded to about
0.418
0.4
300kPa and unload to
0.427
0.5 100kPa, x is expected to be
30
closer to 100kPa. As such, we
1 10 100 1000
Verticalpressure(kPa) 120
can approximately accept the
derived Eur.
log(120) log(30)
gradient _ k 66.9 Eur 2.3 100 66.9 15000 kPa
0.427 0.418
Eur 2.3 y '66.9 Jakis formula:

K 1 sin ' 1 sin 25 .58
NC
0
Poissons ratio ur = 0.2
80
3 = 100kPa for
consolidation,
During shearing, 3 = 0
70
60
Excess pore pressure
Deviatorstress(kPa)
50 accumulates during
40 shearing 3 100kPa
3 = 100 kPa
30
Typically for NC clay, E50ref
20
may be about
Testdata 2~5 times
10 Eoedref or about 2800kPa~7000kPa.
Trial runs to fit the test data gives
0
0 0.01 0.02 0.03 0.04 E50ref =0.06
0.05 3500kPa
0.07
Axialstrain
Can we use the CIU test 1 ~ q test data to

derive the E50ref ??
Singapore 2011
166 of 375
Summary of Hardening Soil Parameters
FEM simulation using Plaxis SoilTest Facility
Singapore 2011
167 of 375
Lets call it a day!
See you tomorrow...
Thank you!
Singapore 2011
168 of 375
Mesh & Geometry Selection
1/35
Contents
Plane strain, Axi-symmetry, 3D
Model boundaries
General considerations
Excavations
Shallow foundations
Embankments
Tunnels
Conclusions
References
169 of 375
Plane strain
Considerations:
One
O dimension
di i iis relatively
l ti l llong
Similar geometry and stress or loading conditions in any cross
section long dimension
Consequences:
No strain long dimension (stress can change!)
No shear stress and arching long dimension
Model represents 1 length unit long dimension
y
Plane strain
Examples:
170 of 375
Plane strain
NOT a plane-strain situation:
45 m
30 m
45 m
8m
Axi-symmetry
Considerations:
Geometry
G t isi circular
i l
Similar geometry and stress or loading conditions in any cross
section that includes the central axis
Consequences:
Stress and strain central axis are radial
Model represents 1 radian around central axis
171 of 375
Axi-symmetry
Examples:
NOT possible with gravity!
Axi-symmetry
NOT an axi-symmetric situation:
Gravity!
172 of 375
3D models
Considerations:
Do I really need a 3D model?
If yes, but still I use a 2D model:
q
What are the consequences?
Would this give conservative or optimistic results?
How large is the error?
Consequences of moving to 3D:

More difficult modelling and interpretation of results
Longer calculation times
Generally less accurate results (due to coarser meshes)
Nevertheless, 3D calculations are quite feasible
3D models
173 of 375
3D models
3D models
174 of 375
3D models
Model boundaries
General considerations
Type of analysis: Deformation, stability, dynamics, flow, .

Type
yp of behaviour: Drained or undrained.
Is the situation (fully) symmetric? Can we model only half the problem?
Boundaries should not influence results.
Changes in stress and strain at boundaries should be low (except for
symmetry boundaries).
What is the consequence of taking boundaries closer or further away?
175 of 375
Model boundaries
Stability analysis:
Mechanism must fit in model

Onlyyp
plastic deformation is relevant
Stress state may not be disturbed by boundaries (arching!)
Model can generally be smaller than for deformation analysis
Model boundaries
Deformation analysis:
Deformations may still occur at a large distance from the action,

especially for undrained analysis (preservation of volume!).
Both elastic and plastic displacements are of influence.
Model should generally be larger than for stability analysis
drained undrained
176 of 375
Model boundaries
Dynamic analysis:
Vibrations may occur at very large distance from the action.

Even veryy small displacements
p ((vibrations)) are of influence.
Even if measures are taken to avoid spurious reflections at
boundaries, it is better to take boundaries far away (considering
wave speed and duration of analysis).
Model should generally be larger than for deformation analysis
~
~ ~ ~ ~
Model boundaries
Stability analysis
Drained
deformation analysis
Undrained
~
~ ~ ~ ~
Dynamic analysis
177 of 375
Model boundaries Shallow foundations

a w a
initial stress after loading

distribution
a
limit
w depth
(0.1 to 0.2) accepted
Suggestions: Stability analysis: a 2w

Deformation analysis: a 3w
Model boundaries Shallow foundations

Take account of the following:
For deformation analysis:

When using g Mohr-Coulomb,, use different layers
y with increasing
g
stiffness; bottom layer with height w should have large small-
strain stiffness.
When using Hardening-Soil,
Hardening Soil, use bottom layer with height w with
large small-strain stiffness for Eurref.
Best results using HSsmall model.
For horizontal loading components: Increased width in loading
direction.
178 of 375
Model boundaries Embankments

a w a
h
Similarity with shallow footings
Suggestions: Stability analysis: a 2w

eo a o a
Deformation analysis:
a ys s a3 3w
Model boundaries Embankments

Embankments are considered to follow similar rules as shallow

foundations with the same base width w
For stability analysis, a can be smaller if mechanism is purely in
embankment itself
179 of 375
Model boundaries Excavations

a w a
l d
Suggestions: Stability or structural analysis: a l and a 2d

Deformation analysis: a 1.5 l and a 3d
U HS
Use HSsmallll or b
bottom
tt llayer with
ith small-strain
ll t i stiffness
tiff f Eurref (height
for (h i ht a))

1. Suggested model depth requires that large small-strain stiffness is

used below the excavation. HSsmall takes care of this.
2. When using Hardening-Soil, use bottom layer with height a with
3. Ignoring small
small-strain
strain stiffness will result in unrealistic heave of
excavation bottom (and wall) and a too wide settlement trough
behind the wall.
4
4. For a < 3d significant settlements may be expected at the upper
model corners. This is even more pronounced for undrained
behaviour.
180 of 375

Considering the wall:
1. Unrealistic heave of excavation bottom gives unrealistic heave of

wall > use large stiffness below excavation
2. For a < 2d vertical model boundaries influence wall displacements
3. Model depth and width seem to have little influence on the wall
forces (bending moments)
Model boundaries Tunnels

w D w w w
TBM or
NATM excavation
w
D D
a a
a
Suggestions: Face stability: a D ; w 2D
Structural analysis: a D ; w 2D
Deformation analysis: a D ; w 3D
Use HSsmall or bottom layer with small-strain stiffness for Eurref (height a)
181 of 375
Model boundaries Tunnels

1. Large unloading and small-strain stiffness below the tunnel

2. Suggested model depth requires that large small-strain stiffness is used
below the tunnel. HSsmall is preferred.
3. When using Hardening-Soil: use bottom layer with height a with a
4. Ignoring small-strain stiffness will result in unrealistic heave of tunnel; a
lower model depth should then be considered (but 2 or 3 is preferred).
5. Ignoring small-strain stiffness will generally result in a too wide
settlement trough above the tunnel, regardless the model width.
6. For w < 3D significant settlements may be expected at the upper model
corners. This is even more pronounced for undrained behaviour.
7. For deep tunnels the overburden may be modelled as load, provided
that at least a height w above the tunnel is included in the model.
Meshing
Type of element:
Two types of volume elements are available in Plaxis 2D:
node (ux, uy)
stress point (, )
x
x
x x
x
y-axis x x
x x x x
x x x x
i
x-axis
6-node triangle 15-node triangle
(quadratic interpolation) (4th order interpolation)
182 of 375
Meshing
Type of element (2D):
Which type of element in which situation?
6-node elements 15-node elements
Plane strain analysis Plane strain

Axi-symmetry
Working
W ki load
l d conditions
diti (SLS) Working
W ki loadl d conditions
diti (SLS)
Failure conditions (ULS)
Phi-c reduction
Updated
U d dM Mesh
h analysis
l i Updated
U d dM Meshh analysis
l i
Note: 15-node elements sometimes fail in Updated Mesh analysis

due to high distortion
Meshing
Type of element 3D: 3
12
1
9
8
6
1
4 7
10
2
15 2
3
14
5
4 6 11
13
5
3DT, 3DF: 15-node wedge New Plaxis 3D: 10-node tetrahedral
((quadratic interpolation)) ((quadratic interpolation))
Do not confuse 15-node wedge in 3D (quadratic) with 15-node triangle in 2D (4th order)!
183 of 375
Meshing
General considerations:
Fine meshes required near stress concentrations or sharp

deformation ggradients ((near structures,, loads,, tunnel faces,, etc.).
)
Coarser meshes may be used towards the model boundaries.
Better to use larger models with relatively large elements (coarse
mesh) near the boundary than to use smaller modelsmodels.
Hint:
Use local element size factors to make meshes fine near loads and
structures and coarse at model boundaries (local element size
factor may be larger than 1.0!).
Meshing
Using local refinement !
184 of 375
Conclusions
Conclusions:
Model size and boundaries depend, a.o., on type of analysis and type of
behaviour (stability analysis, drained deformation undrained deformation,
dynamic analysis).
Small strain stiffness and relatively large models are needed to

Small-strain
accurately predict deformations.
Make use of local refinement or local element size factors to optimise

mesh!
References
Potts D.M., Zdravkovic L. (2001). Finite element analysis in geotechnical engineering
Application. Thomas Telford, London.
Meiner H. (2002). Baugruben Empfehlungen des Arbeitskreises 1.6 Numerik in der

Geotechnik, Abschnitt 3, Geotechnik 25, 44-46.
S h i
Schweiger H.F.
H F (2002).
(2002) Musterlsung
M t l und
dPParameterstudie
t t di f fr dreifach
d if h verankerte
k t B Baugrube,
b
Geotechnik 25, 101-109.
Ruse N.M. (2003). Rumliche Betrachtung der Standsicherheit der Ortsbrust beim
Tunnelvortrieb. PhD thesis. Institut fr Geotechnik. Universitt Stuttgart.
Vermeer P.A., Wehnert M. (2005). Beispiele von FE-Anwendungen Man lernt nie aus. In:
FEM in der Geotechnik (ed. Grabe et.al.). Technische Universitt Hamburg-Harburg.
Brinkgreve R.B.J, Bakker K.J., Bonnier P.G. (2006). The relevance of small-strain stiffness
in numerical simulation of excavation and tunnelling projects. In: NUMGE 2006 (ed.
Schweiger). Taylor & Francis, London. 133-139.
185 of 375
STRUCTURAL ELEMENTS & IN PLAXIS 2D
Presentation by
Dr William Cheang
Principal Geotechnical Consultant
Plaxis AsiaPac Pte Ltd
Some course notes:

Dr Ronald Brinkgreve, Plaxis B.V.
Dr Shen Rui Fu, NUS
Contents
1. StructuralelementsavailableinPlaxis
2. UsageofstructuralelementsinFEmodelling
3. Plate elements(BeamandShellelement)
4. Anchor elements(Springelement)
5
5. Geotextile elements
elements (Membraneelement)
(Membrane element)
6. Interface elements(Zerothicknesselement)
186 of 375
2
1.Structural elements in Plaxis
1. Plateelement Section 3.42 & 14.5
2. Anchorelement Section 3.45, 3.46 & 14.1
3. Geogridselement Section 3.43 & 14.3
4 Interfaceelement
4. Interface element
Section 3.44 & 14.1
2. Application of structural elements
wall strip footing tunnel
geotextile wall ground anchor

g cofferdam
187 of 375
strut anchored wall 4
3.1 Plate Element Plaxis Introductory Course, Danang, Vietnam
Overview:
1. 3or5nodedlineelements(for6nodedor15nodedelementmesh)
2. 3degreesoffreedompernode
3. Plateshave:
o Axialforces
o Shearforces
Shear forces
o Bendingmoments
o Hoopforces(axisymmetry)
4. Elasticorelastoplasticbehaviour
5. Formodellingwalls,floors,tunnels
3.2 Plate Element
Plates elasticparameters
h3 b
EI E (b = 1 m)
12
EA E h b (b = 1 m)
EI (Equivalent rectangular
d h 12
EA plate thickness)
h h
b
b = 1 m in plane strain
b = 1 meter in axisymmetry
188 of 375 b
6
Plates elastoplasticbehaviour
Np
M
Mp
3.4 Plate Element

(Illustration: MpNp
(Illustration:Mp Np.P2D):
P2D):
1200 10090 80 70 60 50 40 30 20 10 0 10 20 30 40 50 60 70 80 90 100

Envelope 0
Elastoplastic plate
1000
Elasticplate
5
800
Elastoplastic plate
10 Elasticplate
600
N
400 15
200 20
189 of 375
0
25 8
200 0 200 400 600 800 1000 1200 M
M
EffectonGlobalFOSbyc/phiReduction
ff l b l b / hi d i
CBPElastic,Failure CBP Elasto-Plastic

withnoPlastic Failure with Plastic
Hinge, Hinge, FOS=1.40
FOS=1.75
1. Elastic wall excludes possibility of wall plastic hinge; and over-estimate FOS=1.75
2. Allowing for wall plastic hinge (Elasto-plastic
(Elasto plastic wall) gave lower FOS
FOS=1.40
1.40 and smaller soil yielded
zone behind the wall
9
3.6 Plate Element
Plates weight,insoil
Actual problem In the model
dreal
wreal = concrete d real wmodel = soil d real wplate Below GT

soil sat
wmodel = wreal w plate = ( concrete - soil ) d real Above GT
Ab
190 of 375
soil 10 unsat
Plates weight,excavation
Actual problem In the model
dreal
1
wreal = concrete d real wmodel = soil d real wplate
2 Below GT
soil sat
1
wmodel = wreal wplate = ( concrete soil ) d real Above GT
2
soil 11 unsat
3.8 Plate Element

Plates connections
Spring data:
Stiffness
6 8 Min/Max moment
Rotation
spring
5 7
Hinged connection
Rigid connection
(default)
191 of 375
12
Illustration: Connection.P2D
Walls thinwallvs.thickwall
Thinwall
h ll
Wallthickness<<walllength
Nomuchendbearing,onlyfriction
Plateelementsuffices
Thickwall
Wallthicknesssignificant
Endbearingcapacityneeded
Usesoilelementswithmaterialsetrepresen ngwallmaterial
Inordertoobtainstructuralforcesaplatewithfictitiousproperties
maybeinserted
b d
13
3.10 Plate Element

Walls thick wall
1. Soilelementswithmaterialsetrepresentingwallmaterial
2 Difficulttoobtainstructuralforcesfromsoilelements,
2. Diffi l b i lf f il l
thereforeintroduceveryflexibleplatewithinthesolidwall
elements:
Noinfluenceondeformation:lowstiffness,noweight
No influence on deformation: low stiffness no weight
Locatedinontheneutralline(usuallythemiddle)
Tightbondingtotheconcreteelements:nointerfaces
(Illustration: Beam.P2D): d
Solid elements: Esoil=Ewall, I = 1/12*d3 , d = wall thickness
Plate element: EI = EsoilI / x, choose x large (e.g. 106)
uplate = usoil Mwall = x*Mplate, Qwall = x*Qplate
192 of 375
14
2.PlaxisAnchor Element
Introductory Course, Danang, Vietnam
Anchors fixedend
a) Tomodelsupports,anchorsandstruts
To model supports, anchors and struts
a) Elastoplasticspringelement
b) Oneendfixedtopointinthegeometry,otherendisfully
fixedfordisplacement
c) Positioningatanyangle
d) Prestressingoption
P i i
Anchors nodetonode
a) To
Tomodelanchors,columns,strutsandrods
model anchors columns struts and rods
a) Elastoplasticspringelement
b) Connectstwogeometrypointsinthegeometry
c)) Nointeractionwiththemeshalongtheanchorrod
h h h l h h d
d) Prestressingoption
15
4.1Anchor Element
Anchors materialproperties
Axialstiffness,EA
Axial stiffness EA (foroneanchor)
(for one anchor) [kN]
Spacing,Ls (outofplanedistancebetweenanchors) [m]
Maximumanchorforceforcompressionandtension,
|Fmax,comp|and|Fmax,tens| [kN]
Ls
193 of 375
16
4.2 Anchor Element
Anchors prestressing
DefinedinStagedconstructionphase
Bothtension(groutanchor)orcompression(strut)
possible
Tension = positive
17
5.1.Geogrid Element
Geogrids
1.
1 3or5nodedlineelement
3 5 d d li l t
2. Elasticorelastoplasticbehaviour
3. Noflexuralrigidity(EI),onlyaxialstiffness(EA)
4. Onlyallowsfortension,notforcompression
194 of 375
18
+ Geogrid Element
5.2 Anchor Element
Groundanchors
1. Combinationofnodetonodeanchorandgeogrid
2. Nodetonodeanchorrepresentsanchorrod(freelength)
(nointeractionwithsurroundingsoil)
3
3. G
Geogridrepresentsgroutedpart(fullinteractionwithsurroundingsoil)
id t t d t (f ll i t ti ith di il)
4. Nointerfacearoundgroutedpart;interfacewouldcreateunrealisticfailuresurface
5
5. Working load conditions onl nopullout
Workingloadconditionsonly no p llo t
6. Ifpulloutforceisknownthiscanbeusedbylimitinganchorrodforce
19
5.3Groundanchors
Axial force distribution along fixed length (modelled using geogrid)
Nrod <>Ngrout duetosharednode

betweenanchor,geotextile
andsoil
Probableactualdistributionofaxial
forcesingroundanchor
axialforcesingeotextileelement
195 of 375
Input geometry Generated mesh
20
5.4GroundAnchors:Influenceofnodenumbers
along structural elements

alongstructuralelements
21
6.1InterfaceElement
Interfaces materialproperties
1. Soilstructureinteraction
1. Wallfriction
2. Slipandgappingbetweensoilandstructure
2. Soilmaterialproperties
Soil material properties
A. TakenfromsoilusingreductionfactorRinter
3. Individualmaterialsetforinterfacepossible
196 of 375
22
6.2 Interface Element
Interfaces reductionfactor
SuggestionsforR
gg inter:
Interactionsand/steel =Rinter 0.6 0.7
Interactionclay/steel =Rinter 0.5
I t
Interactionsand/concrete
ti d/ t =R
Rinter 1.0
1 0 0.8
08
Interactionclay/concrete =Rinter 1.0 0.7
Interactionsoil/geogrid(groutedbody) =Rinter1.0
(interfacemaynotberequired)
Interactionsoil/geotextile =Rinter0.9 0.5(foil,textile)
With reference to BS8002:
23
References
1 Brinkgreve
1. Brinkgreve, R.,
R Engin,
Engin E,
E & Swolf,
Swolf WW. (2010)
(2010), Plaxis 2d 2010
197 of 375
24
25
198 of 375
Tied-back excavation using HSsmall
TIED-BACK EXCAVATION
Using the HSsmall model
199 of 375
200 of 375
INTRODUCTION
A building pit was constructed in the south of the Netherlands. The pit is 15 m deep and 30 m
wide. A diaphragm wall is constructed using 100 cm diameter bored piles; the wall is anchored
by two rows of pre-stressed ground anchors. In this exercise the construction of this building
pit is simulated and the deformation and bending moments of the wall are evaluated.
The upper 40 m of the subsoil consists of a more or less homogeneous layer of medium dense
fine sand with a unit weight of 18 kN/m3 . Triaxial test data of a representative soil sample is
given in figure 2. Underneath this layer there is very stiff layer of gravel, which is not to be
included in the model. The groundwater table is very deep and does not play a role in this
analysis.
AIMS
Using interface elements
Using ground anchors
Pre-stressing of anchors
Combination of structural elements
0 x 4 1
Stage 1
Secant wall
11 12
Stage 2
13 14 Anchor rods
Stage 3
7 8 15
Grout bodies
16 17
9 5 18 10
6
3 2
Figure 1: Geometry for tied-back excavation
201 of 375
202 of 375
MATERIAL PARAMETERS
Determination of stiffness & strength properties (sand)
In this exercise the HSsmall model is used and the model parameters for the sand layer
have been extracted from the triaxial test data (see figure 2). The HSsmall model takes into
account the stress-dependency of soil stiffness, elasto-plastic behaviour under both compres-
sion loading and shear loading and increased stiffness in areas with very low strain levels.
The soil parameters can be found in table 1, while the determination of the soil parameters
can be found in appendix A.
Figure 2: Triaxial test data for the sand layer
Secant wall
The secant wall consists of 100cm diameter bored piles with an intermediate distance of 80cm,
hence there is a 20cm overlap of the piles. This configuration is taken this into account for the
determination of the cross sectional area (A) and moment of inertia (I) per meter out-of-plane
(see Appendix B). The concrete stiffness is Ec =2.7107 kN/m2 with a specific weight =16
kN/m3, which leads to the material parameters as given in Table 2. The determination of the
stiffness parameters can be found in Appendix A.
203 of 375
Table 1: Soil parameters for the HSsmall model

Parameter Symbol Sand(Rinter =0.6) Sand(Rinter =1.0) Unit
Material model Model HSsmall HSsmall
Type of behaviour Type Drained Drained
Unsaturated weight unsat 18.0 18.0 kN/m3
Saturated weight sat 18.0 18.0 kN/m3
ref 4
Drained triaxial test stiffness E50 2.010 2.0104 kN/m2
Drained primary oedometer Eref
oed 2.0104 2.0104 kN/m2
stiffness
Unloading/reloading stiffness Eref
ur 8.0104 8.0104 kN/m2
Power for stress-dependent m 0.5 0.5
stiffness
Cohesion c 1.0 1.0 kN/m2
Friction angle 0 35 35
Dilatancy angle 5 5
ref 4
Small-strain shear modulus G0 10.010 10.0104 kN/m2
Threshold shear strain 0.7 1.5104 1.5104
0
Unloading/reloading Poissons ur default default
ratio
Reference stress pref default default kN/m2
NC
Coefficient for lateral stress K0 default default
under primary loading
Interface strength reduction Rinter 0.6 rigid
Coefficient for lateral initial K0 automatic automatic
stress
Ground anchors
The anchors are made of 32mm diameter steel bars at an intermediate distance of 1m. The
steel bars have a stiffness of Es =2.1*108 kN/m2 . The anchors have an ultimate strength
of 605 kN per anchor. In combination with a secant wall the anchors may be prestressed
to a maximum level of 60% of the ultimate strength, hence up to 363 kN per anchor. The
maximum compression force of the anchor is not important as the anchors will not be loaded
under compression. The grout body that forms the bonded length of the anchor behaves
relatively weak under tension compared to the steel bar inside. Therefore it is assumed that
both stiffness and strength of the bonded part of the anchor are fully determined by the steel
bar. This leads to the material properties for both the anchor rod (free length) and grout body
(bonded length) as given in tables 3 and 4.
204 of 375
Table 2: Properties of the secant wall (plate)

Parameter Symbol Secant wall Unit
Material behaviour Material type Elastic
7
Axial stiffness EA 2*10 kN/m
6
Flexural stiffness EI 1.67*10 kN/m2 /m
Weight w 15.0 kN/m/m
Poissons ratio 0.15
Table 3: Properties of the anchor rods (node-to-node anchors)

Parameter Symbol Anchor rod Unit
Material behaviour Material type Elastoplastic
Axial stiffness EA 1.7*105 kN
Spacing Lspacing 1.0 m
Max. tension force |Fmax,tens | 605 kN
Max. compression force |Fmax,comp | 605 kN
Table 4: Properties of the grout bodies (geotextiles)

Parameter Symbol Grout body Unit
Material behaviour Material type Elastoplastic
5
Axial stiffness EA 1.7*10 kN/m
Max. tension force Np 605 kN/m
205 of 375
206 of 375
GEOMETRY INPUT
Start a new project
Project properties
Accept the default values in the Project tab sheet of the Project properties (15-node
elements). For the dimensions see figure 3.
Figure 3: Project propeties, tabsheet Model
Geometry
(15,0)
(0,0) (70,0)
0 x 4 1
(0,-5) 11 12
(0,-10) 13 14
(30,-15)
(0,-15) 7 8 15
(37.5,-20)
(30,-20) 16 17
(0,-25) (70,-25)
9 5 18 10
6 (37.5,-25)
(15,-27)
(0,-60) (70,-60)
3 2
Figure 4: Geometry of the model
Click the Geometry line button and draw the geometry contour and soil layers as
specified in figure 4.
207 of 375
Click the Plate button and draw the secant wall from (15, 0) to (15, -25).
Click the Interface button and draw the interface from (15,0) to (15, -27) and back to
(15,0). This creates an interface on both sides of the secant wall.
Click the Geotextile button and insert both grout bodies.
Click the Node-to-node anchor button and insert both anchor rods. These anchors
connect the beginning of the grout bodies to the wall.
Finally, click the Geometry line button again to introduce the two levels of excavation.
Hints: As interfaces can be introduced on both sides of a geometry line, one

should pay attention to the arrows on the cursor. These arrows indicate
where the program will locate the interfaces.
> Please note that the interface is extended for a short distance underneath
the beam. This is done to overcome a singular point at the bottom of the
wall.
Hint: It is not necessary to create a geometry line before creating plates,
geogrids or anchors. When drawing a plate or geogrid, a geometry line is
automatically added. Anchors do not create corresponding geometry
lines. This is not necessary since anchors do not interact with the
underlying soil.
Fixities
Click the Standard fixities button to apply standard boundary conditions.
Material properties
Enter the material properties for the four soil data sets, as determined in table 1of this
exercise.
After entering all properties for the three soil types, drag and drop the properties to the
appropriate clusters.
Enter material properties for the plates, anchors and geogrids as indicated in tables 2,
3and 4.
208 of 375
Mesh generation
From the Mesh menu, set the Global coarseness to Medium and press the Generate
button. This will result in a mesh as shown in figure 5.
Figure 5: Medium finite element mesh
Select the geogrid and plate elements and press Refine line from the Mesh menu. This
will result in a refinement around the selected lines as shown in Figure 6.
Figure 6: Refined finite element mesh
209 of 375
210 of 375
CALCULATION
When starting the calculation program choose the Classical mode.
The entire construction process consists of five phases. Define the phases, as shown graph-
ically below. For each phase, use the Plastic calculation, Staged construction.
Initial phase
For the initial phase choose the K0 procedure for calculating the initial stresses. As the phreatic
line is located below the geometry the generation of initial pore pressures can be skipped and
since its not necessary to switch off any soil for the initial situation it is not needed to define
the initial phase.
Phase 1
In the first phase, the diaphragm wall is activated and the first excavation takes place.
Note that though the the interfaces along the wall are activated automatically with the
activation of the wall, the extensions below the diaphragm wall have to be activated
manually.
Figure 7: Phase 1: activation of the wall Figure 8: Phase 2: activation and

and 1st excavation presstressing of the 1st anchor
Phase 2
In the second phase, a new option is used, namely the prestressing of anchors.
First the grout-body (the geogrid) is switched on by clicking on the geogrid element.
The element will appear in yellow as soon as it is switched on. The light grey colour
indicates non-active elements.
211 of 375
Now that the grout-body is active, the anchor element needs to be prestressed. By
double clicking on a node-to-node anchor a window will appear as shown in figure 9.
Select the option Adjust prestress, fill in a prestress force of 300 kN/m (tension) and
press OK.
In the geometry a black node-to-node anchor indicates that the anchor is activated. The
letter P indicates that a prestress force will be active in the anchor.
Figure 9: Node-to-node anchor properties
Phase 3, 4 and 5
Now define the remaining phases according to figures 10, 11 and 12.
In phase 3 excavate the second part of the excavation
In phase 4 activate the lower anchor and prestress it to 300 kN/m
In phase 5 excavate the remaining 3rd part.
Hint: When processing an anchor in a certain calculation phase the anchor

force will exactly match the prestress force at the end of that phase. In
following calculation phases without prestressing, the anchor force will be
influenced by the excavation process
212 of 375
Figure 10: Phase 3: Second excavation Figure 11: Phase 4: Activation and
prestressing of 2nd anchor
Figure 12: Phase 5: Final excavation
213 of 375
214 of 375
INSPECT OUTPUT
The results of fase 5 is presented in Figure 13. After this final stage the excavation bottom
heave calculated is about 5 cm.
Figure 13: Deformed mesh (phase 5)
By double clicking on the node-to-node anchors, Plaxis will present a table, in which the
stress in all anchors may be inspected. Anchor forces are approximately 340 kN where
the lower anchor has a slightly higher anchor force than the upper anchor.
When double-clicking on one of the geogrids the change of axial forces within the grout body
can be investigated. What is immediately noticeable is that the axial force at the connection
with the anchor rod is significantly lower than the force in the anchor rod itself. This is due the
fact that the end of the anchor rod is not only connected to the grout body, but also to several
soil elements surrounding the end of the anchor rod. Therefore part of the anchor force is
transferred directly to those soil elements while part of the anchor force is transferred to the
geotextile representing the grout body. The amount of force transferred to the soil depends on
the stiffness of the soil; in this exercise it is 25-35% of the anchor force. However, this effect
has very little influence on other calculation results. That is, it is not so important for other
calculation results how the anchor rod transfers its force; directly to the soil or by means of the
grout body.
By double-clicking on the wall the structural forces in the wall can be inspected. The
maximum bending moment should be in the order of 350 kNm/m (figure 14)
When double-clicking on an interface only the results of part of the interface can be seen.
In order to see the results for the whole interface chain, keep Ctrl + Shift pressed on the
keyboard while double-clicking on the interface. In figure 15the left side are the passive
earth pressures and the right side are the active earth pressures. It can be seen that
only a small part of the maximum passive earth pressures has been mobilized at this
stage.
215 of 375
Figure 14: Bending moments in the sec- Figure 15: Effective normal stresses in the
ant wall interface
Geometry size
For any project the geometry has to be made sufficiently large so that the boudary conditions
have no influence on the calculation results. This means in practice that close to the boundar-
ies (with exception of a axis of symmetry) displacements should be small and stresses should
be undisturbed. When using the HSsmall model there is an interesting plot that can be used
to check this.
From the Stresses menu choose the option State parameters and then G/Gur .
This plot shows the actual shear stiffness divided by the unloading/reloading shear stiffnes
at engineering strain level. For areas with very small deformations the stiffness will be high
(small strain stiffness) and so the value of G/Gur > 1. Hence, the geometry is sufficiently large
if next to the boundaries, with exception of the axis of symmetry, G/Gur > 1, which indeed is
the case.
Hint: State parameters are additional quantities that relate to the state of the
material in the current calculation step, taking into account the stress
history. Examples of state parameters are the isotropic overconsolidation
pressure (pp ) and the hardening parameter p that specifies the maximum
shear strain level reach in the stress history.
216 of 375
Surface settlements
In Plaxis Output it is possible to see calculation results in a user-defined cross section. This
feature will be used to check the surface settlements behind the secant wall.
Click the Cross section button . The Cross section points window appears, see figure
16.
It is possible to draw a cross section by hand and check in the Cross section points window
what the coordinates are of the start and end point of the cross section. However, it is also
possible to position the cross section at a specific location by defining the coordinates of the
start and end point manually.
Move the mouse to the Cross section points window and fill in the coordinates (15, -0.1)
for the first point and (70, -0.1) for the second point and press OK. This will create a
cross section from the secant wall until the right boundary of the model just below the
soil surface. The cross section will open in a new window.
From the Deformations menu select Total displacements and then u y to see the vertical
displacements of the soil surface. The maximum settlement is 12-13 mm, see figure 17.
Figure 16: Cross section points window
Figure 17: Vertical displacements behind the secant wall
217 of 375
218 of 375
APPENDIX A: DETERMINATION OF SOIL PARAMETERS FROM

TRIAXIAL TEST
Figure 18: Triaxial test for sand layer
Strength parameters
Fill in 1 and 3 in the Mohr-Coulomb criteria:
1 3 = (1 + 3 ) sin + 2c cos
Since the cohesion will be small, assume c = 0:
1 3
1 +3
= sin
370100
370+100
= sin
= 35o
= 30 = 5o
For reasons of numerical stability, use c = 1 kPa
219 of 375
Stiffness parameters
Since excavation is considered in this exercise, the input of Youngs modulus E should be
based on unloading, rather than on primary loading. For the same reason, Poissons ratio
should also be based on unloading, which results in a somewhat lower value.
The triaxial test has a cell pressure 3 = 100 kPa. This corresponds with reference pressure,
so E50 = Eref
50 .
ref v 135
E50 = v
= 0.675%
= 2.0 104 kP a
For Sand it can be assumed that

ref ref
Eoed = E50 = 2.0 104 kP a
ref ref
Eur 4 E50 = 8.0 104 kP a
m = 0.5
Additionally it is assumed that:
Gref
0
ref
= 1.25 Eur = 1 105 kP a
0.7 = 1.5 104
220 of 375
APPENDIX B: MATERIAL PROPERTIES SECANT WALL

For a plane strain model material properties for the secant wall have to specified per meter
length of the wall. In order to do so we first recognize the secant wall as consisting of repetitive
parts at a certain intermediate distance, as shown in Figure 19.
Figure 19: Secant wall as repetitive equal sections
Compared to the original bored piles the repetitive sections have a reduced cross sectional
area. Though it can be analytically derived how much the reduction is, the fastest way to
determine this is to draw the repetitive section on paper with a fine grid based on the original
bored piles with a diameter of 1000mm and an overlap of 200mm and count squares. Using
this method the cross sectional area of the repetitive section is determined as As = 0.74 m2 .
Since the sections are at a distance D apart where D is given as 800mm, the cross sectional
area of the wall per meter is given as:
Awall = ADs = 0.74
0.8
= 0.93 m2 /m
For the moment of inertia is assumed that the influence of the reduced cross sectional area
is negligble as the reduction is close to the axis of bending and symmetric. Therefore the
moment of inertia per meter wall is determined as:
Ipile 4 (0.5)4
Iwall = D
= r
4D
= 40.8
= 61.3 103 m4 /m
With Econcrete = 2.7 107 kN/m2 this gives
EA = (2.7107 )(0.93) = 2.5107 kN/m

EI = (2.7107 )(61.3103 ) = 1.67106 kNm2 /m
And for the weight:
w = A = 16 0.93 = 15 kN/m/m
221 of 375
Undrained Soil Behaviour
Some of the used material was originally created by:

Prof. Helmut Schweiger, Technical University of Graz, Austria
1
Contents
Drained / undrained (conditions and analysis)
Drained / undrained soil behaviour
Typical results from drained and undrained triaxial tests
Strength parameters
What is the critical case: drained or undrained?
Modeling undrained behaviour with Plaxis
Three methods
Effective stress analysis:
y how does it actually
y work
Undrained shear strength
Undrained behaviour with Mohr-Coulomb Model
Undrained behaviour with Hardening Soil Model
Influence of dilatancy
Summary
2
222 of 375

In drained analysis water is assumed to immediately flow out
upon loading
l di and
d therefore
th f no pore pressures are built
b ilt up.
This is appropriate when
Permeability is high
Rate of loading is low
Short term behaviour is not of interest for problem considered
In undrained analysis no water movement takes place and

therefore pore pressures are built up.
This is appropriate when
Permeability is low and rate of loading is high
Short term behaviour has to be assessed

Suggestion by Vermeer & Meier (1998) for deep excavations:
T < 0.10 (U < 10%) use undrained conditions
T > 0.40 (U > 70%) use drained conditions
k = Permeability
k E oed Eoed = Oedometer modulus
T 2
t w = Unit weight of water
w D
D = Drainage length
t = Construction time
T = Dimensionless time factor
U = Degree of consolidation
4
223 of 375
Contents
Strength parameters
Three methods
y work
Summary
Triaxial test (NC) drained / undrained

Typical results from drained (left) and undrained (right) triaxial tests on normally
consolidated soils (from Atkinson & Bransby
Bransby, 1978)
6
224 of 375
Triaxial test (OC) drained / undrained

Typical results from drained (left) and undrained (right) triaxial tests
on overconsolidated soils
Triaxial test stress paths (NC/OC)
8
225 of 375
Strength parameters
Mohr-Coulomb parameters in terms of effective stress (real soil behaviour)

c tan

c
3 ' s 1 '
1 3 3 c c
1 sin ; t s sin
2 2 tan tan 9
Strength parameters
MC parameters in terms of total stresses (only undrained conditions!)
c tan total stresses
Cu
1 3 1 3
2
F
2
F cu ,
-Cu
Effective
ff stresses
Soil behaves as if it was purely cohesive (zero friction)

Cu : undrained shear strength
Cu only changes if drainage occurs (no change if undrained conditions prevail)
10
226 of 375
t
1.
short
short long
2.
ESP
long
s, s
Note that for soils in general:

1. factor of safety against failure is lower for short term (undrained)
conditions for loading problems (e.g. embankment)
2. factor of safety against failure is lower for long term (drained)
conditions for unloading problems (e.g. excavations) 11

t
2.
1
1.
ESP
s, s
For very soft NC soil, factor of safety against failure may be lower for
short term (undrained) conditions for unloading problems (e.g.
excavations)
For very stiff OC soil, factor of safety against
g failure may be lower for
short term (undrained) conditions for loading problems (e.g.
embankment) 12
227 of 375
Contents
Strength parameters
Three methods
y work
Summary
13
Undrained behaviour with PLAXIS

Method A (analysis in terms of effective stresses):
t
type off material
t i l behaviour:
b h i undrained
d i d
effective strength parameters c, ,
effective stiffness parameters E50,
Method B (analysis in terms of effective stresses):

type of material behaviour: undrained
d i d strength
undrained t th parameters
t c = cu, = 0,
0 =0
effective stiffness parameters E50,
Method C (analysis in terms of total stresses):

type of material behaviour: drained
total strength parameters c = cu, = 0, = 0
undrained stiffness parameters Eu, u = 0.495
14
228 of 375

PLAXIS automatically adds stiffness of water when undrained material
t
type is
i chosen
h using
i ththe ffollowing
ll i approximation:
i ti
Kw Eu 2 G 1 u
K total K'
n 31 2 u 31 2 u
E ' 1 u 3 ' B (1 2 ')
K total u
3 1 2 u 1 ' 3 B (1 2 ')
Notes:
Skempton B-value can be entered explicitely for undrained materials in order to
simulate effect of partially saturated soil on the effective and excess pore
pressures.
This procedure gives reasonable relation between u and B only for f < 0.35 !
Real value of Kw/n ~ 1.10 kPa (for n = 0.5)
6
15
FE modeling of undrained behaviour (method A)
uf
u
TSP
ESP
cu
s s
s,
single
g set of parameters
p in terms of effective stress (consistent)
( )
realistic prediction of pore pressures (if model is appropriate)
the undrained analysis can be followed by a consolidation analysis
Cu is a consequence of the model, not an input parameter!!
16
229 of 375
FE modeling of undrained behaviour (method C)
TSP=ESP
cu
s s
s, s
pa
parameters
a ete s in te
terms
soof tota
total st
stress
ess
no prediction of pore pressures (only total stresses are obtained)
the undrained analysis can not be followed by a consolidation analysis
Cu is an input parameter!!
17
FE modeling of undrained behaviour (method B)
ESP TSP
cu
s s
s, s
parameters
p in terms of total stress and effective stress
prediction of pore pressures (generally unrealistic)
the undrained analysis should not be followed by a consolidation
analysis (pore pressures unrealistic)
Cu is an input parameter!! 18
230 of 375

Notes on different methods:
Method A:
Recommended, but be careful with MC model
Soil behaviour is always governed by effective stresses
Increase of shear strength during consolidation included
Essential for exploiting features of advanced models such as the
Hardening Soil model, the Soft Soil model and the Soft Soil Creep model
Method B:
Only when no information on effective strength parameters is available
May be a safer choice than Method A when using MC-model
Cannot be used with the Soft Soil model and the Soft Soil Creep model
Method C:
NOT recommended
No information on excess pore pressure distribution (total stress
analysis)
19
Contents
Skempton's parameters A and B
Strength
g parameters
p
Three methods
Effective stress analysis: how does it actually work
Undrained behaviour with Mohr
Mohr-Coulomb
Coulomb Model
Summary
20
231 of 375
Undrained behaviour of Mohr-Coulomb

t
cu,MC tan()
cu,real
Mohr-Coulomb
Real soil
1
cu c ' cos ' s 'sin ' c ' cos ' vo ho sin '
2
1
cu c ' cos ' 'v 0 1 K 0 sin
i '
2 21
Undrained behaviour of Hardening Soil

t
tan()
cu,HS {
Hardening Soil
cu is a result of the analysis depending on c, , Eur/Eoed and other parameters

Its important to simulate triaxial tests and compare them with real soil tests
Not all cu values can be achieved with a particular model
22
232 of 375
Influence of constitutive model

Parameter sets for Hardening Soil model
ref ref ref ref nc
Model Number E50 Eur Eoed c ur p m K0 Rf
2 2 2 2 2
kN/m kN/m kN/m kN/m - kN/m - - -
HS_1 30 000 90 000 30 000 35 0 / 10 0.0 0.2 100 0.75 0.426 0.9
HS_2 50 000 150 000 50 000 35 0 0.0 0.2 100 0.75 0.426 0.9
HS_3
_ 15 000 45 000 15 000 35 0 0.0 0.2 100 0.75 0.426 0.9
HS_4 30 000 90 000 40 000 35 0 0.0 0.2 100 0.75 0.426 0.9
HS_5 30 000 90 000 15 000 35 0 0.0 0.2 100 0.75 0.426 0.9
HS_6 50 000 150 000 30 000 35 0 0.0 0.2 100 0.75 0.426 0.9
P
Parameters
t for
f MC M
Model
d l
see also Schweiger (2002)
2
E = 30 000 kN/m , = 0.2, = 35, = 0 and 10 23
Parameter variation Hardening Soil

Simulation of undrained triaxial compression test HS model - q vs p
150
HS_1
HS_2
125 HS_3
HS_4
HS 5
HS_5
100 HS_6
total stress path
q [kN/m ]
2
75
50
25
0
0.00 25.00 50.00 75.00 100.00 125.00 150.00
2
24
p' [kN/m ]
233 of 375

Simulation of undrained triaxial compression test HS model - q vs 1
150
125
100
N/m ]
2
75
q [kN
50 HS_1
HS_2
HS_3
25 HS_4
HS_5
HS_6
0
0.00 1.00 2.00 3.00 4.00 5.00 6.00
25
1 [%]

Simulation of undrained triaxial compression test HS model - pw vs 1
80
70
excess pore pressure [kN//m ]
2
60
50
40
30
HS_1
HS_2
20
HS_3
HS_4
10 HS_5
HS_6
0
0.00 1.00 2.00 3.00 4.00 5.00 6.00
26
1 [%]
234 of 375
if we set 0 then, negative volumetric plastic deformations
occur at failure:
v ve vp (elastic-plastic
(elastic plastic behavior)
v 0 (undrained conditions)
vp 0 ve 0 p ' K ve 0
At failure: q M p ' q 0
result: unlimited increase of q, i.e. infinite strength!!

g
27
Simulation of undrained triaxial compression test MC / HS model - q vs 1
300
275
250
225
200
175
N/m ]
2
150
q [kN
125
100
75
MC non dil
50 MC dil
HS_1 non dil
25 HS_1 dil
0
0.00 0.25 0.50 0.75 1.00 1.25 1.50 1.75 2.00 2.25 2.50 2.75 3.00
28
1 [%]
235 of 375
Simulation of undrained triaxial compression test MC / HS model - q vs p
300
275 MC non dil

MC dil
250 HS_1 non dil
HS_1 dil
225 total stress p
path
200
175
q [kN/m ]
2
150
125
100
75
50
25
0
0.00 25.00 50.00 75.00 100.00 125.00 150.00 175.00 200.00 225.00 250.00
29
p' [kN/m2]
Simulation of undrained triaxial compression test MC / HS model - pw vs 1
100
90 MC non dil
MC dil
80
HS_1 non dil
excess pore pressure [kN//m ]
2
70 HS_1 dil
60
50
40
30
20
10
-10
-20
20
0.00 0.25 0.50 0.75 1.00 1.25 1.50 1.75 2.00 2.25 2.50 2.75 3.00
30
1 [%]
236 of 375
Summary
Undrained analysis should be performed in effective stresses and
with
ith effective
ff ti stiffness
tiff and
d strength
t th parameters
t (Method
(M th d A)
Undrained shear strength is result of the constitutive model
The Mohr-Coulomb model generally overestimates the undrained
shear strength in a Method A calculation. This can be solved by
doing a Method B analysis, but this is a trick that generally
generates incorrect excess pore pressures
One should not use dilatancy in an undrained analysis
31
References
Atkinson, J.H., Bransby, P.L. (1978)
The Mechanics of Soils,
Soils An Introduction to Critical State Soil Mechanics
Mechanics. McGraw Hill
Ortigao, J.A.R. (1995)
Soil Mechanics in the Light of Critical State Theories An Introduction. Balkema
Schweiger, H.F. (2002)
Some remarksk on pore pressure parameters A andd B ini undrained
d i d analyses
l with
i h the
h Hardening
d i Soil il
Model. Plaxis Bulletin No.12
Skempton, A.W. (1954)
The Pore-Pressure Coefficients A and B. Geotechnique, 4, 143-147
Vermeer, P.A., Meier, C.-P. (1998)
Proceedings Int. Conf. on Soil-Structure Interaction in Urban Civil Engineering, Darmstadt, 177-191
32
237 of 375
Modelling of Groundwater in PLAXIS
MODELLING OF GROUNDWATER IN PLAXIS

Dr William Cheang, Plaxis AsiaPac
CONTENTS
A. Introduction
1. Groundwater in Geotechnical Engineering
2
2. Plaxis
B. Definitions Porewater Pressures in Plaxis
1. Active
2
2. Steady-state
3. Excess
C. Generation of Porewater Pressures in Plaxis
1.
1 Porewater Pressure due to Hydrostatic Condition
2. Pore Pressures due to Groundwater Flow (Steady or Transient States)
D. Hydraulic models
1.
1 Fully Saturated Soils
2. Partially Saturated Soils
E. Case Histories
1. Excavations
2. Embankments and Dams
3. Slopes
F. References
Plaxis Introductory Course, July 10-12,

2
Danang.
238 of 375
Modelling of Groundwater in Plaxis

A. INTRODUCTION

3
Danang.
Groundwater Analysis
A.
A Geotechnical problems are related to groundwater
B. Two extreme conditions of porewater response are normally considered,
they are:
1.
1 Drained
D i d
2. Undrained (Method A, B & C)
C. Real soil behaviour is related to time , i.e. transient, with the porewater
pressure being dependent on imposed:
1. Permeability
2. Rate of loading
3. Hydraulic boundary
D. The interstitial voids of the soil skeleton can be fully or partially filled with
pore fluid and therefore effective stresses are influenced by this action
E. This lecture will look into the following issues:
1. The setup of pore pressures in Plaxis
2. Input parameters
3. Some examples of groundwater regimes

4
Danang.
239 of 375

B. DEFINITIONS OF POREWATER
PRESSURES IN PLAXIS

5
Danang.
A. DEFINITIONS AND MODES
A. Definition of Porewater Pressure Terms in Plaxis
1. Active State Porewater Pressures
2. y State Porewater Pressures (Background)

Steady ( g )
3. Excess Porewater Pressures
B. Calculation Modes (since 2010)
1. Classical (same with previous versions prior 2010)
2. Advanced
3
3. Fl
Flow

6
Danang.
240 of 375
A1.ACTIVE POREWATER PRESSURE
A. In Classical model there three porewater pressure terms
B
B. Active (Total) = Steady
Steady-state
state + Excess
1. ACTIVE porewater pressures is combination of STEADY-STATE and EXCESS porewater
pressures (see Reference Manual 5.9).
2. Steady-state pre pressures are generated due to water conditions (hydraulic boundaries)
assigned to soil clusters (layers)
3. Excess pore pressures are calculated as a result of undrained or consolidation analysis

7
Danang.
A2.STEADY-STATE POREWATER PRESSURE
A. Steady-state porewater pressures (Background PWP)

1. Represent a stable groundwater condition that remain constant over time
2
2. Situation 1 : Standing,
Standing stable or static water
water-table
table
3. Situation 2 : Permanent Groundwater flow or seepage in stable state
B. Steady-state pore pressures in situation 1(see SECTION C):
1.
1 General Phreatic line
2. Local Phreatic line
3. Cluster
4. Interpolated between clusters
C. Steady-state pore pressures in situation 2 (see SECTION C):
1. Groundwater Flow :Steady-State Seepage
2. Groundwater Flow: Transient Seepage taking into account of changing hydraulic
boundary. Porewater pressures calculated from Transient Seepage is taken as a
Steady-state.

8
Danang.
241 of 375
A3.EXCESS POREWATER PRESSURE
A. Excess pore pressures are generated during calculation:

1. Plastic Calculation (Elastoplastic time-independent analysis)
2
2. Consolidation Calculation (Elastoplastic time
time-dependent
dependent analysis)
B. Plastic Calculation (Reference Manual 5.5.2)
1. Classical mode
2
2. Excess porewater pressure generated due to undrained drainage type (Undrained A or B)
3. Constitutive soil model dependent
4. Time independent
C
C. Consolidation (EPP) Calculation (Reference Manual 5.5.4)
5 5 4)
1. Classical mode
2. Time dependent
3. Can increase or decrease with time
4. Constitutive soil model dependent
5. Influence permeability

9
Danang.
B. CALCULATION MODES

10
Danang.
242 of 375
B1.CLASSICAL MODE
A. Steady-state
Steady state pore pressures
1. Phreatic lines
2. Steady-state groundwater flow analysis
3. Transient-state g
groundwater flow analysis
y =
Steady-state background pore pressure
B. Excess pore pressures
1. Undrained material type in combination with
Pl ti calculation
Plastic l l ti
2. Consolidation analysis
INPUT KERNEL
St d State
Steady St t E
Excess Porewater
P t D f
Deformation
ti
Note:
Note:
1. Undrained Analysis
1 Hydrostatic
1.
2. Consolidation Analysis
2. GWF calculation
3. Excess (Soil Model)
Active Porewater = Steady-state + Excess

11
Danang.
B2.ADVANCED MODE
A. Consolidation analysis
B
B. Transient groundwater flow analysis
KERNEL
Active Porewater Deformation

N t
Note:
1. Undrained Analysis
2. Consolidation Analysis
3. Excess (Soil Model)

12
Danang.
243 of 375
B3.FLOW MODE
Flow mode:
Similar to PlaxFlow but with huge improvements in the
kernel (see Galavi, 2010)
All functionalities of PlaxFlow rewritten in PLAXIS code
(
(new) )
Steady state groundwater flow
Transient groundwater flow
All types of boundary conditions
New features in wells and drains
Faster calculation (new)

13
Danang.

B. GENERATION OF POREWATER
PRESSURES

14
Danang.
244 of 375
GENERATION OF STEADY-STATE PORE PRESSURES
Steady-state (Background) pore pressures can be generated by:
1. Phreatic and Cluster Approach (Hydrostatic)
2. Groundwater Flow Analysis

15
Danang.
PHREATIC AND CLUSTER APPROACH
A. Phreatic Level (Reference 5.9.2)

1. General Phreatic Level
2. Cluster Phreatic Level

B. Cluster Pore Pressure Distribution (Reference 5.9.5)
1. Interpolation (Adjacent to clusters or phreatic lines)
2. Cluster Dry
3. User-defined Pore Pressure Distribution

16
Danang.
245 of 375
C1. PHREATIC LINE
a) Porewater pressures are hydrostatic

b) Calculated based on gamma-water
gamma water * height of the water column
c) Simple situations (water-table is horizontal)
d) No flow
e) For cases, i.e. simple excavations, foundations or embankments

Hydrostatic steady-state pressure along interface
17
Danang.
GENERAL PHREATIC LINE

18
Danang.
246 of 375
GENERAL & CLUSTER PHREATIC LINES

19
Danang.
Cluster: Dry

20
Danang.
247 of 375
Cluster: User-defined Pore Pressure Distribution

21
Danang.
Combination: Phreatic and Cluster Options

22
Danang.
248 of 375
COMBINATION: PHREATIC & CLUSTER OPTIONS

23
Danang.
CLUSTER DRY + INTERPOLATION COMBO
Cl ster Dry
Cluster Dr
Interpolated
Case Histories

24
Danang.
249 of 375
GROUNDWATER FLOW ANALYSIS

A. Steady-state Pore Pressure Generation based on Groundwater Flow
Calculation
B
B. Influence by:
1. Soil Permeability
2. Boundary Conditions (External or Internal)
C
C. Phreatic line is calculated for
1. Confined flow problems
2. Unconfined flow problems
D
D. St d t t groundwater
Steady-state d t flow:
fl
1. No change in flow field with time
2. Position of phreatic is fixed (influence by k and geometry of hydraulic passage)
3. Long-term flow field condition
E. Transient-state groundwater flow :

1. Flow field influence by time
2. Position of phreatic line changing with time
3. Applicable to problems where pore pressure and hydraulic boundaries are changing with
time.

25
Danang.

26
Danang.
250 of 375
GROUNDWATER FLOW: STEADY STATE
Calculation based on setup of:
1. y conditions:
Boundary
a) Prescribed water levels (constant)
b) Closed flow boundaries (bottom

(bottom, axis of symmetry)
c) Wells and drains (constant)
d) Interface elements (on=impermeable,

(on=impermeable off=permeable)
e) Inflow / outflow (constant)
2. Soil permeabilities
3. Phreatic level in the soil is being calculated for t=

27
Danang.
GROUNDWATER FLOW: TRANSIENT-STATE
Transient groundwater flow:

1. Boundary conditions
a) Prescribed water levels (changing with time)
b) Closed flow boundaries (bottom, axis of symmetry)
c) Wells and drains (changing with
ith time)
d) Interface elements (on=impermeable, off=permeable)
e) Inflow / outflow (changing with time)
2. Flow field changes in time:

a) Constantl changing nat
Constantly natural
ral water
ater conditions
b) Relatively fast building process, pumping, wells
3. Embankments with river changes, tidal change
4. Reservoir impoundment and drawdown
5. Precipitation problems

28
Danang.
251 of 375
Pore pressures steady-state
Steady-state flow
3 28 29 6 9 30 31 2
4 8 11 5
General General
General
16 17
21 26
19
9 18
8
20 23 24 27
22 25
13 14 12
15 7 10
0 1

29
Danang.
SOME POINTS: FINITE ELEMENT MODELING
1. GWF calculation generally needs finer mesh than deformation analysis

2. GWF calculation generally needs large number of steps than deformation
y
analysis
3. GWF calculation usually converges, but can be problematic when:
a) Mesh is too coarseness
b) Elements are distorted
c) Large differences in permeabilities

30
Danang.
252 of 375
SOME POINTS- FINITE ELEMENT MODELING
1. Qualitative evaluation:
Flow field
Location of phreatic line
2. Quantitative evaluation:
Heads, pore pressures compared to hydrostatic,
Compare with measurements or field experience

31
Danang.

D. HYDRAULIC MODELS IN PLAXIS

32
Danang.
253 of 375
Groundwater flow flow in unsaturated soil
Water content and permeability in unsaturated zone
k k rel k sat , k rel f h p , S

(h p )
S (h )
n
hp= => pressure head

33
Danang.

A
A. Linear Model
krel
1
hp = -
hp = 0
m
0 hp

1 hp 0 Saturated For numerical stability

k rel 1 m h p 0 hp Partially saturated
hp 1 Ae
Dry
3 N int

34
Danang.
254 of 375

Soil Water Characteristic Curve (SWRC or RC)

35
Danang.
van Genuchten model
1 g n

S ( h p ) S res ( Ssat Sres ) 1 g a h p
gn ( g )
n
2
g n 1

g n

g n
krel S Se l 1 1 Se n
g g 1
with S Sres
Se
Ssat Sres
Ssat,S
Sres: saturated and residual saturation
ga, gn and gl: curve fitting parameters
36
Danang.
255 of 375
Approximate van Genuchten model
1 if hp 0
Linear in Saturation

hp
S hp 1 iff hps h p 0
h ps
0 if hp hps

1 if hp 0
4h p Log-linear in Permeability
h
k rel h p 10 ppk if h pk h p 0
4
10 if h p h pk

hps: length of partially saturated zone under hydrostatic conditions
hpk: pressure head at krel=10-44

37
Danang.
Groundwater flow - material data sets

A. Parameters:
1. Permeabilities (kx, ky)
2. Void ratio (to calculate storage)
3. Elastic storage coefficient
(The volume of water that a unit volume of saturated soil loses due to
a unit reduction in the applied water head)
4. Maximum unsaturated zone height
B. Soil classification
1. Particle fractions
2. Predefined series (Staring, Hypres, USDA) with Van Genuchten and
Approx van Genuchten parameters
Approx. parameters.
3. User-defined

38
Danang.
256 of 375
Soil classification - Staring

Dutch soil classification system
18 upper soils data sets
18 lower soil data sets
Upper soils:
U il
< 1m below soil surface
Lower soils:
all deeper
p soils

39
Danang.
Soil classification: Hypres
Hydraulic Properties of
European Soils
Particle distribution:
< 2m
2m - 50m
50m 2mm
5 upper soils data sets

5 lower soil data sets
1 organic soil data set

40
Danang.
257 of 375
Soil classification: USDA

United States Department
of Agriculture
Particle distribution:
< 2m
2m - 50m
50m 2mm
12 soils data sets
No difference between
upper and lower soils

41
Danang.
Soil classification and Van Genuchten parameters
Relative permeability
Degree of saturation

42
Danang.
258 of 375

E. EXAMPLES

43
Danang.

44
Danang.
259 of 375

45
Danang.

46
Danang.
260 of 375

47
Danang.

48
Danang.
261 of 375

49
Danang.

50
Danang.
262 of 375
REFERENCES
A. Galavi, V. (2010), Groundwater flow, fully coupled flow deformation and undrained analyses in
Plaxis 2D and 3D. Technical Report, Plaxis B.V.

51
Danang.
263 of 375
Excavation and dewatering
EXCAVATION AND DEWATERING
264 of 375
265 of 375

INTRODUCTION
This example involves the dry construction of an excavation. The excavation is supported by concrete diaphragm
walls. The walls are tied back by pre-stressed ground anchors. The Hardening Soil model is used to model the
soil behaviour. Special attention is focused on the output, which provides us insight in the soil behaviour and its
interaction with structural elements. It is noted that the dry excavation involves a groundwater ow calculation
to generate the new water pressure distribution.
INPUT
The excavation is 20 m wide and 10 m deep. 15 m long concrete diaphragm walls of 0.35 m thickness are used to
retain the surrounding soil. Two rows of ground anchors are used at each wall to support the walls. The upper
anchor has a total length of 14.5 m and an inclination of 33.7o (2:3). The lower anchor is 10 m long and is installed
at an angle of 45o . The excavation is symmetric so only one half of the problem needs to be modelled.
Figure 1: Excavation supported by tie back walls
The relevant part of the soil consists of three distinct layers. From the ground surface to a depth of 3 m there is
a ll of relatively loose ne sandy soil. Underneath the ll, down to a minimum depth of 15 m, there is a more or
less homogeneous layer consisting of dense well graded sand. This layer is particular suitable for the installation
of the ground anchors. In the initial situation there is a horizontal phreatic level at 3 m below the ground surface,
(i.e. at the base of the ll layer) Below the sand layer there is a loam layer which extends to large depth.
Geometry model
The symmetric problem can be modelled with a geometry model of 32 m width and 20 m depth. The proposed
geometry model is given in gure 2. A ground anchor can be modelled by a combination of a node-to-node anchor
and a geogrid (yellow line). The geogrid simulates the grout body whereas the node-to-node anchor simulates
the anchor rod. The diaphragm wall is modelled as a plate. The interfaces around the plate are used to model
soil-structure interaction eects. They are extended under the wall for 1.0 m to allow for sucient exibility and
accurate reaction forces. Interfaces should not be used around the geogrids that represent the grout body. In
general, it is a good habit to extend interfaces around corners of structures in order to allow for sucient freedom
of deformation and to obtain a more accurate stress distribution. When doing so, make sure that the extended
part of the interface is always turned o in the water conditions mode.
266 of 375

(10,0)
(0,0) (45,0)
0 1 2
(0,-3) 10 12 3
(45,-3)
(0,-7)
9 13 (19,-9)
(0,-10) 16
8 14 (22,-11)
17 18
19
(17,-14)
(0,-17) 7 11 4
(45,-17)
15
(10,-18)
(14,-11)
(0,-35) (45,-35)
6 5
Figure 2: Geometry model of building pit
Material properties
The soil consists of three distinct layers. The parameters of the dierent layers are shown in table 1. The interfaces
around the wall will be left impermeable in order to block the ow through it. Since the interfaces in the loam
layer below the wall (the extended part of the interfaces) do not inuence the soil behaviour, therefore their
strength is not reduced and the permeability must be changed to permeable. This will be achieved during the
denition of the staged construction phases.
Table 1: . Soil and interface properties.

Parameter Symbol Fill Sand Loam Unit
Material model Material model HSsmall HSsmall HSsmall
Drainage type Drainage type Drained Drained Drained
Unsaturated soil weight unsat 16.0 17.0 17.0 kN/m3
Saturated soil weight sat 20.0 20.0 19.0 kN/m3
ref
Reference secant stiness from triaxial test E50 20.5 103 38.5 103 20.0 103 kN/m2
ref
Reference tangent stiness from oedometer test Eoed 20.5 103 35.0 103 20.0 103 kN/m2
Reference unloading/reloading stiness ref
Eur 61.5 103 115.5 103 60.0 103 kN/m2
Power for stress-dependent stiness m 0.5 0.5 0.7
Cohesion c' 1.0 1.0 8.0 kN/m2
Friction angle 30.0 34.0 29.0 o
Dilatancy angle 0.0 4.0 0.0 o
Threshold shear strain 0.7 1.0 104 1.0 104 1.5 104
Reference small-strain shear modulus Gref
0 180.0 103 350.0 103 180.0 103 kN/m2
Advanced parameters Default Default Default
Horizontal permeability kx 1.0 0.5 0.1 m/day
Vertical permeability ky 1.0 0.5 0.1 m/day
Interface strength reduction Rinter 0.65 0.7 Rigid
Coecient for initial horizontal stress K0 Automatic Automatic Automatic
267 of 375

The properties of the concrete diaphragm wall are entered in a material set of the plate type. The concrete has a
Young's modulus of 35 GPa and the wall is 0.35 m thick. The properties are listed in table 2.
Table 2: Properties of the diaphragm wall

Parameter Symbol Diaphragm wall Unit
Material type Material type
Axial stiness EA 1.2 107 kN/m
Flexural stiness EI 1.2 105 kN/m2 /m
Weight w 8.3 kN/m/m
Poisson's ratio 0.15
For the properties of the ground anchors, two material data sets are needed: One of the Anchor type (anchor rod)
and one of the Geogrid type (grout body). The Anchor data set contains the properties of the anchor rod and
the Geogrid data set contains the properties of the grout body. The data are listed in tables 3 and 4.
Table 3: Properties of the anchor rod

Parameter Symbol Anchor rod Unit
Material type Material type Elastic
Axial stiness EA 2.5 105 kN
Spacing Ls 2.5 m
Table 4: Property of the grout body

Parameter Symbol Grout Unit
Material type Material type Elastic
Axial stiness EA 1.0 105 kN/m
Mesh generation
For the generation of the mesh it is advisable to set the Global coarseness parameter to Medium. In addition, it
is expected that stress concentrations will occur around the two grout bodies and in the lower part of the wall,
hence local renements are proposed there.
After generating the mesh, continue to the calculation.
268 of 375

CALCULATION
The calculation consists of the initial phase and six phases.
In the rst phase the wall is constructed.
In the second phase the rst 3 m of the excavation are constructed without connection of anchors to the
wall. At this depth the excavation remains dry.
In the third phase the rst anchor is installed and prestressed.
The fourth phase involves further excavation to a depth of 7 m, including the de-watering of the excavation.
This involves a groundwater ow analysis to calculate the new pore water pressure distribution, which is a
part of the denition of the third calculation phase.
In the fth phase the second anchor is installed and prestressed.
The sixth phase is a further excavation (and de-watering) to the nal depth of 10 m.
The calculation will be done using 2 alternative methods. In the rst method the water will be lowered using
steady-state groundwater ow analysis. This method assumes that excavation is suciently slow that the ow
eld will reach a steady-state situation for every excavations step. For rather slow excavations in high permeable
soils this is a reasonable assumption. In the second method the water will be lowered using a transient ow
analysis. This method is the preferred method if the excavation is suciently fast that no steady-state situation
will be reach during excavation.
Method 1: Steady-state groundwater ow

In this method a so-called semi-coupled analysis will be performed. This means that the groundwater ow eld
is generated rst and used as input to the deformation analysis. In other words, the groundwater ow will have
an eect on the deformations in the soil, but the deformations in the soil will not change the ow eld. This
assumption is reasonable if the ow eld will not be disturbed by excess pore pressures resulting from undrained
behaviour or by signicant changes in permeability due to large deformations. In this excavation problem indeed
permeabilities are high and undrained behaviour should be of little or no importance.
For this method the Calculation mode should be set to Classical mode in the Select calculation mode window that
appears directly after opening PLAXIS Calculations. If the incorrect mode is chosen one can still change this by
selecting the Calculation mode option from the Tools menu.
All calculation phases are dened as Plastic calculations of the Staged construction type and standard settings
for all other parameters. The instructions given below are limited to a description of how the phases are dened
within the Staged construction mode.
Initial phase
Set the Calculation type to K 0 procedure for calculating the initial stresses.
Press the Dene button on the Parameters tabsheet to dene the initial situation
In Staged construction mode make sure that all soil is activated and all structural elements are deactivated,
then continue to Water conditions mode.
Draw a horizontal phreatic level from (x,y) = (-2,-3) to (20,-3), (30,-3) and (47,-3).
Pore pressures will be generated based on this phreatic line. To do so, make sure the Generate by phreatic
level button is selected.
Press the Water pressures button to view the pore pressures.

After inspecting the initial pore pressures, close the Output program and press the Update button to return
to the Calculations program.
269 of 375

Phase 1: Construction of the diaphragm wall
Construction of the diaphragm wall takes 5 days. Therefore, ll in a Time interval of 5 days on the
Parameters tabsheet.
Go to the phase denition by pressing the Dene button on the Parameters tabsheet.
In Staged construction mode activate the wall and the interface extensions below the wall. The interfaces
along the wall are activated automatically.
Phase 2: First excavation stage

On the Parameters tabsheet select the option Reset displacements to zero as we're not interested in the
displacements caused by the installation of the wall.
Also on the Parameters tabsheet, enter a construction time of 7 days in the Time interval eld.
Go to the phase denition by pressing the Dene button on the Parameters tabsheet.
In Staged construction mode deactivate the rst excavation part.
Press Update to return to the Calculations program.
Phase 3: Prestress rst anchor

Prestressing the rst row of anchors will take 1 day, hence enter a 1 day construction time.
In Stage construction mode of the phase denition activate the upper geotextile representing the grout body
of the rst anchor.
Double-click on the top node-to-node anchor, the properties window for the node-to-node anchor appears.
Select the option Adjust prestress and enter a 120 kN/m prestress force.
Close the properties window and return to the Calculations program.
Phase 4: Second excavation stage and dewatering

The second excavation stage including dewatering will take 10 days, hence enter a 10 days construction time.
In Staged construction mode deactivate the second excavation stage.
Switch to Water conditions mode.
No water ow can occur through a axis of symmetry. Therefore the axis of symmetry must be a closed ow
boundary. To do so, select the Closed boundary button and draw a closed boundary from (x,y) = (0,0)
to (0,-35). Check that the bottom of the geometry is also a closed boundary.
During excavation the water level will be lowered. Due to high permeabilities water will be drawn from
outside the excavation, hence a groundwater ow analysis has to be performed. Therefore, make sure the
Groundwater button is set to Groundwater ow steady state by clicking the down arrow and choosing
the correct option.
The groundwater head boundary conditions needed for the groundwater ow analysis can be applied in a
simple manner by using the general phreatic level. In order to do so, make sure no cluster is selected (for
instance by clicking completely outside the geometry so that the general phreatic line is red) and then draw
a new general phreatic level from (x,y) = (-2,-7) to (20,-7), (30,-3) and (47,-3).
Press Update to return to the Calculations program.
270 of 375

Phase 5: Prestress second anchor
Prestressing the second row of anchors will take 1 day, hence enter a 1 day construction time.
In Stage construction mode of the phase denition activate the lower geotextile representing the grout body
of the second anchor.
Double-click on the lower node-to-node anchor, the properties window for the node-to-node anchor appears.
Select the option Adjust prestress and enter a 200 kN/m prestress force.
Close the properties window and continue to Water conditions mode.
The phreatic line should be still the same as in the previous calculation phase and also the option Ground-
water ow steady-state should still be selected.
Return to the Calculations program.
Phase 6: Third excavation stage and dewatering

The third excavation stage including dewatering will take 7 days, hence enter a 7 days construction time.
In Staged construction mode deactivate the third excavation stage.

Switch to Water conditions mode.
Check that both the axis of symmetry and the bottom of the model are closed boundaries.
Make sure the Groundwater button is set to Groundwater ow steady-state .
Draw a new general phreatic level from (x,y) = (-2,-10) to (20,-10), (30,-3) and (47,-3).
Nodes for load displacement curves

Select some nodes for load displacement curves, for instance the top of the wall at (x,y) = (10,0) and the middle
of the excavation bottom at nal depth at (x.y) = (0, -10).
Now start the calculation.
271 of 375

Method 2: Transient groundwater ow
In this method a fully coupled analysis will be performed. This analysis couples transient groundwater ow,
consolidation and deformations implying that the groundwater ow eld, development and dissipation of excess
pore pressures and deformation are calculated simultaneously all inuencing each other. This type of analysis
should be performed if developement of excess pore pressures is expected inuencing the ow eld or when
signicant changes in permeability due to large deformations are likely to occur. In this excavation problem the
main reason to use this analysis is to take into account that the ow eld will not reach a steady-state during
excavations. The addidional eects of coupling the ow eld with undrained behaviour will probably be small as
this project deals with high permeabilities. Note that a fully coupled analysis requires that the calculation type
is Consolidation.
It is possible to re-use the project made for the calculation using the method of steady-state analysis:
In PLAXIS Calculation save the project under a dierent name
From the menu Tools select the option Calculation mode. In the window that now appears select Advanced
mode.
Change the calculation phases according to the description below. Note that only the changes relative to
the steady-state calculation method are mentioned.
Initial phase
No changes have to be made
Phase 1: Construction of the diaphragm wall

Set the calculation type to Consolidation on the Parameters tabsheet.
Phase 2: First excavation stage

Phase 3: Prestress rst anchor

Phase 4: Second excavation stage and dewatering

Dene the staged construction phase and switch to Water conditions mode.
During excavation the water level will be lowered. However, due to the short construction time it's unlikely
that the ow eld will be steady state and therefore a transient groundwater ow analysis will be done.
Therefore, make sure the Groundwater button is set to Groundwater ow transient by clicking the
down arrow and choosing the correct option. The phreatic level remains unchanged.
272 of 375

Phase 5: Prestress second anchor
Though the phreatic level in the excavation doesn't change, the ow eld is not steady-state yet outside the
excavation. Therefore this phase needs transient ow analysis without making further changes.
Make sure the option Groundwater ow transient is selected.

Phase 6: Third excavation stage and dewatering

Also on the Parameters tabsheet, set the number of Additional steps to 500.
Make sure the Groundwater button is set to Groundwater ow transient .

Start the calculation
273 of 375

OUTPUT
Figure 3 gives the total displacements for the nal phase for both the calculation with steady-state groundwater
ow and the transient groundwater ow.
The excavation using steady-state ow gives a maximum displacements of about 24 mm while excavation using
transient ow gives a maximum displacement of about 23 mm.
Figure 3: Total displacements for the steady state ow analysis (left) and the transient ow analyis (right)
Figure 4 shows the vertical displacements for the nal phase for both calculations. For the displacements behind
the wall the excavation using steady-state analysis clearly gives more vertical displacements over a larger distance
from the excavation than the excavation with transient ow.
Figure 4: Vertical displacements for the steady state ow analysis (left) and the transient ow analyis (right)
The extreme bending moments are about -165 kNm/m and 75 kNm/m for the excavation using steady-state
groundwater ow analysis while the extremen bending moments for the excavation using transient groundwater
ow are about -170 kNm/m and 95 kNm/m.
Figure 6 shows the horizontal displacements of the top of the wall as a function of construction time for both the
excavation using steady-state ow and transient ow.
274 of 375
Figure 5: Bending moments in the wall for the steady state ow analysis (left) and the transient ow analyis
(right)
Figure 6: Horizontal wall displacements for the excavation
275 of 375
GENERATION OF INITIAL STRESSES

Dr Phung Du
uc Long
Some slides from :William Cheang

C Plaxis AsiaPac
CONTENTS
Generation of Initial Stresses
1
1. Ko Procedure
Ko-Procedure
2. Gravity Switch On
276 of 375
2
Ko and Gravity Switch-On Proced

dure
INITIAL STRESSES
INITIAL STRESSES
1. Initial stresses represent the equilibrium sttate of the undisturbed soil and consist of:
a) Soil weight
b) Loading history
2. In Plaxis two possibilities exist:
a) K0 procedure
b) Gravity loading
277 of 375
K0-PROCEDURE
1. Generation of initial stresses during Initial Phasse Material Set
2. Require coefficient of earth pressure at rest Ko

o
Default automatically calculated using 1-ssin phi, or
Manual
3 Disadvantage:
3. Di d t N equilibrium
No ilib i ffor iinclined
li d surfa
face
4. Advantage: No displacements are generated, only

o
stresses
t (with
( ith reference
f to
t Gravity
G it method)
th d)
'h 'v K 0
Effective Principle Stress (Gravity Switch On)
GRAVITY LOADING
1. Calculation of initial stresses by weight
loading.
2. Disadvantage: Non-physical displacements
are created.
3. Advantage: Equilibrium satisfied in all cases.
For 1D compression:

'n 'v Non-physical displacements
1 reset in subsequent phase)
so

K0
1
278 of 375
GRAVITY LOADING
1. Procedure
a. Initial phase
a Skip K0 procedure,
a. procedure Mweight remains zero
b. Phase 1
a. Choose Plastic calculation, Total multip
pliers
b Set
b. S t weight
i ht multiplier
lti li Mweight
M i ht = 1
c. Phase 2
a. Select Reset displacements to zero to discard all displacements from raising the gravity
GRAVITY LOADING
Points
1. Undrained material
a. Select Ignore undrained be

ehaviour in Phase 1 to prevent
the generation of unrealisticc excess pore pressures
2. K0 procedure has been used firsst
1 In the Initial phase redo the

1. e K0 procedure,
procedure but with
Mweight = 0; this will rese
et all initial stresses to zero.
279 of 375
Since Plaxis 2D version 2010
1. We have streamlined the

procedure
2. By using Gravity
Gravity Loading
Loading
reset displacement is
automatically used in
subsequent phase
3. Also some statements are
given in the remarks window
GRAVITY LOADING
Cases where gravity loading should be ussed instead of K0-procedure:
280 of 375
SPECIAL CASES
Gravity loading needed due to geometrry, but

K0
1
Gravity loading needed due to geomettry, but initial OCR or POP

required
K0 procedure using Phase 1: Phase 2:

d i d K0, OCR,
desired OCR POP E
Excavate exccess soil
il R
Reset di
displacements
l
Check K0, OCR !
Initial stresses
SPECIAL CASES
1. For complex initial situations like inn
ner city building projects it may be
needed to use several calculation ph hases to model the current situation
b f
before starting
t ti theth actual
t l project.
j t
existing buildings
our
project
our project
initial p ase 1
pha pphase 2 our p
project
j
reset displacements
281 of 375
FACTOR OF SAFETY ANALYSIS

PHI-C REDUCTION
Dr William WL Cheang, Plaxis AsiaPac
Dr Boonchai Ukritchon,
Ukritchon Chulalongkorn University
PlaxisIntroductoryCourse,Danang1012July2013 1
Phi-c Reduction in Plaxis (Strength Reduction Method1,2,3,4,5&6)
Main advantages:
1. Requires no a-priori assumptions on the failure mechanism
2. Critical surface is found automatically as slope failure occurs naturally through the zones due to
insufficient shear strength to resist shear stresses.
3. No requirement of assumptions on the inter-slice shear force distribution
4. Applicable to complex conditions
5. Information such as stresses, movements and pore-pressures and numerical tool as for
6
6. Powerful alternative approach
282 of 375
PRELIMINARIES
PHICREDUCTION
PlaxisIntroductoryCourse,Danang1012
3
July2013
Safetyfactor
Many possible definitions
Manypossibledefinitions
a v a ila b le s o il r e s is ta n c e
1 .8
m o b iliz e d s o il r e s is ta n c e
fa ilu r e lo a d
5 .9
w o r k in g lo a d
PLAXIS:safetyfactoronsoilresistance
f f
4
July2013
283 of 375
FactorofSafety(PhiCReduction
Phi/creduction:
a. Reductionofstrengthparameterscandtan()until
failure is reached.
failureisreached.
b. Thefactorofsafetyistheratioofinitialandreduced
strength
c tan
Msf
creduced tan reduced
c
cm
FS
tan
t m
tan
FS
PlaxisIntroductoryCourse,Danang1012 5
July2013
Safety factor analysis

Calculation procedure:
1. Create a phi/c reduction phase (in version 2010 select Safety)
2. Accept the default increment for MSF=0.1 from the multiplier
t b h t
tab-sheet.
3. Calculate
4. Carefully examine Msf vs. displacement
curve in Plaxis Curves
Notes:
a
a. Select control point within (expected) failing body
b. Use sufficient number of load steps (250-500?)
c. Use a sufficiently fine mesh (Check for mesh sensitivity)
d. Limit the maximum structural forces by choosing elastoplastic behaviour
for walls, anchors and geotextiles.
284 of 375

Safetyfactoranalysis
Number of load steps
1.16
1.16
1.12
1.12
Sum-Msf
um-Msf
1.08
1 08
1.08
S
Su
1.04
1.04
11.00
1.0
0.0 0.3 0.6 0.9 1.2 1.5 0.3 0.6 0.9 1.2
0.0
displacement
displacement
Insufficient steps!!! (no steady line)
7
July2013
Use different plots to check failure mechanism
1.Arrowsofincremental 2.Shadingsofincremental 3.Shadingsofincremental

displacements displacements shearstrains
285 of 375
PART 1: TOTAL STRESS ANALYSIS

PART1:TOTALSTRESSANALYSIS
PHICREDUCTION
9
July2013
Total stress analysis:

y undrained stability
y
(13.737,5) (25,5)
Clay
((10
10,0)
0) t =16kN/m3
(0 0)
(0,0) 20
20
c=su =20kN/m2
=0
Eu/s
/ u =500
500
(10,5) Rigidstratum (25,5)
286 of 375
Taylors stability chart
Su/(H)
/( H)
Taylors stability chart
Su/(H)
/( )
5m Clay
20
10
10m t =16kN/m
k / 3
c=su =20t/m2
0.165
D/H=2;slopeangle=20
Chart:1/Nf =0.165,Nf =6.061
Nfield =H/s
H/ u =(16)(5)/20=4
(16)(5)/20 4
FS=6.061/4=1.515
287 of 375
PART 2: EFFECTIVE ANALYSIS

PART2:EFFECTIVEANALYSIS
PHICREDUCTION
13
July2013
AutoSLOPE: LEM
SpencersMethod:
Classicalmethod: FS=1.547
Taylorsstabilitychart
FS=1.515
288 of 375
AutoSLOPE: Comparisons
AutoSLOPE:Comparisons
15
July2013
Geometry & mesh
289 of 375
Factor of safety
Plaxis:FS=1.537
Classicalmethod:
Taylorsstabilitychart
y y
FS=1.515
SpencersMethod:
FS=1.547
Failure mechanism: deformed mesh
290 of 375
Failure mechanism: FS analysis
Plaxis:FS=1.537
Comparisons
Plaxis:FS=1.537
SpencersMethod:
FS=1.547
291 of 375
Effective Stress Analysis
(20,10) (30,10)
Soil(drained)
Soil (drained)
t =16kN/m3
(0,5) (10,5)
E=10000kN/m2
c =5kN/m2
=20
(0,0) (30,0)
Limit equilibrium method

M th d off Slice:
Method Sli AutoSLOPE
A t SLOPE
SpencersMethod:
FS=1.409
292 of 375
AutoSLOPE: Comparisons
AutoSLOPE:Comparisons
23
July2013
Modeling
Excavate1
Excavate 2
Excavate2 Excavation
Excavation
problem
Fill2
Fillproblem
Fill 1
Fill1
293 of 375
Geometry & mesh
Define: K0 initial stress
Defaultvalue:
K0=1sin v =z
(Jakyseqn.) v=
vu
u
h=K0v
294 of 375
Analysis Stage
stage1
excavation1
stage2
excavation2
Excavation1
Deformed
mesh:h
excavation 1,2
,
Excavation2
295 of 375
Factor of safety:
e ca ation
excavation
Plaxis:FS=1.373
SpencersMethod:FS=1.409
Failure
mechanism
Plastic
p
points
296 of 375
Comparisons
Plaxis: FS = 1 373
Plaxis:FS=1.373
Spencer sMethod:
Spencers Method:
FS=1.409
Analysis: Fill embankment
Fill2
Fill1
297 of 375
Geometry & mesh
Elementsnotusedintheinitialgeometry
Elementsnotused
intheinitialgeometry
Define initial geometry &

i iti l stress
initial t
298 of 375
Stage1:Fill1
Stage
construction
Stage 2 Fill2
Stage2:Fill2
Fill1
Deformed
mesh: h
Fill 1,2
12
Fill2
299 of 375
Factor of safety:
Fill problem
bl
Plaxis:FS=1.376
Plaxis FS 1 376
SpencersMethod:FS=1.409
Failure
mechanism
Plastic
points
300 of 375
Comparisons
Plaxis:FS=1.376
SpencersMethod:
FS=1.409
Comparison: Deformed mesh

Comparison:Deformedmesh
Excavation 2
Excavation2
Fill 2
Fill2
40
July2013
301 of 375
Comparison: Failure mechanism

Comparison:Failuremechanism
Excavation:FS
Fill:FS
41
July2013
Comparison: Failure zone

Comparison:Failurezone
Excavation:FS
Fill:FS
42
July2013
302 of 375
PART 3: ADDITIONAL TEST CASES (2D & 3D)

PART3:ADDITIONALTESTCASES(2D&3D)
PHICREDUCTION
43
July2013
Example 2: homogenous slope with and without foundation layer
Model Slope G1 (753 elements-15n) :

FOS 1.347
FOS= 1 347 (1.400
(1 4001, 1.380
1 3802)
Note:
1 Griffiths & Lane (1999)
Model Slope G2 (757 elements-15n) : 2 Bishop & Morgenstern (1960)
FOS= 1.323 (1.4001, 1.7522,1.2794,1.3795, 1,3756) 3.Taylor (1937)
4.Janbu
5.Bishop
6 Morgenstern-Price
6.Morgenstern Price
303 of 375
EXAMPLE 3: UNDRAINED STABILITY OF A SLOPE WITH A WEAK UNDERLAYER
Automatic detection of most critical

shear surface:
Toe becomes critical when lower soil layer is much
Cu1 stronger
Cu2
Cu2/Cu1 = 2
Mixed mechanisms toe or deep-seated?

Cu1
Plaxis FOS = 1.34
Cu2
Cu2/Cu1 = 1.5
Deep seated becomes critical when lower soil layer has

the same strength with upper soil
cu = 50 kPa
Cu1 Plaxis FOS = 1.19
Cu2/Cu1 = 1.0
Cu2
c = 60 kPa
u
45
July2013
Example 4: undrained clay slope with a thin layer
2.0
1.8
1.6
1.4
Factor of safety)
1.2
Model Slope G3A cu2/cu1=1 (3436 elements- Model Slope G3D cu2/cu1=0.8 (3436 elements-
1.0
15n) 15n)
0.8
FOS 1.428
FOS= 1 428 (1.47
(1 471, 1.47
1 473) FOS= 1.384 (1.451) 0.6
0.4
0.2
0.0
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1
Cu2 /Cu1)
Note:
Model Slope G3B cu2/cu1=0.6 (3436 elements- Model Slope G3F cu2/cu1=0.5 (3436 elements- 1 Griffiths & Lane (1999)
15n) 15n) 2 Bishop & Morgenstern (1960)
FOS= 1.319(1.401, 1.404) FOS= 1.112(1.251) 3 T l (1937) ((green liline))
3.Taylor
4.Janbu
Model Slope G3E cu2/cu1=0.4 (3436 elements- Model Slope G3c cu2/cu1=0.2 (3436 elements-
15n) 15n) :
FOS 0.903(1.05
FOS= 0 903(1 051) FOS= 0.470 (0.591, 1.304)
304 of 375
EffectonGlobalFOSbyc/phiReduction
CBP Elastic Failure

CBPElastic,Failure CBP Elasto-Plastic
withnoPlasticHinge, Failure with Plastic
FOS=1.75 Hinge, FOS=1.40
1. Elastic wall excludes possibility of wall plastic hinge; and over-estimate FOS=1.75
2
2. All i for
Allowing f wallll plastic
l ti hinge
hi (El
(Elasto-plastic
t l ti wall) ll) gave llower FOS
FOS=1.40
1 40 and
d smaller
ll soilil yielded
i ld d zone
behind the wall
Some points
1. Always inspect the incremental displacements or strains as computed in the last
load step to make sure that failure is reached.
2. The mesh used in the calculation needs to be sufficiently fine.
3. Mesh: Refine and redo the phi-c analysis until the factor of safety remains constant
upon further refinement of the mesh.
4 Al
4. Always use th
the arc-length
l th titime stepping
t i procedure
d within
ithi th
the Phi C reduction
Phi-C d ti
(default)
5. Use a small tolerated error (maximum should be the default error of 1% )
305 of 375
Appendix:
Appendix:2
2Dvs.3
Dvs.3D(Benz)
1 3Danalysismayyieldsubstantiallydifferentresultsfrom2Danalysis
1. 3D analysis may yield substantially different results from 2D analysis
2. TheadvantageofFEMoverclassicaldesigntoolsisobvious.I
3. ntheexample:stabilityofabentoniteslurrytrench.
Plane strain (14 m) 3D (40 m) Axis symmetric (40 m)

FOS = 1.17 FOS = 1.90 FOS= 2.69
49
July2013
TheCaseofGriffithandMarquez2007(Case1)
50
July2013
306 of 375
Case1 (GriffithandMarquez,2007) Asimple2:1slopestabilityverification
Incremental disp
Incremental shear strains
51
July2013
SlipSurface:Spoonunderisosurfaces
p p
L/H=1 L/H=2 L/H=4
L/H=8 L/H=12
52
July2013
307 of 375
Case:InfluenceofSlopeLength/HeightratioonFOSin3D
53
July2013
Case2:InclinedSideFace(tobecontinued)
54
July2013
308 of 375
References
1
1. M
Matsui, i T
T. & San,
S K.C.
K C (1992) Finite
Fi i element
l slope
l stability
bili analysis
l i b by shear
h strength
h reduction
d i technique.
h i S il and
Soils d
Foundation, Vol.32 (1),pp.59-70
2. Zienkiewicz, O.C., Humpheson, C. & Lewis, R.(1975) Associated and non-associated visco-plasticity and plasticity in
soil mechanics. Geotechnique 25(4).pp. 671-689.
3
3. Ugai K.(1989).
Ugai, K (1989) A method of calculation of total safety factor of slope by elastoplastic FEM.
FEM Soils and Foundation 29(2).
29(2)
pp.190-195.
4. Farias, M.M., Naylor, D.J.(1998). Safety analysis using finite elements. Computer and Geotechnics, Vol 22(2) p.p. 165-
181.
5. Griffiths, D.V., Lane, P.A. (1999). Slope stability analysis by finite elements. Geotechnique 49 (3), pp.387-403.
pp.387 403.
6. Griffith, D.V. & Marquez, R,M (2007). Three-dimensional slope stability analysis by elastoplastic finite elements,
Geotechnique 57, No. 6, 537546.
309 of 375
Slope stability for a road construction project
SLOPE STABILITY FOR A ROAD

CONSTRUCTION PROJECT
310 of 375
311 of 375
INTRODUCTION
On the North Island of New Zealand a new road section has to be constructed along the shore line of a tidal bay,
see gure 1.
Figure 1: Situation overview for the newly constructed road
Though the easiest solution would have been to construct the road at a larger distance from the bay as the slope
gradients are easier there, this is not possible as the upper land is privately owned which for historic reasons
cannot be changed. The new road therefore had to be constructed along the steeper gradient just next to the
shore line of the tidal bay.
The hillside is mainly siltstone, weathered at the surface but intact at certain depth. Construction will take place
in summer when the ground water level is low. However, in winter the hillside side almost fully saturates due to
heavy rainfall, which has a signicant inuence on the stability. For the construction of the new road part of the
slope was excavated. The excavated material is crushed and mixed with sand and gravel to make ll material to
support the road.
During the rst winter after the road construction the road started to tilt towards the tidal bay and after assessing
the winter situation the factor of safety was considered too low. The decision was taken to stabilize the ll and
hillside below the road using so-called launched soil nails: long steel reinforcement bars that are shot with high
speed into the ground.
MAIN GOAL OF THE ANALYSIS

Determine the factor of safety of the original hillside
Construct the new road under dry (summer) conditions and calculate its factor of safety
Simulate wet (winter) conditions and calculate its factor of safety
Apply stabilising soil nails and calculate the factor of safety in wet conditions
INPUT
Start a new project and select appropriate General settings according to the size of the geometry (see gure 2)
and make sure to use a snap distance of 0.25m. Use 15-node elements as basic element type since in this exercise
we will deal with failure behaviour.
312 of 375
Figure 2: Geometry model (a) and position of the road surface and soil nails (b)
Geometry
Enter the geometry as indicated in gure 2a. The order in which geometry points are created is arbitrary.
Introduce the 3 soil nails by using geogrids according to the coordinates given in gure 2b.
Introduce the road surface by using a plate element from (22,16) to (28,16)
Introduce the trac load by applying a vertical distributed load of -10 kN/m2 on the road surface.
Apply default boundary conditions.
Material properties
Soil and interfaces
Enter the material properties for the three soil data sets specied in table 1.
After entering all properties for the three soil types, drag and drop the properties to the appropriate clusters,
as indicated in gure 3.
313 of 375
Table 1: Soil material set parameters

Parameter Symbol Intact siltstone Weathered siltstone Reinforced ll Units
Material model Model Mohr-Coulomb Mohr-Coulomb Mohr-Coulomb

Type of behaviour Type Drained Drained Drained
3
Dry weight unsat 16.0 16.0 19.0 kN/m
3
Wet weight sat 17.0 17.0 21.0 kN/m
2
Young's modulus Eref 12000 12000 20000 kN/m
Poisson's ratio 0.3 0.3 0.3
2
Cohesion c 8 8 8 kN/m

Friction angle 35 19 30

Dilatancy angle 0 0 0
3
Permeability x-dir. kx 110 0.01 0.1 m/d
3
Permeability y-dir. ky 110 0.01 0.1 m/d
Tension cut-o Tension cut-o Disabled Enabled Enabled -
Weath
ered s
iltston
e
A A A
Intact siltstone
Figure 3: Geometry model showing the soil material sets
Road surface
The road surface is modelled with a plate element. Therefore, create a new plate material set using the parameters
as specied in table 2 and assign it to the plate representing the road surface.
Table 2: Properties of the road surface (plate)

Parameter Symbol Road surface Unit
Material model Model Elastic

5
Axial stiness EA 2.510 kN/m
2
Flexural stiness EI 500 kNm /m
Weight w 3.0 kN/m/m
Poisson's ratio 0.0
Soil nails
The 3 soil nails are modelled using geogrid elements. Hence, create a new geogrid material set with parameters
as specied in table 3 and assign the material to all 3 soil nails.
314 of 375
Table 3: Properties of the soil nails (geogrids)

Parameter Symbol Soil nail Unit
Material model Model Elastoplastic

Axial stiness EA 5275 kN/m
Max. tension force Np 150 kN/m
Mesh generation
Set the Global coarseness to Medium.
Select all clusters that fall within the boxed area (12 clusters in total) while keeping the <Shift> button
pressed and then select the option Rene cluster from the Mesh menu in order to rene the mesh in the
selected area. This will give a mesh as given in gure 4.
Figure 4: Medium coarse generated mesh with renement
315 of 375
CALCULATION
The calculation consists of the initial phase and 12 calculation phases more in order to model the proper con-
struction sequence and the determination of the factors of safety at key moments in the construction process.
When starting the Calculations program select Classical mode for calculating undrained behaviour and consoli-
dation.
Initial phase
The initial situation consists of the intact hill side and a phreatic level representing typical summer conditions as
construction starts in summer. In order to dene the initial situation, follow these steps:
The geometry has a non-horizontal soil layering, hence the K0 -procedure cannot be used. Therefore, set the
Calculation type to Gravity loading.
Dene the Staged construction settings and make sure only the clusters representing the original hillside are
activated. Also make sure all structural elements (road surface and soil nails) are switched o.
In Water conditions mode, eEnter a phreatic level by two coordinates (-1, 10) and (56, 10).
Click the Update button to return to the Calculations program.
Phase 1 - Stability prior to the construction

Before the construction is started the factor of safety is determined of the initial situation
One the General tabsheet make sure this calculation phase is Safety.
Accept all default settings
Phase 2 - Road excavation

The road excavation should continue from the initial situation and not from the results of the safety factor
determination. Therefore, on the General tabsheet, set the eld Start from phase on Initial phase .
Set the Calculation type to Plastic of loading type Staged construction.
In order to discard the displacements during gravity loading make sure the option Reset displacements to
zero is selected on the Parameters tabsheet.
On the Parameters tabsheet press the Dene button to dene the phase
Switch o the upper part of the road excavations, see gure 5.
Press the Update to return to the Calculations program.
316 of 375
Figure 5: Phase 2, road excavation
Phase 3 - Construction of the ll

This calculation phase is again a Plastic calculation, loading type Staged construction.
For all the other settings the defaults should be used.
Switch on the additional ll
Open the material set database and assign the reinforced ll material set to the 4 clusters of the ll area,
see gure 6.
Close the material sets database and press the Update to return to the Calculations program.
Figure 6: Phase 3, Construction of the ll
Phase 4 - Construction of the road

This calculation phase again is a Plastic calculation, loading type Staged construction. Keep all default
settings.
Switch on the plate representing the road by clicking on it. Make sure the distributed load representing the
trac load remains switched o.
317 of 375
Phase 5 - Apply the trac load

Again a Plastic calculation, topmost level of loading type Staged construction. Keep all default settings.
Switch on both parts (left nd right) of the distributed load representing the trac load. The plate repre-
senting the road surface remains switched on.
We are now nished with the road construction.
Phase 6 - Factor of safety of the road in summer conditions

In order to determine the factor of safety directly after constructing the road use a Safety phase. Keep all
default settings.
Phase 7 - Winter conditions

The increase of water level should occur after nishing the road construction and not after determination of
the factor of safety of this situation. Therefore, on the General tabsheet, set the Start from phase parameter
to Phase 5 Apply the trac load . Keep all other settings to their default.
Set the Calculation type to Plastic, loading type Staged construction.

Go to the Staged construction denition and from there to the Water conditions mode.
Select the Phreatic level button and draw a new phreatic line from (-1,20) to (5,20) and further to (20,10)
and (56,10).
If there is no closed ow boundary yet on the bottom of the geometry (indicated with a thick black line)
then select the Closed boundary button and draw a closed ow boundary at the full bottom of the geometry.
Select Steady state as groundwater analysis type and press the Update button to return to the Calculations
program.
Phase 8 - Factor of safety of the road in winter conditions

In order to determine the factor of safety directly in winter conditions create a Safety phase. Keep all default
settings.
In winter conditions the factor of safety appears to be rather low and therefore it is decided to improve stability
by applying launched soil nails.
Phase 9 - Apply top level soil nails

The application of the rst level of soil nails should occur after calculating winter conditions and not after
determination of the factor of safety of this situation. Therefore, on the General tabsheet, set the Start
from phase parameter to Phase 7 Winter conditions . Keep all other settings to their default.
Set the Calculation type to Plastic, loading type Staged construcion.

Switch on the topmost soil nail, see gure 7.
Return to the Calculations program
318 of 375
Figure 7: Phase 9, Road construction with trac load and topmost level of soil nails
Phase 10 - Factor of safety in winter conditions with top level soil nails
In order to determine the factor of safety directly in winter conditions with the topmost level of soil nails
installed create a Safety phase. Keep all default settings
Phase 11 - Apply additional soil nails

The application of the rst level of soil nails should occur after installing the top level of soil nails and not
after determination of the factor of safety of this situation. Therefore, on the General tabsheet, set the Start
from phase parameter to Phase 9 Apply top level soil nails . Keep all other settings to their default.
Set the Calculation type to Plastic, loading type Staged construction.

Switch on the 2 other soil nails
Return to the Calculations program
Phase 12 - Factor of safety in winter conditions with all soil nails in-
stalled
In order to determine the factor of safety directly in winter conditions with the all soil nails installed create
a Safety phase. Keep all default settings
Load-displacement curves
Before starting the calculation choose some points for node-displacement curves. In order to check failure for the
phi/c reduction phases the chosen points should be in the expected failure zone. As there are several possible
slope instabilities, chose at least points at (15,20), (25,16), (28,16) and (33,11).
Now start the calculation by pressing the Calculate button.
319 of 375
INSPECT OUTPUT
Factors of safety
Check the factors of safety in the Curves program. Create a new curve of displacements vs. Sum-Msf for the
point at coordinates (25,16). See gure 8.
1.70
1.60
FoS 1.6 (all nails installed)
1.50
FoS 1.4 (top nails installed)

1.40
Msf

1.30
FoS 1.2 (summer conditions, no nails)

1.20
FoS 1.15 (before construction)
1.10
FoS 1.15 (winter conditions, no nails)
1.00
0.00 0.50 1.00 1.50 2.00
|U| [m]
Figure 8: Factors of safety for key moments in the project.
Critical failure mechanisms
For all situations check the failure mechanism. This can be done by for instance with the graph of incremental
shear strains (s ) of every Safety phase. This will show the change in shear strains in the last calculation step,
hence when failure occurred, and will show any shear bands that may have occurred. See gures 9-13. It becomes
clear that installing the top nails disturbs the failure mechanism. However, it is only after having installed the
lower nails as well that the sliding of the road ll no longer is the most critical mechanism. Both failure of the
slope above the road and a very large hillside sliding mechanism with considerably higher factor of safety (almost
1.6) are now the critical mechanisms.
320 of 375
Figure 9: Most critical failure mechanisme prior to construction
Figure 10: Most critical failure mechanisme after construction, summer conditions.
Figure 11: Most critical failure mechanisme after construction, winter conditions.
321 of 375
Figure 12: Most critical failure mechanisme after installing top soil nails
Figure 13: Most critical failure mechanisme after installing all soil nails
322 of 375
Overview of Soil Models,, Parameters,,

Possibilities & Limitations
Content
Overview of soil models,
models parameters,
parameters possibilities & limitations
in PLAXIS 3D
Mohr-Coulomb model
Hardening Soil model (HS + HSsmall)
Soft Soil model
Soft
S f S
Soil C
Creep model
Hoek-Brown model
Standard soil tests with different models
Which model in which situation?
323 of 375
Overview of soil models in PLAXIS 3D
Mohr-Coulomb model
Li
Linear-elastic
l ti perfectly-plastic
f tl l ti b behaviour
h i
Hookes law:
d xx 1 0 0 0 d xx
d 1 0 0 0 d yy
yy
d zz E 1 0 0 0 d zz

d xy (1 )(1 2 ) 0 0 0 1
2 0 0 d xy
d yz 0 0 0 0 1
0 d yz
2

d zx 0 0 0 0 0 1
2 d zx
324 of 375
Mohr-Coulomb model
Linear elastic perfectly-plastic
Linear-elastic perfectly plastic behaviour
Yield function:
f 12 '3 '1 12 '3 '1 sin ' c 'cos
cos '
Plastic potential function:
g 12 '3 '1 12 '3 '1 sin c 'cos
Mohr-Coulomb model
Parameters:
E Youngs modulus
Poissons
Poisson s ratio
c Cohesion
Friction angle
Dilatancy angle
325 of 375
Mohr-Coulomb model
Possibilities:
First order approach of soil behaviour in general
(Drained) failure behaviour quite well described
Limitations:
Linear elastic behaviour until failure (no strain- or stress- or stress path-
dependent stiffness behaviour)
Be careful with efffective strength in undrained behaviour
Hardening Soil model

Characteristics:
Stress-dependent stiffness behaviour according to
a power law
Hyperbolic stress-strain relationship in axial
compression
Plastic strain by mobilising friction (shear hardening)
Plastic strain by primary compression (compaction hardening)
Elastic unloading / reloading
Failure behaviour according to the Mohr-Coulomb criterion
Small-strain stiffness ((HSsmall model only)y)
326 of 375

Shear hardening:
Elastoplastic formulation of hyperbolic q-1 relationship

Elastic
q q
MC failure line MC failure line
plastic
m 3p,fric
2p,fric
p fric
1p,fric
q
Hardening Soil model MC failure line
Compaction hardening:
Elastoplastic formulation of
p - v relationship in Cap
primary compression pc
fc = 0
pc p
c 1
v
327 of 375

S ll t i stiffness:
Small-strain tiff
Strain-dependent stiffness
Hysteresis
Energy dissipation
Damping

Parameters:
E50ref Secant stiffness from triaxial test at reference pressure
Eoedref Tangent stiffness from oedometer test at pref
Eurref Reference stiffness in unloading / reloading
G0ref Reference shear stiffness at small strains (HSsmall)
0.7 Shear strain at which G has reduced to 70% (HSsmall)
m Rate of stress dependency in stiffness behaviour
pref Reference pressure (100 kPa)
ur Poissons ratio in unloading / reloading
c Effective cohesion
Effective friction angle
Dilatancy angle
Rf Failure ratio qf /qa like in Duncan-Chang model (0.9)
K0nc Stress ratio xx/yy in 1D primary compression
328 of 375

Possibilities:
Better non-linear formulation of soil behaviour in general
Distinction between primary loading and unload./reloading
Memory of pre-consolidation stress
Different stiffness in different stress paths
Well suited for unloading situations with simultaneous deviatoric loading
(excavations)
Strain-dependent stiffness (hysteresis, damping) (HSsmall only)
Li it ti
Limitations:
No peak strength and softening
No secondary y compression
p ((Creep);
p); No anisotropy
py
E50 / Eoed > 2 difficult to input
Soft Soil model

Characteristics:
Based on Cam-Clay theory
Logarithmic stress
stress-strain
strain relationship
(stiffness linearly dependent on p)
Plastic strain in primary compression (compaction hardening)
M based on K0nc in 1D compression instead of on
Failure behaviour according to the Mohr
Mohr-Coulomb
Coulomb criterion
329 of 375
Soft Soil model
Soft Soil model

Parameters:
* Modified compression index
* Modified swelling index

ur Poissons ratio for unloading / reloading
Dilatancy angle
K0nc Horiz./vertical
Horiz /vertical stress ratio in normally consolidated
1D compression (determines M)
330 of 375
Soft Soil model

Possibilities:
Reasonable model for primary loading of normally
consolidated clay and soft soils
Failure behaviour better than (modified) Cam-Clay
Limitations:
Less suitable for over-consolidated clay and in certain unloading stress
paths; not suitable for sand
No time-dependent behaviour (secondary compression)
Soft Soil Creep model

Characteristics:
Apparently comparable behaviour with Soft Soil model:
Logarithmic stress
stress-strain
strain relationship
Memory of preloading (pre-consolidation stress)
Irreversible volume strain upon primary loading
Failure behaviour according to the Mohr-Coulomb criterion
Time-dependent
Time dependent deformations (secondary compression)
Irreversible strains by means of visco-plasticity (creep strain) instead of
plasticity
331 of 375
332 of 375

Parameters:
* Modified compression index
* Modified swelling index

* Modified creep index
ur Poissons ratio for unloading / reloading
Dilatancy angle
K0nc Horiz./vertical stress ratio in normally consolidated
1D compression
p ((determines M))

Possibilities:
Reasonable model for primary loading of normally consolidated clay
and soft soils
Time-dependent behaviour (secondary compression; creep)
Limitations:
Side role of over-consolidation ratio OCR
Influence of K0nc-parameter (M)
No softening
333 of 375
Hoek-Brown model -
Hoek Brown failure criterion (2002):
Hoek-Brown
a
'3
'1 '3 ci mb s -
ci -
GSI 100
mb mi exp
28 14 D ci = Intact uniaxial compressive
1 1 GSI 20 strength
a exp exp
2 6 15 3 GSI = Geological Strength Index
GSI 100 mi = Intact rock parameter
s exp D = Disturbance Factor
9 3D
Hoek-Brown model -
-1
Uni axial compressive strength:
Uni-axial
c ci s a
- -
Tensile strength:
s ci
t
mb
-c
-3
-
t
334 of 375
Hoek-Brown model -
Parameters:
ci Intact uniaxial compressive strength

- -
GSI Geological Strength Index
mi Intact rock parameter
D Disturbance factor
Hoek-Brown model -
Possibilities:
Continuum approach of rock strength
- -
Limitations:
Only applicable to rock
No anisotropy
335 of 375

Model 1 (Mohr
(Mohr-Coulomb)
Coulomb) Model 2 (Hardening-Soil)
(Hardening Soil) Model 3 (Soft Soil)
E = 20000 kN/m2 E50 = 20000 kN/m2 * = 0.012
= 0.3 Eoed = 20000 kN/m2 * = 0.0024
c = 0 kN/m2 Eur = 60000 kN/m2 = 0.2
= 30 m = 1.0 c = 0 kN/m2
= 0 pref = 100 kN/m2 = 30
= 0.2 = 0
c = 0 kN/m2 K0nc = 0
0.5
5
= 30
= 0
K0nc = 0.5

Standard drained triaxial test at 3 = 100 kPa
336 of 375

Undrained triaxial test at 3 = 100 kPa

Undrained triaxial test at 3 = 100 kPa
337 of 375

Oedometer test with unloading

Direct Simple Shear test at xx = yy = 100 kPa
338 of 375

Direct Simple Shear test at xx = yy = 100 kPa
Which model in which situation?

Soft soil (NC-clay
(NC clay, Hard soils (OC-
(OC
peat) clay, sand, gravel)
Primary
y load. Soft Soil ((Crp),
p) HS / HSsmall
(surcharge) HS / HSsmall
Unloading + HS / HSsmall HS / HSsmall
d i t i lload
deviatoric d
(excavation)
Deviatoric Soft Soil (Crp),
(Crp) HS / HSsmall
loading HS / HSsmall
Secundary
y Soft Soil Creep
p n/a
compression
339 of 375
340 of 375
Consolidation analysis using finite elements

Ronald Brinkgreve
(with input from Vahid Galavi)
Plaxis BV, The Netherlands
Content
Theory of consolidation
FEM for consolidation analysis
Validation: One-dimensional consolidation
New features in PLAXIS 2D 2010
Conclusions
341 of 375
C
Considering:
id i
Homogeneous, fully saturated, isotropic linear elastic soil

Incompressible particles
Volumetric strain of saturated soil is caused by:

n
1. Pore pressure change: v1 pw v = volumetric strain
Kw pw = (excess) pore pressure
n = porosity
2. Storage of pore water: v 2 qt Kw = bulk stiffness of water
q = pore water flow
t = time increment
P
Pore water
t flow:
fl
k k
Darcys law: q pw v 2 pw t
w w
Total change of volumetric strain in time, k = permeability
considering homogeneous permeability: w = unit weight of water
v ( v1 v 2 ) n pw k 2
pw
t t K w t w
342 of 375
G
General
l 3D case:
v 1 p ' 1 ( p pw ) 1 p 1 pw

t K ' t K ' t K ' t K ' t
1 p 1 n pw k 2
pw
K ' t K ' K w t w
E'
where K' = bulk stiffness of soil skeleton and p = mean total stress
3 1 2 '
kK' pw p
Considering incompressible water: 2 pw
w t t
1D consolidation:
lid ti

v 1 ' 1 ( pw )
2H
t Eoed t Eoed t
1 1 n pw k 2
pw
Eoed t Eoed K w t w
(1 ') E '
where Eoed = constrained modulus of soil skeleton
(1 ')(1
)(1 2 '))
k Eoed pw
Considering incompressible water: 2 pw
w t t
343 of 375

1D consolidation,
lid ti considering
id i a constant
t t ttotal
t l stress
t : 0
t
pw k Eoed
cv 2 pw where cv = consolidation coefficient =
w
t
cv t
T
H2

C
Consolidation
lid ti analysis
l i b based
d on excess pore pressure (EPP)
(EPP):
pactive psteady pexcess
Assumptions:
Steady state pore pressure is constant in time (horizontal phreatic level or
steady state pore pressure from groundwater flow calculation)
Excess pore pressure can change in time
Fully saturated soil (above and below phreatic level)
Limitation:
Time dependent hydraulic boundary is not possible (variable phreatic level)
344 of 375

Consolidation
C lid ti analysis
l i b based
d on Bi
Biots
t th
theory off consolidation:
lid ti
Coupling between deformation and flow of pore water (excess pore pressures)
Differences with Terzaghis uncoupled or one-dimensional consolidation:

Instantaneous settlement in undrained loading
p
Spreading g of load and therefore smaller excess p
pore p
pressures
Complex flow of excess pore pressures
Mandell-Cryer effect: Pore pressure may not immediately reduce

I t t
Instantaneous settlement:
ttl t
axisymmetry after consolidation
345 of 375

M d ll C
Mandell-Cryer effect:
ff t
A C B D
C
axisymmetry
A
B E
D F E

Geometry
G t creation:
ti
Soil layers
Structures
Drains pw=0 in consolidation analysis
346 of 375

Soilil properties:
S i
Unit weight (sat,unsat)
Stiffness (E,)
( )
Strength (,c)
Material type (drained, undrained, non-porous), Permeability (k)
Drained behaviour is appropriate when T>2.0 (U>99%)

Permeability is high
Rate of loading is low
cv t k Eoed
Short term behaviour not relevant T Cv
No excess pore pressures H2 w
T = hydrodynamic period Cv = consolidation coefficient

Soilil properties:
S i
Stiffness (E,)
( )
Strength (,c)
Undrained analysis appropriate when T<10-4 (U<1%)

Permeability is low and rate of loading is high
Short term behaviour has to be assessed
cv t k Eoed
Calculation of excess pore pressures T Cv
H2 w
347 of 375

Soilil properties:
S ti
Stiffness (E,)
( )
Strength (,c)
Consolidation analysis appropriate when 10-4<T<2.0 (1%<U<99%)

Permeability is relatively low
Time
Time-dependent
dependent behaviour has to be assessed kE cv t
T Cv oed
Note: In consolidation analysis, drainage is ruled H2 w

byy Permeability
y rather than Material type
y

Soilil properties:
S i
Stiffness (E,)
( )
Strength (,c)
Change of permeability (Ck) k e
log
k 0 Ck
Note: Consolidation coefficient is more or less constant
B t k decreases
But: d with
ith load
l d (compaction
( ti off soil)
il) k Eoed
C
Eoed increases with load (stress level) v
w
Use realistic Ck only with advanced models
348 of 375

Soilil properties:
S i
Stiffness (E,)
( )
Strength (,c)
Change of permeability (Ck)
Creep
Note: Adding creep leads to extra generation of pore pressures

pressures,
which is perceived as delayed consolidation

setttlement
pexcess
no creep
creep
no creep time (log-scale)

creep
time (log-scale)
349 of 375

Boundary
B d conditions:
di i
Fixities
Loads
Closed consolidation boundaries (no outflow; otherwise draining boundaries: pw=0)
FEM for consolidation analysis Node

Element
Mesh:
M h Stress
Elements: Interpolation of primary variables point
Nodes: Primary y variables ((displacements,
p p
pore p
pressures))
Stress points: Derived variables (strains, stresses, Darcy velocities)
Same order of interpolation in PLAXIS
350 of 375

Calculations:
C l l i
Consolidation Staged construction > Time interval t
Consolidation Minimum ppore p
pressure > |p-stop|
|p p|
Consolidation Incremental multipliers > Time increment
Typical: Plastic calculation (staged construction) with undrained materials

materials,
followed by consolidation analysis
Advanced: All phases consolidation: SC > t > SC > t > > SC > |p-stop|
Rate loading: Time increment and load increment give loading rate

Calculations:
C l l i
Consolidation Staged construction > Time interval t
Consolidation Minimum ppore p
pressure > |p-stop|
|p p|
Consolidation Incremental multipliers > Time increment
l2
tcritical
Cv
Note: smaller steps
p may yggive
stress oscillations
l = element length
= 80 ffor 15
15-node
d ttriangles
i l
= 40 for 6-node triangles
351 of 375

C l l i
Calculations:
K v L p f Equilibrium
dv dp
H pL S q
T
Continuity
dt dt
K L v 0 0 v0 f

LT S p 0 t H p 0 t q
* * System of equations

Solution: Displacements and (excess) pore pressures

C l l i
Calculations:
Stiffness matrix
Coupling matrix
K v L p f Forces Equilibrium
Pore pressures
Flow matrix
Transposed coupling matrix
dv dp
H pL S q
T Displacements
Net flow Continuity
dt dt C
Compressibility
ibili off water
K L v 0 0 v0 f

LT S p 0 t H p 0 t q
* * System of equations

S t H S q q0 q
* *
Solution: Displacements and (excess) pore pressures
352 of 375

Output:
O t t
Deformations
Stresses
Excess pore pressure
History curves
(e g pore pressure as function of time)
(e.g.
353 of 375
New features of PLAXIS 2D 2010

N
New ttypes off calculations:
l l ti
Fully coupled flow-deformation analysis

Consolidation based on total pore pressure (groundwater flow + deformation)
Bishop stress (unsaturated behaviour)
Possibility to model unsaturated soil behaviour
V i
Various b
boundary
d conditions
diti ffor flflow ((seepage, infiltration,
i filt ti d
drain,
i well,)
ll )
New groundwater flow code (steady state + transient)
New K0 procedure (Bishop stress + Terzaghi stress)
354 of 375
New features of PLAXIS 2D 2010

U
Unsaturated
t t d soilil modelling:
d lli
Bishop stress
Suction (a new variable)
Retention curves (Mualem-Van Genuchten + user defined models)
Existing Plaxis soil models (Bishop stress)
User defined soil models (Bishop stress and suction)
Calculation modes in PLAXIS 2D 2010

Classical
Cl i l mode
d
Terzaghi stress

Semi-coupled analysis (pore pressure is
independent of deformation)
Advanced mode
Bishop stress
Unsaturated soil modelling
Fully coupled analysis
Flow mode
Steady state groundwater flow
Transient groundwater flow
355 of 375
Conclusions
FEM is quite suitable for 2D and 3D consolidation analysis
2D or 3D coupled consolidation is different from 1D or uncoupled consolidation
PLAXIS has several options for consolidation based on excess pore pressure
Adding creep gives more realistic time-dependent behavour and leads to
delayed consolidation
R
Recent t development:
d l t F
Fully
ll coupled
l d flflow-deformation
d f ti analysis
l i andd unsaturated
t t d
soil behaviour
356 of 375
Geotextile reinforced embankment
GEOTEXTILE REINFORCED
EMBANKMENT WITH CONSOLIDATION
357 of 375
2 Computational Geotecnics
358 of 375
INTRODUCTION
In 1979 a test embankment was constructed in the Netherlands near the town of Almere. The
objective of this test was to measure the influence of geotextile reinforcement on the short
term stability of an embankment on soft soil. Two test embankments were constructed on top
of a layer, one with and one without geotextile. The construction procedure was such that a
ditch was excavated in the clay layer while at the same time a retaining bank was made with
the excavated clay. A cross-section of the reinforced test embankment is given in figure 1.
line of symmetry
geotextile
retaining bank
1
sand fill 2
2
soft clay
1,5
strong sand layer
1 3.5 3.5 1 3 14 7
model width approx. 33 m
Figure 1: Cross-section of the reinforced embankment
Cone penetration tests gave an average cone resistance of qc = 150 kPa for the clay. The clay
is considered to be normally consolidated. The behaviour is assumed to be undrained (the
retaining bank should be drained, however). The saturated weight of the clay is 13.5 kN/m3.
A plasticity index of Ip = 50% is assumed. Due to the limited soil data, parameters should
be selected using engineering judgement and by using the correlations given in the lecture
"Evaluation of soil stiffness parameters". To obtain an undrained shear strength for the clay
layer it is suggested to use the correlation su qc /15. Having no data for the effective cohesion
and the effective friction angle, they have to be estimated from the undrained shear strength in
order to do a consolidation analysis. For the determination of a stiffness parameter for the clay
layer it is suggested to use the correlation Eu 15000 su /Ip (%). The shear modulus G is one
third of the undrained Youngs modulus Eu . The effective Poissons ratio should be chosen
such that a realistic K0nc is obtained in one-dimensional compression (K0nc = 0 /(1 0 ) 0.5).
The effective Youngs modulus is calculated from the shear modulus E 0 = 2G(1 + 0 ). The
fill was reported to be fully saturated loose sand with a saturated weight of 18 kN/m3 . The
behaviour is considered to be drained. The effective strength properties are estimated at 0 =
30 and c = 3 kPa. K0nc is assumed at 0.5. For the stiffness one should take E = 4000 kPa
and 0 =0.33.
359 of 375
AIMS
Calculation of two alternatives within one project.
Simulation of embankment construction in stages.
Application of geogrid elements
Review of undrained behaviour and pore pressures.
Perform consolidation analysis.
Determination of the factor of safety using phi/c reduction
1. Determination of stiffness & strength properties (clay)
2. Geometry input
(a) Start a new project

(b) Enter general settings
(c) Enter geometry
(d) Enter fixities
(e) Enter material properties for soil and geotextile
(f) Mesh generation + refine line
3. Calculation
(a) Initial conditions (Pore pressure generation, Initial geometry configuration, Genera-
tion of initial stresses)
(b) Switch on geotextile, excavate ditch + raise retaining embankment
(c) Apply first hydraulic fill
(d) Apply second hydraulic fill
(e) Determine factor of safety
(f) Repeat this using consolidation phases instead of plastic phases.
4. Inspect output
5. Suggestion for extra exercise: non-reinforced embankment
360 of 375
Note: The main purpose of the exercise is to assess the failure mechanism and
the factor of safety, which has the following consequences for the model:
There is no need to use an advanced soil model as the main

advantage of advanced models is a better prediction of
displacements.
The geometry size is chosen such that the failure mechanism fits
within the model boundaries. This means the geometry can be fairly
small.
If a deformation analysis has to be performed for this case it is

recommended to use an advanced soil model, for instance the Hardening
Soil or HSsmall model, and to choose the geometry considerably larger to
avoid influence from the boundary conditions on the results.
361 of 375
GEOMETRY INPUT
General settings
Start a new project and select appropriate General settings. Use 15-node elements as basic
element type since in this exercise we will deal with failure behaviour.
Geometry and boundary conditions
(9.5,7.5) (12,8.5)
10 11
(33,8.5)
(8,7.5)
7 9 8 (33,7.5)
(4.5,5.5) (12,5.5) (26,5.5)
(0,5.5) 1 6 12 13 2
(33,5.5)
(0,3.5) 4 5 (1,3.5)
(0,2) 14 15
(33,2)
y
(0,0) 0 x 3
(33,0)
Figure 2: Geometry model with coordinates
Enter the geometry as indicated in the previous graph. The order in which geometry
points are created is arbitrary.
Click the Geogrid button to introduce the geotextile (from (4.5, 5.5) to (26.0, 5.5)).
Click the Standard fixities button for the standard boundary conditions.
Material properties (clay)

Determine the Mohr-Coulomb strength parameters ( and c) as well as the elastic parameters
(E and ) for the clay layer from the data as given in the introduction of this exercise. The
procedure on how to determine the parameters for clay are provided at the end of this exercise.
For this exercise, we will continue with the parameters as given in table 1.
Soil and interfaces

Enter the material properties for the three soil data sets, as indicated in table 1.
After entering all properties for the three soil types, drag and drop the properties to the
appropriate clusters, as indicated in figure 3.
362 of 375
Table 1: Soil parameters
Parameter Symbol Clay Retaining Fill Stiff layer Unit

bank
Material model Model Mohr- Mohr- Mohr- Mohr-
Coulomb Coulomb Coulomb Coulomb
Type of behaviour Type Undrained A Drained Drained Drained
Unsaturated weight unsat 13.5 13.5 18.0 18.0 kN/m3
Saturated weight sat 13.5 13.5 18.0 18.0 kN/m3
Youngs modulus E 2667 2667 4000 40000 kN/m2
Poissons ratio 0.33 0.33 0.33 0.33
Cohesion c 8.0 8.0 3.0 3.0 kN/m2

Friction angle 20.0 20.0 30.0 32.0

Dilatancy angle 0.0 0.0 0.0 2.0
Permeability x-dir kx 1.0103 1.0 1.0 1.0 m/day
Permeability y-dir ky 1.0103 1.0 1.0 1.0 m/day
K0 Automatic Automatic Automatic Automatic
3
2 3
1
1
Figure 3: Geometry model with soil material sets

(1) Clay, (2) Retaining bank, (3) Fill and (4) Stiff layer
Geotextile
In the project database select the data type Geogrids and create a new material set. In
this material set, enter 2500 kN/m as stiffness. Note that this is the stiffness in extension.
In compression no stiffness is used.
Drag the geogrid data set to the geotextile in the geometry and drop it there. The
geotextile should flash red once, indicating the properties have been set.
Mesh generation
From the Mesh menu select the option Global coarseness. In the window that appears,
set the mesh coarseness to Medium and click on the Generate button, which will present
the following FE mesh composed of 15-node elements.
363 of 375
Figure 4: Medium coarse generated mesh
Select the clay layer (this consists of two clusters, see also hint) and press Refine cluster
from the Mesh menu. This will result in a refinement in the clay layer that will be needed
for the consolidation analysis. See figure 5.
Close the window showing the generated mesh and continue to the Calculations program.
Figure 5: Mesh with cluster refinement
364 of 375
CALCULATION
The calculation consists of two alternatives for the construction of the embankment: without
and with consolidation taken into account. After both alternatives the factor of safety is
determined. In the calculations list 8 phases are needed, 4 phases for each alternative. First
start with the fully undrained construction, that is without taking consolidation into account.
When starting Plaxis Calculations, choose Classical mode.
Initial conditions
Select the initial phase in the phase list and then press the Define button on the Para-
meters tabsheet in order to define the initial phase. The input window now opens in
Staged Construction mode.
Deselect all material clusters and geotextile elements that are not present at the start of
the analysis. As we want to model the entire construction sequence from the beginning,
switch off:
Geotextile elements
Material clusters for the fill
Material cluster for retaining bank
Now continue to the Water conditions mode by clicking the equally named button.
Enter a phreatic level at ground level by two coordinates (0, 5.5) and (33, 5.5). Click on
the Water pressures button to generate the pore pressures.
Phase 1: Excavation of the ditch and construction of the retaining bank

This calculation phase is a Plastic analysis, with loading type Staged construction. For all the
other settings the defaults should be used. In this phase:
Activate the full geotextile
Construct the retaining bank
Excavate the ditch (left of the embankment)
Phase 2: First fill

This calculation phase is also a Plastic analysis with the Staged construction loading
type. For all the other settings the defaults should be used. In this phase the first layer
of fill must be switched on.
365 of 375
Phase 3: Second fill

This calculation phase is again a Plastic analysis, loading type Staged construction. For
all the other settings the defaults should be used. Switch on the second layer of fill.
Phase 4: Safety factor determination

This calculation phase is a Safety phase. The loading type will be set automatically.
Keep all default settings.
After this, we will construct the embankment taking into account consolidation:
Phase 5: Consolidated construction of the ditch and retaining bank

This phase starts an alternative calculation, so phase 5 should NOT follow on phase 4 as is the
default, but it should follow on the initial phase. To do so, on the General tabsheet set Start
from phase to the Initial phase. This calculation phase is a Consolidation analysis, loading
type Staged construction. We assume that construction of the ditch and retaining bank will
take 3 days. Hence, in the Loading Input box fill in a Time Interval of 3 days. During this time
interval construction will take place, as well as consolidation. For all the other settings the
defaults should be used. In this phase again:
Switch on the full geotextile
Construct the retaining bank
Excavate the ditch (left of the embankment)
Phase 6: First fill - consolidated

This calculation phase is also a Consolidation analysis, loading type Staged construction. We
assume that making the hydraulic fill will take 7 days, so the Time interval should be set on 7
days. For the rest this phase is equal to phase 2; hence the first layer of fill must be switched
on.
Phase 7: Second fill - consolidated

This calculation phase is again a Consolidation analysis, loading type Staged construction.
This second fill will take 3 days, so the Time interval should be set on 3 days. For all the other
settings the defaults should be used. In staged construction, switch on the second layer of fill.
366 of 375
Phase 8: Safety factor determination

This calculation phase is a Safety phase. The loading type will be set automatically. Keep all
default settings.
Select points for load-displacement curves

As node for load-displacement curves, select the toe of the embankment and start the calcu-
lation.
367 of 375
INSPECT OUTPUT
In order to get a good idea of the displacement mechanism, one can view the contours
of incremental displacements. Figure 6 shows this plot of the final calculation step for the
undrained construction. It clearly shows the effect of the geotextile reinforcement. Figure 7
shows the incremental displacement for the consolidated construction. Here the embankment
has a more gradual settlement without showing an upcoming failure mechanism.
Figure 6: Incremental displacements contours, undrained (phase 3)
Figure 7: Incremental displacement contours, consolidated (phase 7)
The axial forces of the geotextile can be visualised by double clicking on the geotextile. This
will first present the displacement of the geotextile. On using the menu item Forces, one can
select Axial forces N.
Figure 8: Axial forces in geotextile, undrained (phase 3)
At the ends of the geotextile the axial force must be zero, but due to the discretisation and
some numerical inaccuracy this is not completely achieved. The maximum axial forces is
approx. 8 kN/m. figure 9 shows the axial forces for the consolidated construction. The
maximimum axial force here is only 5-6 kN/m.
Finally, the factors of safety are checked. In order to do so follow these steps:
Start the curves manager by selecting the Curves manager option from the Tools menu.
368 of 375
Figure 9: Axial forces in geotextile, consolidated (phase 7)
In the curves manager (see figure 10) select New in the Charts tabsheet. This presents
the Curve Generation window as shown in figure 11.
On the x-axis we want to show the displacements of the point at the toe of the embank-
ment, hence choose Point A and Deformations Total displacements |u|.
On the y-axis we want to show the strength reduction factor, hence select Project and
Multiplier M sf on the y-axis.
Figure 10: Curves manager
The created curve indicates a safety factor around 1.4 for the undrained construction and a a
safety factor of 2.1 for the consolidated construction of the embankment, as can be seen in
figure 12.
From the graph above, the factor of safety can be determined. Always look for a steady state
solution (slight variations in the load multipliers, increasing displacements). In most case, the
phi/c reduction calculation shows some variation at the beginning of the calculation. Note
that the displacements resulting from a Safety analysis are non-physical. Hence the total
displacements are not relevant. An incremental displacement plot of the last step, however,
shows the failure mechanism that corresponds the calculated value for M sf .
Addicionally, figures 13 and 14 show the failure mechanisms with the lowest factor or safety
for both the undrained and consolidated case.
369 of 375
Figure 11: Curve generation window
Consolidated: Msf=2.1
Undrained: Msf=1.4
Figure 12: Safety factor curve for reinforced embankment
370 of 375
Figure 13: Incremental displacements, undrained (phase 4)
Figure 14: Incremental displacements, consolidated (phase 8)
371 of 375
SUGGESTION FOR EXTRA EXERCISE: NON-REINFORCED

EMBANKMENT
For the undrained construction of an embankment, now introduce phase (9). In the Start
from phase list box select <0 initial phase>. This phase as well as phases 10 and 11
are Plastic analyses. Excavate the ditch and construct the embankment, but do NOT
activate the geotextile.
In the next phase (10) the first part of the fill is activated.
In the next phase (11) the second part of the fill is activated.
In the next phase (12) perform a safety analysis. In principle we can keep the 100
additional steps for this calculation. However, 50 additional steps is already sufficient
here.
For the consolidated construction of the embankment, now introduce phase (13). In the
Start from phase list box select <0 initial phase>. This phase as well as phases 14 and
15 are Consolidation analyses. Set the Time interval to 3 days, excavate the ditch and
construct the embankment, but do NOT activate the geotextile.
In the next phase (14) the first part of the fill is activated. Set the Time interval to 7 days.
In the next phase (15) the second part of the fill is activated. Set the Time interval to 3
days.
Finally, in the last phase (16) perform a Safety analysis again. In principle we can keep
the 100 additional steps here as well. However, 30 additional steps is already sufficient
to obtain a reliable value.
Presented below is both the incremental displacement plot as well as the incremental shear
strain plot of both the drained and consolidated non-reinforced embankment after safety ana-
lysis. Hence, the plots show the failure mechanisms.
Figure 15: Incremental displacements, undrained (phase 12)
FACTORS OF SAFETY
The factors of safety are checked with the Curves program, see figure 19.
372 of 375
Figure 16: Incremental shear strains, undrained (phase 12)
Figure 17: Incremental displacements, consolidated (phase 16)
Figure 18: Incremental shear strains, consolidated (phase 16)
Consolidated: Msf=1.4
Undrained: Msf=1.1
Figure 19: Safety factor curve for non-reinforced embankment
373 of 375
374 of 375
SUGGESTIONS FOR THE DETERMINATION OF THE CLAY

PARAMETERS
qc
su 15 = 150
15
= 10 kPa
su = 12 x,0 + ,0
0 ,0 ,0
y sin() + c cos() with x = K0 y (1 sin()) y
In the middle of the clay layer at about 2m below ground level:
y,0 = h (sat water ) = 2 3.5 = 7 kPa = x,0 = (1 sin(20)) y,0 = 4.6 kPa
For this clay estimate = 20, then c 8 kPa
Eu 15000s
50
u
= 1500010
50
= 3000 kPa
1 1
G = 3 Eu = 3 3000 = 1000 kPa
E 0 = 2G(1 + ) = 83 G = 2667 kPa
K0
= 1+K 0
= 0.5
1.5
= 0.33
Use Undrained A as the type of material behaviour.
375 of 375

Plaxis Introductory Course, Danang, Vietnam

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

Plaxis Introductory Course, Danang, Vietnam

Uploaded by

Copyright:

Available Formats

PLAXIS INTRODUCTORY COURSE

10-12 JULY 2013

Day 1: Wednesday 10.7.2013

10:30 12:00 Exercise 1 Simple Foundation on Elastoplastic Soil Dr Cheang

1:30 2:15 Lecture 3 Non-linear Computation Dr Phung

3:15 4:45 Exercise 2 Simulation of Laboratory Tests Dr Cheang

Day 2: Thursday 11.7.2013

10:30 12:00 Exercise 3 Modelling of Anchored Excavation Dr Phung

1:30 2:15 Lecture 7 Drained and Undrained Analysis Dr Cheang

3:15 4:45 Exercise 4 Dewatering of Excavation Dr Phung

Day 3: Friday 12.7.2013

1:30 2:15 Lecture 11 Overview of Soil Models Dr Phung

3:15 4:45 Exercise 6 Geotextile Reinforced Embankment & Consolidation Dr Phung

Exercise Exercise2:SimulationofTriaxialandOedometerTests 124

Lecture5 Geometry&MeshSelection 169

Lecture6 StructuralElementsinPlaxis2D 186

Exercise Exercise3:AnchoredExcavation 199

Lecture7 UndrainedAnalysisinPlaxis 222

Lecture8 ModellingofGroundwaterinPlaxis 238

Exercise Exercise4:ExcavationandDewatering 264

Lecture9 InitialStresses 276

Lecture10 FactorofSafetyAnalysisviaPhi=C'Reduction 282

Exercise Exercise5:StabilityAnalysisofASlopeStabilisedbySoilNails 310

Lecture11 OverviewofSoilModels 323

Lecture12 ConsolidationAnalysis 341

Exercise Exercise6:GeotextileReinforcedEmbankmentwithConsolidation 357

Plaxis Introductory Course, Danang, Vietnam

CG1: GEOTECHNICAL FINITE ELEMENT ANALYSIS

Technical University of Catalunya, Barcelona

some off the

Stability (local and general)

Admissible deformation and displacements

geotechnical analysis: basic solution requirements

Potts & Zdravkovic

methods for numerical analysis

geotechnical finite element analysis

Objectives of the numerical (finite element) analysis

Advantages of numerical (finite element) analysis

geotechnical finite element analysis

Some requirements for successful numerical modelling

2 the quality of a tool is important

3. the design process involves considerably more than analysis

Borrowed from C. Viggiani, with thanks

introduction: the Finite Element Method

the FEM is a computational procedure that may be used to obtain an

the governing mathematical equations are approximated by a series of

this lecture presents only a basic outline of the method

the first stage in any FE analysis is to generate a FE mesh

a mesh consists of elements connected together at nodes

introduction: the Finite Elements Method

the nodes are the points where values Footing

the values of nodal displacements are

and strain throughout the complete mesh

a constitutive law is then used to relate strains to stresses and this

the nodal forces are related to the nodal displacements by equations

two-dimensional analysis of continua is generally based on the use of

the most used elements are based on an iso-parametric approach

primary unknowns: values of the nodal displacements

how does it work? p

1 2 12 coefficients, depending on the values of the 12 nodal

constitutive relation (elasticity)

elasticity: one-to-one relationship between stress and strain